diff --git a/data/sampled_jsons/1e-4_OR_0.0001_EntityErasure_AECM_learning_rate_implementation.jsonl b/data/sampled_jsons/1e-4_OR_0.0001_EntityErasure_AECM_learning_rate_implementation.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..950f95d4dc99a4eda85a3a441c29bb5ca625f1fd
--- /dev/null
+++ b/data/sampled_jsons/1e-4_OR_0.0001_EntityErasure_AECM_learning_rate_implementation.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Learning rate annealing with tune.run() - RLlib - Ray", "date": "", "ddg_snippet": "Is there a way to set learning rate annealing with tune.run()? The grid search below will run two training: one LR at 1 e -5 and second at 1 e -6. How do I, for example, have a LR schedule where at the end of 1000 iterations is LR is reduced from 1 e -5 to 1 e -6 and from 1 e -6 to 1 e -7?", "subpage_snippet": "", "source": "discuss.ray.io", "link": "https://discuss.ray.io/t/learning-rate-annealing-with-tune-run/1928", "content": "Is there a way to set learning rate annealing with tune.run()? The grid search below will run two training: one LR at 1 e -5 and second at 1 e -6. How do I, for example, have a LR schedule where at the end of 1000 iterations is LR is reduced from 1 e -5 to 1 e -6 and from 1 e -6 to 1 e -7?"}
+{"idx": 1, "title": "An Analysis of Batch Size vs. Learning Rate on Imagenette – Vishal...", "date": "", "ddg_snippet": "With a batch size of 64 and three different LRs ( 1 e - 4 , 1 e -3, 1 e -2) I achieved the following results using three models (xresnet34, xse_resnext50, xresnet18)The best training run goes to xresnet34 with a batch size of 16 and a learning rate of 1 e -3.", "subpage_snippet": "", "source": "vishalbakshi.github.io", "link": "https://vishalbakshi.github.io/blog/posts/2025-06-18-imagenette/", "content": "With a batch size of 64 and three different LRs ( 1 e - 4 , 1 e -3, 1 e -2) I achieved the following results using three models (xresnet34, xse_resnext50, xresnet18)The best training run goes to xresnet34 with a batch size of 16 and a learning rate of 1 e -3."}
+{"idx": 2, "title": "[Bug]: Learning Rate setting in the GUI does not change if one does...", "date": "", "ddg_snippet": "When I type E - 4 in the Learning Rate field, it does not change it, preserving the previous value (e.g. 8E-5). So you start the training, but it doesn't do as expected.Either it accepts it as 1 E - 4 or 0 . 0001 , or at least shows a validation error explaining the value is not acceptable.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/Nerogar/OneTrainer/issues/836", "content": "When I type E - 4 in the Learning Rate field, it does not change it, preserving the previous value (e.g. 8E-5). So you start the training, but it doesn't do as expected.Either it accepts it as 1 E - 4 or 0 . 0001 , or at least shows a validation error explaining the value is not acceptable."}
+{"idx": 3, "title": "Strange learning rate finder plot for custom model - Part 1 (2019)...", "date": "", "ddg_snippet": "learn = learn .split([[ learn .model.frontal_cnn, learn .model.lateral_cnn], learn .model.joined_cnn]). Afterwards I ran the following code to find a good learning rate for my modelIn these cases, I would rather suggest that 1 e -2 and 1 e - 4 are the best learning rates , respectively.", "subpage_snippet": "", "source": "forums.fast.ai", "link": "https://forums.fast.ai/t/strange-learning-rate-finder-plot-for-custom-model/45441", "content": "learn = learn .split([[ learn .model.frontal_cnn, learn .model.lateral_cnn], learn .model.joined_cnn]). Afterwards I ran the following code to find a good learning rate for my modelIn these cases, I would rather suggest that 1 e -2 and 1 e - 4 are the best learning rates , respectively."}
+{"idx": 4, "title": "Fine-Tuning T5 With Hugging Face | Restackio", "date": "", "ddg_snippet": "Learning rates of 1 e - 4 and 3 e - 4 are generally effective for tasks such as classification, summarization, translation, and question answering.", "subpage_snippet": "", "source": "d2wozrt205r2fu.cloudfront.net", "link": "https://d2wozrt205r2fu.cloudfront.net/p/fine-tuning-answer-t5-hugging-face-cat-ai", "content": "Learning rates of 1 e - 4 and 3 e - 4 are generally effective for tasks such as classification, summarization, translation, and question answering."}
+{"idx": 5, "title": "Example Usage | runpod-workers/worker-lora_trainer | DeepWiki", "date": "", "ddg_snippet": "\"lr_scheduler_num_cycles\": 1, \" learning _ rate \": 0 . 0001 lr_scheduler: LR decay pattern. network_dim: LoRA rank dimension. unet_lr: UNet learning rate . save_precision: Model storage format.", "subpage_snippet": "", "source": "deepwiki.com", "link": "https://deepwiki.com/runpod-workers/worker-lora_trainer/5.2-example-usage", "content": "\"lr_scheduler_num_cycles\": 1, \" learning _ rate \": 0 . 0001 lr_scheduler: LR decay pattern. network_dim: LoRA rank dimension. unet_lr: UNet learning rate . save_precision: Model storage format."}
+{"idx": 6, "title": "Perfect LoRA Training parameters human character - Models - Hugging...", "date": "", "ddg_snippet": "--text_encoder_lr= 0 . 0001 : Slightly lower learning rate for text encoder stability.-- learning _ rate = 1 e - 4 : Lower learning rate to prevent overfitting. --max_train_steps=1500: Sufficient steps for limited data without overtraining.", "subpage_snippet": "", "source": "discuss.huggingface.co", "link": "https://discuss.huggingface.co/t/perfect-lora-training-parameters-human-character/147211", "content": "--text_encoder_lr= 0 . 0001 : Slightly lower learning rate for text encoder stability.-- learning _ rate = 1 e - 4 : Lower learning rate to prevent overfitting. --max_train_steps=1500: Sufficient steps for limited data without overtraining."}
+{"idx": 7, "title": "Songbird Loss function plot looks like (a bad) EKG... - QIIME 2 Forum", "date": "", "ddg_snippet": "If you don't want a noisy loss, the best approach is to lower the learning rate --p- learning - rate . The default is now 1 e -3, you can try 1 e - 4 or 1 e -5 to see if that will give you a cleaner loss (although you will need to train for more epochs). 1 Like.convergence-plot_LR- 0 . 0001 -1.", "subpage_snippet": "", "source": "forum.qiime2.org", "link": "https://forum.qiime2.org/t/songbird-loss-function-plot-looks-like-a-bad-ekg/14329", "content": "If you don't want a noisy loss, the best approach is to lower the learning rate --p- learning - rate . The default is now 1 e -3, you can try 1 e - 4 or 1 e -5 to see if that will give you a cleaner loss (although you will need to train for more epochs). 1 Like.convergence-plot_LR- 0 . 0001 -1."}
+{"idx": 8, "title": "StephanHeijl.com - Notes on BERT: Pre-training of Deep Bidirectional...", "date": "", "ddg_snippet": "Adam settings. Learning rate . 1 e - 4 .The pretraining learning rate is set to 1 e - 4 , not an uncommon learning rate for Adam. The first 10.000 steps are subject to learning rate warm-up, where the lr is linearly increased from 0 to the target.", "subpage_snippet": "", "source": "stephanheijl.com", "link": "https://stephanheijl.com/notes_on_bert.html", "content": "Adam settings. Learning rate . 1 e - 4 .The pretraining learning rate is set to 1 e - 4 , not an uncommon learning rate for Adam. The first 10.000 steps are subject to learning rate warm-up, where the lr is linearly increased from 0 to the target."}
+{"idx": 9, "title": "Simple Linear Regression not converging – Python", "date": "", "ddg_snippet": "Your code seems actually to work fine; except for learning rate , really! Just reduce it from 0.01 to e.g. 0 . 0001 and everything works fine (well, I would also reduce tolerance to something much much smaller, like 1 e -5, to make sure it actually converges to the right solution).", "subpage_snippet": "", "source": "python.tutorialink.com", "link": "https://python.tutorialink.com/simple-linear-regression-not-converging/", "content": "Your code seems actually to work fine; except for learning rate , really! Just reduce it from 0.01 to e.g. 0 . 0001 and everything works fine (well, I would also reduce tolerance to something much much smaller, like 1 e -5, to make sure it actually converges to the right solution)."}
diff --git a/data/sampled_jsons/35068_XLRS-Bench-_Could_Your_Multimodal_LLMs_Understand_Extremely_Large_Ultra-High-Resolution_Remote.jsonl b/data/sampled_jsons/35068_XLRS-Bench-_Could_Your_Multimodal_LLMs_Understand_Extremely_Large_Ultra-High-Resolution_Remote.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8366b4edba9e9225adf35c800fd51c0192f2a692
--- /dev/null
+++ b/data/sampled_jsons/35068_XLRS-Bench-_Could_Your_Multimodal_LLMs_Understand_Extremely_Large_Ultra-High-Resolution_Remote.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "XLRS-Bench: Could Your Multimodal LLMs Understand Extremely ... GitHub - AI9Stars/XLRS-Bench: [CVPR 2025 HIghlight] XLRS ... CVPR 2025 Open Access Repository XLRS-Bench XLRS-Bench|遥感数据集|图像分析数据集 arXiv:2503.23771v1 [cs.CV] 31 Mar 2025 XLRS-Bench: Could Your Multimodal LLMs Understand Extremely ...", "date": "", "ddg_snippet": "Mar 31, 2025 · View a PDF of the paper titled XLRS - Bench : Could Your Multimodal LLMs Understand Extremely Large Ultra - High - Resolution Remote Sensing Imagery ?, by Fengxiang Wang and 11 other authors Apr 1, 2025 · We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far. To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios. We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far. Apr 1, 2025 · XL RS - Ben ch 数据集由国防科技大学、清华大学、武汉大学等机构的联合研究团队于2025年推出，旨在评估多模态大语言模型（ ML LM s ）在 超高 分辨率 遥感影像中的感知与推理能力。 To address these issues, we present XLRS - Bench : a com- prehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios. To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.23771", "content": "Mar 31, 2025 · View a PDF of the paper titled XLRS - Bench : Could Your Multimodal LLMs Understand Extremely Large Ultra - High - Resolution Remote Sensing Imagery ?, by Fengxiang Wang and 11 other authors Apr 1, 2025 · We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far. To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios. We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far. Apr 1, 2025 · XL RS - Ben ch 数据集由国防科技大学、清华大学、武汉大学等机构的联合研究团队于2025年推出，旨在评估多模态大语言模型（ ML LM s ）在 超高 分辨率 遥感影像中的感知与推理能力。 To address these issues, we present XLRS - Bench : a com- prehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios. To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios."}
+{"idx": 1, "title": "GitHub - AI9Stars/XLRS-Bench: [CVPR 2025 HIghlight] XLRS ...", "date": "", "ddg_snippet": "Apr 1, 2025 · We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/AI9Stars/XLRS-Bench", "content": "Apr 1, 2025 · We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far."}
+{"idx": 2, "title": "XLRS-Bench", "date": "", "ddg_snippet": "We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far.", "subpage_snippet": "", "source": "xlrs-bench.github.io", "link": "https://xlrs-bench.github.io/home_page.html", "content": "We present XLRS - Bench , a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios, featuring the largest average image size of 8,500 × 8,500 observed thus far."}
+{"idx": 3, "title": "XLRS-Bench: Could Your Multimodal LLMs Understand Extremely ...", "date": "", "ddg_snippet": "To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/2503.23771", "content": "To address these issues, we present XLRS - Bench : a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in ultra - high - resolution RS scenarios."}
+{"idx": 4, "title": "( PDF ) XLRS - Bench : Could Your Multimodal LLMs Understand ...", "date": "", "ddg_snippet": "XLRS - Bench focuses on extremely large ultra - high - resolution RS imagery , integrating. over 10 multimodal vision-language perception and reasoning tasks within the same image .", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/390354847_XLRS-Bench_Could_Your_Multimodal_LLMs_Understand_Extremely_Large_Ultra-High-Resolution_Remote_Sensing_Imagery", "content": "XLRS - Bench focuses on extremely large ultra - high - resolution RS imagery , integrating. over 10 multimodal vision-language perception and reasoning tasks within the same image ."}
+{"idx": 5, "title": "XLRS - Bench : Could Your Multimodal LLMs Understand Extremely ...", "date": "", "ddg_snippet": "XLRS - Bench focuses on extremely large ultra - high - resolution RS imagery , integrating over 10 multimodal vision-language perception and reasoning tasks within the same image .", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Wang_XLRS-Bench_Could_Your_Multimodal_LLMs_Understand_Extremely_Large_Ultra-High-Resolution_Remote_CVPR_2025_paper.pdf", "content": "XLRS - Bench focuses on extremely large ultra - high - resolution RS imagery , integrating over 10 multimodal vision-language perception and reasoning tasks within the same image ."}
+{"idx": 6, "title": "XLRS - Bench : Could Your Multimodal LLMs Understand Extremely ...", "date": "", "ddg_snippet": "XLRS - Bench boasts the largest average image size (8500$\\times$8500) observed thus far, with all evaluation samples meticulously annotated manually, assisted by a novel semi-automatic captioner on ultra - high - resolution RS images .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=ThgEfn5CSU&referrer=[the+profile+of+Maosong+Sun](/profile?id=~Maosong_Sun1)", "content": "XLRS - Bench boasts the largest average image size (8500$\\times$8500) observed thus far, with all evaluation samples meticulously annotated manually, assisted by a novel semi-automatic captioner on ultra - high - resolution RS images ."}
+{"idx": 7, "title": "CVPR.2025 - Highlight | Cool Papers - Immersive Paper Discovery", "date": "", "ddg_snippet": "Large -scale multimodal representation learning successfully optimizes for zero-shot transfer at test time. Yet the standard pretraining paradigm (contrastive learning on large amounts of image -text data) does not explicitly encourage representations to support few-shot adaptation.", "subpage_snippet": "", "source": "papers.cool", "link": "https://papers.cool/venue/CVPR.2025?group=Highlight", "content": "Large -scale multimodal representation learning successfully optimizes for zero-shot transfer at test time. Yet the standard pretraining paradigm (contrastive learning on large amounts of image -text data) does not explicitly encourage representations to support few-shot adaptation."}
+{"idx": 8, "title": "【论文翻】 XLRS - Bench : Could Your Multimodal LLMs Understand ...", "date": "", "ddg_snippet": "XLRS - Bench stands out from existing multimodal understanding benchmarks with four key features: i) the largest average image size of 8,500×8,500, featuring extensive 10,000×10,000 imagery ; ii) 16 sub-tasks designed to evaluate MLLMs’ capabilities in...", "subpage_snippet": "", "source": "blog.csdn.net", "link": "https://blog.csdn.net/l963852k/article/details/149417749", "content": "XLRS - Bench stands out from existing multimodal understanding benchmarks with four key features: i) the largest average image size of 8,500×8,500, featuring extensive 10,000×10,000 imagery ; ii) 16 sub-tasks designed to evaluate MLLMs’ capabilities in..."}
+{"idx": 9, "title": "Paper Digest: CVPR 2025 Papers & Highlights – Paper Digest", "date": "", "ddg_snippet": "XLRS - Bench : Could Your Multimodal LLMs Understand Extremely Large Ultra - High - Resolution Remote Sensing Imagery ?", "subpage_snippet": "", "source": "www.paperdigest.org", "link": "https://www.paperdigest.org/2025/06/cvpr-2025-papers-highlights/", "content": "XLRS - Bench : Could Your Multimodal LLMs Understand Extremely Large Ultra - High - Resolution Remote Sensing Imagery ?"}
diff --git a/data/sampled_jsons/35068_XLRS-Bench_Table_2_Qwen2-VL_scores_year_2024.jsonl b/data/sampled_jsons/35068_XLRS-Bench_Table_2_Qwen2-VL_scores_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..780748d7f0ffa92b9e7d49d3bdb0b5dd8b1fe802
--- /dev/null
+++ b/data/sampled_jsons/35068_XLRS-Bench_Table_2_Qwen2-VL_scores_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Qwen - Wikipedia", "date": "", "ddg_snippet": "Toggle the table of contents. Qwen. 25 languages.The Qwen-VL series is a line of visual language models that combines a vision transformer with a LLM.[3][14] Alibaba released Qwen 2 - VL with variants of 2 billion and 7 billion parameters.[15][16][17].", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Qwen", "content": "Toggle the table of contents. Qwen. 25 languages.The Qwen-VL series is a line of visual language models that combines a vision transformer with a LLM.[3][14] Alibaba released Qwen 2 - VL with variants of 2 billion and 7 billion parameters.[15][16][17]."}
+{"idx": 1, "title": "microsoft/GUI-Actor-7B- Qwen 2 - VL · Hugging Face", "date": "", "ddg_snippet": "Table 1. Main results on ScreenSpot-Pro, ScreenSpot, and ScreenSpot-v 2 with Qwen 2 - VL as the backbone. † indicates scores obtained from our own evaluation of the official models on Huggingface.", "subpage_snippet": "", "source": "huggingface.1319lm.top", "link": "https://huggingface.1319lm.top/microsoft/GUI-Actor-7B-Qwen2-VL", "content": "Table 1. Main results on ScreenSpot-Pro, ScreenSpot, and ScreenSpot-v 2 with Qwen 2 - VL as the backbone. † indicates scores obtained from our own evaluation of the official models on Huggingface."}
+{"idx": 2, "title": "Qwen 2 .5 VL ! Qwen 2 .5 VL ! Qwen 2 .5 VL ! | Qwen", "date": "", "ddg_snippet": "Qwen 2 .5- VL has significantly enhanced its general image recognition capabilities, expanding the categories of images to an ultra-large number.Watch the video and list the paper titles in a table . Qwen 2 .5- VL .", "subpage_snippet": "", "source": "qwenlm.github.io", "link": "https://qwenlm.github.io/blog/qwen2.5-vl/", "content": "Qwen 2 .5- VL has significantly enhanced its general image recognition capabilities, expanding the categories of images to an ultra-large number.Watch the video and list the paper titles in a table . Qwen 2 .5- VL ."}
+{"idx": 3, "title": "Qwen3-Max 2025 Complete Release Analysis: In-Depth Review of...", "date": "", "ddg_snippet": "Enhanced Coding Capabilities: SWE- Bench Verified score of 69.6, significantly improved agent capabilities. Thinking Version: Qwen3-Max-Thinking achieves 100% accuracy on AIME25, HMMT and other mathematical reasoning benchmarks.", "subpage_snippet": "", "source": "qwq32.com", "link": "https://qwq32.com/blog/2025-qwen3-max", "content": "Enhanced Coding Capabilities: SWE- Bench Verified score of 69.6, significantly improved agent capabilities. Thinking Version: Qwen3-Max-Thinking achieves 100% accuracy on AIME25, HMMT and other mathematical reasoning benchmarks."}
+{"idx": 4, "title": "Qwen3-Next-80B-A3B vs Qwen3-235B-A22B-2507:Why Small Excels", "date": "", "ddg_snippet": "Table Of ContentsQwen3-Next-80B vs Qwen3-235B：Key Differences in ArchitecturesTakeaway: Qwen3-Next-80B-A3B is designed for efficiency without sacrificing much performance.", "subpage_snippet": "", "source": "blogs.novita.ai", "link": "https://blogs.novita.ai/qwen3-next-80b-a3b-vs-qwen3-235b-a22b-2507why-small-excels/", "content": "Table Of ContentsQwen3-Next-80B vs Qwen3-235B：Key Differences in ArchitecturesTakeaway: Qwen3-Next-80B-A3B is designed for efficiency without sacrificing much performance."}
+{"idx": 5, "title": "GitHub - QwenLM/ Qwen 2 .5- VL : Qwen 2 .5- VL is the multimodal large...", "date": "", "ddg_snippet": "In the past five months since Qwen 2 - VL 's release, numerous developers have built new models on the Qwen 2 - VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/QwenLM/Qwen2.5-VL", "content": "In the past five months since Qwen 2 - VL 's release, numerous developers have built new models on the Qwen 2 - VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models."}
+{"idx": 6, "title": "Qwen/ Qwen 2 - VL - 2 B-Instruct · Hugging Face", "date": "", "ddg_snippet": "Naive Dynamic Resolution: Unlike before, Qwen 2 - VL can handle arbitrary image resolutions, mapping them into a dynamic number of visual tokens, offering a more human-like visual processing experience.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct", "content": "Naive Dynamic Resolution: Unlike before, Qwen 2 - VL can handle arbitrary image resolutions, mapping them into a dynamic number of visual tokens, offering a more human-like visual processing experience."}
+{"idx": 7, "title": "qwen/ qwen 2 .5- vl -7b • LM Studio", "date": "", "ddg_snippet": "Qwen 2 .5- VL -7B-Instruct. Qwen 2 .5- VL -7B-Instruct is a vision-language model that processes images, text, and video, supporting structured outputs and visual localization. It can analyze charts, graphics, and layouts, and is capable of temporal reasoning over long video sequences.", "subpage_snippet": "", "source": "lmstudio.ai", "link": "https://lmstudio.ai/models/qwen/qwen2.5-vl-7b", "content": "Qwen 2 .5- VL -7B-Instruct. Qwen 2 .5- VL -7B-Instruct is a vision-language model that processes images, text, and video, supporting structured outputs and visual localization. It can analyze charts, graphics, and layouts, and is capable of temporal reasoning over long video sequences."}
+{"idx": 8, "title": "togetherai/ qwen 2 - vl -72b-instruct", "date": "", "ddg_snippet": "togetherai. qwen 2 - vl -72b-instruct.Preview. YAML. together Qwen 2 - VL (72B) Instruct model icon.", "subpage_snippet": "", "source": "hub.continue.dev", "link": "https://hub.continue.dev/togetherai/qwen2-vl-72b-instruct", "content": "togetherai. qwen 2 - vl -72b-instruct.Preview. YAML. together Qwen 2 - VL (72B) Instruct model icon."}
+{"idx": 9, "title": "Qwen", "date": "", "ddg_snippet": "Qwen3-Max-Preview. Скачать приложение.Задайте вопрос Qwen, узнайте больше.", "subpage_snippet": "", "source": "chat.qwen.ai", "link": "https://chat.qwen.ai/", "content": "Qwen3-Max-Preview. Скачать приложение.Задайте вопрос Qwen, узнайте больше."}
diff --git a/data/sampled_jsons/6yBhoJn6qy_Causal_Modeling_of_Climate_Activism_on_Reddit_Related_Work_section_events.jsonl b/data/sampled_jsons/6yBhoJn6qy_Causal_Modeling_of_Climate_Activism_on_Reddit_Related_Work_section_events.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3bfbdc5991ac635014cbf1c5c4507ef5716bd2f6
--- /dev/null
+++ b/data/sampled_jsons/6yBhoJn6qy_Causal_Modeling_of_Climate_Activism_on_Reddit_Related_Work_section_events.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Causal Modeling of Climate Activism on Reddit - OpenReview", "date": "", "ddg_snippet": "In this 16 work , we develop a comprehensive causal model of how and why 17 Reddit users engage with activist communities driving mass climate 18 protests (mainly the 2019 Earth Strike, Fridays for Future, and Ex- 19 tinction Rebellion). Our framework, based on Stochastic Variational 20 Inference applied to Bayesian Networks, learns the causal ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=6yBhoJn6qy", "content": "In this 16 work , we develop a comprehensive causal model of how and why 17 Reddit users engage with activist communities driving mass climate 18 protests (mainly the 2019 Earth Strike, Fridays for Future, and Ex- 19 tinction Rebellion). Our framework, based on Stochastic Variational 20 Inference applied to Bayesian Networks, learns the causal ..."}
+{"idx": 1, "title": "[2410.10562] Causal Modeling of Climate Activism on Reddit", "date": "", "ddg_snippet": "Climate activism is crucial in stimulating collective societal and behavioral change towards sustainable practices through political pressure. Although multiple factors contribute to the participation in activism , their complex relationships and the scarcity of data on their interactions have restricted most prior research to studying them in isolation, thus preventing the development of a ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.10562", "content": "Climate activism is crucial in stimulating collective societal and behavioral change towards sustainable practices through political pressure. Although multiple factors contribute to the participation in activism , their complex relationships and the scarcity of data on their interactions have restricted most prior research to studying them in isolation, thus preventing the development of a ..."}
+{"idx": 2, "title": "Causal Modeling of Climate Activism on Reddit | Article Information | J ...", "date": "", "ddg_snippet": "Article \"Causal Modeling of Climate Activism on Reddit \" Detailed information of the J-GLOBAL is an information service managed by the Japan Science and Technology Agency (hereinafter referred to as \"JST\"). It provides free access to secondary information on researchers, articles, patents, etc., in science and technology, medicine and pharmacy. The search results guide you to high-quality ...", "subpage_snippet": "", "source": "jglobal.jst.go.jp", "link": "https://jglobal.jst.go.jp/en/detail?JGLOBAL_ID=202402219833468228", "content": "Article \"Causal Modeling of Climate Activism on Reddit \" Detailed information of the J-GLOBAL is an information service managed by the Japan Science and Technology Agency (hereinafter referred to as \"JST\"). It provides free access to secondary information on researchers, articles, patents, etc., in science and technology, medicine and pharmacy. The search results guide you to high-quality ..."}
+{"idx": 3, "title": "Causal Modeling of Climate Activism on Reddit - Researchr", "date": "", "ddg_snippet": "Causal Modeling of Climate Activism on Reddit . In Guodong Long, Michale Blumestein, Yi Chang 0001, Liane Lewin-Eytan, Zi Helen Huang, Elad Yom-Tov, editors, Proceedings of the ACM on Web Conference 2025, WWW 2025, Sydney, NSW, Australia, 28 April 2025- 2 May 2025. pages 590-600, ACM, 2025. [doi]", "subpage_snippet": "", "source": "researchr.org", "link": "https://researchr.org/publication/LentiAMM25", "content": "Causal Modeling of Climate Activism on Reddit . In Guodong Long, Michale Blumestein, Yi Chang 0001, Liane Lewin-Eytan, Zi Helen Huang, Elad Yom-Tov, editors, Proceedings of the ACM on Web Conference 2025, WWW 2025, Sydney, NSW, Australia, 28 April 2025- 2 May 2025. pages 590-600, ACM, 2025. [doi]"}
+{"idx": 4, "title": "Causal Modeling of Climate Activism on Reddit", "date": "", "ddg_snippet": "In this work , we develop a comprehensive causal model of how and why Reddit users engage with activist communities driving mass climate protests (mainly the 2019 Earth Strike, Fridays for Future, and Extinction Rebellion).", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1145/3696410.3714684", "content": "In this work , we develop a comprehensive causal model of how and why Reddit users engage with activist communities driving mass climate protests (mainly the 2019 Earth Strike, Fridays for Future, and Extinction Rebellion)."}
+{"idx": 5, "title": "Causal Modeling of Climate Activism on Reddit - arXiv.org", "date": "", "ddg_snippet": "We developed a rich and comprehensive causal model to study the interplay between different determinants of climate activism on Reddit . This work represents a first attempt to apply a multi-causal model to social media data.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.10562v1", "content": "We developed a rich and comprehensive causal model to study the interplay between different determinants of climate activism on Reddit . This work represents a first attempt to apply a multi-causal model to social media data."}
+{"idx": 6, "title": "Causal networks for climate model evaluation and constrained ...", "date": "", "ddg_snippet": "Here we apply causal discovery algorithms to sea level pressure data from a large set of climate model simulations and, as a proxy for observations, meteorological reanalyses.", "subpage_snippet": "", "source": "www.nature.com", "link": "https://www.nature.com/articles/s41467-020-15195-y", "content": "Here we apply causal discovery algorithms to sea level pressure data from a large set of climate model simulations and, as a proxy for observations, meteorological reanalyses."}
+{"idx": 7, "title": "PDF Spatial-temporal Causal Modeling for Climate Change Attribution", "date": "", "ddg_snippet": "In the present work , we employ a relatively simple approach to combining the two: using our Bayesian hierarchical model we estimate the N-year event magnitudes associated with the climate metrics of interest, and we incorporate these estimated variables as additional variables in causal modeling and attribution in the spatio-temporal modeling ...", "subpage_snippet": "", "source": "www.niculescu-mizil.org", "link": "http://www.niculescu-mizil.org/papers/KDD09Climate-final.pdf", "content": "In the present work , we employ a relatively simple approach to combining the two: using our Bayesian hierarchical model we estimate the N-year event magnitudes associated with the climate metrics of interest, and we incorporate these estimated variables as additional variables in causal modeling and attribution in the spatio-temporal modeling ..."}
+{"idx": 8, "title": "Causal Modeling of Climate Activism on Reddit - Semantic Scholar", "date": "", "ddg_snippet": "Climate activism is crucial in stimulating collective societal and behavioral change towards sustainable practices through political pressure. Although multiple factors contribute to the participation in activism , their complex relationships and the scarcity of data on their interactions have restricted most prior research to studying them in isolation, thus preventing the development of a ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/Causal-Modeling-of-Climate-Activism-on-Reddit-Lenti-Aiello/c4c7c3972ba102db37738c082f27ebfcd3983057", "content": "Climate activism is crucial in stimulating collective societal and behavioral change towards sustainable practices through political pressure. Although multiple factors contribute to the participation in activism , their complex relationships and the scarcity of data on their interactions have restricted most prior research to studying them in isolation, thus preventing the development of a ..."}
+{"idx": 9, "title": "PDF A methodological approach for inferring causal relationships from ...", "date": "", "ddg_snippet": "A study closely related to our work on extracting causal relationships from climate change social media data is proposed by Lenti et al. (2024). Their approach analyzes Reddit data from 2016 to 2022 using stochastic variational inference and Bayesian networks to examine the impact of sociodemographics, climate news exposure, extreme weather ...", "subpage_snippet": "", "source": "peerj.com", "link": "https://peerj.com/articles/cs-2964.pdf", "content": "A study closely related to our work on extracting causal relationships from climate change social media data is proposed by Lenti et al. (2024). Their approach analyzes Reddit data from 2016 to 2022 using stochastic variational inference and Bayesian networks to examine the impact of sociodemographics, climate news exposure, extreme weather ..."}
diff --git a/data/sampled_jsons/American_Political_Science_Review_journal_domain.jsonl b/data/sampled_jsons/American_Political_Science_Review_journal_domain.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..38f494136dc9b2f88a90f029aeb0dc99562d42fd
--- /dev/null
+++ b/data/sampled_jsons/American_Political_Science_Review_journal_domain.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "American Political Science Review", "date": "", "ddg_snippet": "Scope. American Political Science Review is political science's premier scholarly research journal, providing peer-reviewed articles and review essays from ...", "subpage_snippet": "", "source": "www.scimagojr.com", "link": "https://www.scimagojr.com/journalsearch.php?q=15557&tip=sid", "content": "Scope. American Political Science Review is political science's premier scholarly research journal, providing peer-reviewed articles and review essays from ..."}
+{"idx": 1, "title": "FirstView articles | American Political Science Review", "date": "", "ddg_snippet": "Journals · American Political Science Review ; FirstView articles. English; Français ... We illustrate this in the domains of policing and gender, examining ...", "subpage_snippet": "", "source": "www.cambridge.org", "link": "https://www.cambridge.org/core/journals/american-political-science-review/firstview", "content": "Journals · American Political Science Review ; FirstView articles. English; Français ... We illustrate this in the domains of policing and gender, examining ..."}
+{"idx": 2, "title": "Vol. 22, No. 1, Feb., 1928 of The American Political ...", "date": "", "ddg_snippet": "The American Political Science Review (APSR) is the longest runningpublication of the American Political Science Association (APSA). APSR,first published in ...", "subpage_snippet": "", "source": "www.jstor.org", "link": "https://www.jstor.org/stable/i333476", "content": "The American Political Science Review (APSR) is the longest runningpublication of the American Political Science Association (APSA). APSR,first published in ..."}
+{"idx": 3, "title": "Vol. 11, No. 1, Feb., 1917 of The American Political ...", "date": "", "ddg_snippet": "The American Political Science Review (APSR ) is the longest runningpublication of the American Political Science Association (APSA). APSR,first published in ...", "subpage_snippet": "", "source": "www.jstor.org", "link": "https://www.jstor.org/stable/i306597", "content": "The American Political Science Review (APSR ) is the longest runningpublication of the American Political Science Association (APSA). APSR,first published in ..."}
+{"idx": 4, "title": "American Political Science Review, Cambridge University ...", "date": "", "ddg_snippet": "463-478 The Coevolution of American Political Science and the American Political Science Review ... 613-618 The Study of Political Parties, 1906–2005: The View ...", "subpage_snippet": "", "source": "ideas.repec.org", "link": "https://ideas.repec.org/s/cup/apsrev7.html", "content": "463-478 The Coevolution of American Political Science and the American Political Science Review ... 613-618 The Study of Political Parties, 1906–2005: The View ..."}
+{"idx": 5, "title": "FAQs - American Political Science Association (APSA)", "date": "", "ddg_snippet": "APSA Preprints is a free-to-access pre-publication platform dedicated to early research outputs in political science and related disciplines.", "subpage_snippet": "", "source": "apsanet.org", "link": "https://apsanet.org/publications/apsa-preprints/faqs/", "content": "APSA Preprints is a free-to-access pre-publication platform dedicated to early research outputs in political science and related disciplines."}
+{"idx": 6, "title": "This Journal's Future Is Female", "date": "", "ddg_snippet": "29 Jul 2019 — The American Political Science Review's new editors want to preserve its strong reputation while broadening its readership, relevance and ...", "subpage_snippet": "", "source": "www.insidehighered.com", "link": "https://www.insidehighered.com/news/2019/07/30/political-science-association-pleases-and-surprises-members-its-flagship", "content": "29 Jul 2019 — The American Political Science Review's new editors want to preserve its strong reputation while broadening its readership, relevance and ..."}
+{"idx": 7, "title": "Popular culture in (and out of) American political science", "date": "", "ddg_snippet": "by N Dorzweiler · 2017 · Cited by 14 — Abstract. Historically, American political science has rarely engaged popular culture as a central topic of study, despite the domain's outsized ...", "subpage_snippet": "", "source": "journals.sagepub.com", "link": "https://journals.sagepub.com/doi/abs/10.1177/0952695116684314", "content": "by N Dorzweiler · 2017 · Cited by 14 — Abstract. Historically, American political science has rarely engaged popular culture as a central topic of study, despite the domain's outsized ..."}
+{"idx": 8, "title": "Political Science — Whole Tale 1.2.0.0 documentation", "date": "", "ddg_snippet": "Top journals , including the American Journal of Political Science (AJPS) and the American Political Science Review ( APSR ) require authors reporting empirical ...", "subpage_snippet": "", "source": "wholetale.readthedocs.io", "link": "https://wholetale.readthedocs.io/en/latest/project/cases/political-science.html", "content": "Top journals , including the American Journal of Political Science (AJPS) and the American Political Science Review ( APSR ) require authors reporting empirical ..."}
+{"idx": 9, "title": "American Political Science Review - - Latest issue", "date": "", "ddg_snippet": "The Women, Peace, and Security (WPS) sector assumes increasing the number of women involved in peace negotiations drives better outcomes for local women.", "subpage_snippet": "", "source": "www.cambridge.org", "link": "https://www.cambridge.org/core/journals/american-political-science-review/latest-issue", "content": "The Women, Peace, and Security (WPS) sector assumes increasing the number of women involved in peace negotiations drives better outcomes for local women."}
diff --git a/data/sampled_jsons/Beautiful_Soup_Python_library_primary_function_HTML_parsing.jsonl b/data/sampled_jsons/Beautiful_Soup_Python_library_primary_function_HTML_parsing.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..27b19ed1c1f4d70eafa37781a403e4cdb343c7ee
--- /dev/null
+++ b/data/sampled_jsons/Beautiful_Soup_Python_library_primary_function_HTML_parsing.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "BeautifulSoup – Scraping Paragraphs from HTML |", "date": "", "ddg_snippet": "... Beautiful Soup (bs4) is a Python library for pulling data ... BeautifulSoup() function helps us to parse the html file or you say the encoding in html .", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/beautifulsoup-scraping-paragraphs-from-html/", "content": "... Beautiful Soup (bs4) is a Python library for pulling data ... BeautifulSoup() function helps us to parse the html file or you say the encoding in html ."}
+{"idx": 1, "title": "Python Web Scraping Tutorial | GeeksforGeeks", "date": "", "ddg_snippet": "Beautiful Soup sits on top of famous Python parsers like LXML and HTML , allowing you to try different parsing strategies or trade speed for ...", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/python-web-scraping-tutorial/", "content": "Beautiful Soup sits on top of famous Python parsers like LXML and HTML , allowing you to try different parsing strategies or trade speed for ..."}
+{"idx": 2, "title": "Implementing Web Scraping in Python with BeautifulSoup |", "date": "", "ddg_snippet": "A really nice thing about the BeautifulSoup library is that it is built on the top of the HTML parsing libraries like html5lib, lxml, html . parser ...", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/implementing-web-scraping-python-beautiful-soup/", "content": "A really nice thing about the BeautifulSoup library is that it is built on the top of the HTML parsing libraries like html5lib, lxml, html . parser ..."}
+{"idx": 3, "title": "Beautifulsoup Installation - Python - GeeksforGeeks", "date": "", "ddg_snippet": "Beautiful Soup is a Python library for pulling data out of HTML and XML files. ... BeautifulSoup BeautifulSoup(bs4) is a Python library for pulling ...", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/beautifulsoup-installation-python/", "content": "Beautiful Soup is a Python library for pulling data out of HTML and XML files. ... BeautifulSoup BeautifulSoup(bs4) is a Python library for pulling ..."}
+{"idx": 4, "title": "Python Parse HTML", "date": "", "ddg_snippet": "Before you start parsing , you need to bring in the Beautiful Soup library with this code. ... Beautiful Soup is excellent for parsing HTML and ...", "subpage_snippet": "", "source": "Serply.io", "link": "https://Serply.io/blog/python-parse-html", "content": "Before you start parsing , you need to bring in the Beautiful Soup library with this code. ... Beautiful Soup is excellent for parsing HTML and ..."}
+{"idx": 5, "title": "4 Best Python Libraries for Efficient Web Scraping (Updated)", "date": "", "ddg_snippet": "BeautifulSoup : This is a Python library used to parse ... Beautiful Soup is one of the best Python libraries for parsing HTML and XML documents.", "subpage_snippet": "", "source": "www.scrapingdog.com", "link": "https://www.scrapingdog.com/blog/best-python-web-scraping-libraries/", "content": "BeautifulSoup : This is a Python library used to parse ... Beautiful Soup is one of the best Python libraries for parsing HTML and XML documents."}
+{"idx": 6, "title": "Web Scraping with Python | A Beginner's Guide | PromptCloud", "date": "", "ddg_snippet": "Requests can be used to retrieve HTML content from web pages, which can then be parsed using Beautiful Soup . ... Beautiful Soup is a convenient-to-use ...", "subpage_snippet": "", "source": "www.promptcloud.com", "link": "https://www.promptcloud.com/blog/web-scraping-python-guide/", "content": "Requests can be used to retrieve HTML content from web pages, which can then be parsed using Beautiful Soup . ... Beautiful Soup is a convenient-to-use ..."}
+{"idx": 7, "title": "Open Source Software in Python", "date": "", "ddg_snippet": "Beautiful Soup - Beautiful Soup is a Python HTML /XML parser designed for quick turnaround projects like screen-scraping.", "subpage_snippet": "", "source": "pythonsource.com:443", "link": "https://pythonsource.com:443/", "content": "Beautiful Soup - Beautiful Soup is a Python HTML /XML parser designed for quick turnaround projects like screen-scraping."}
+{"idx": 8, "title": "Beautiful Soup 4 Python - PythonForBeginners.com", "date": "", "ddg_snippet": "Beautiful Soup is a Python library for pulling data out of HTML and XML files. ... Python script, import the library Now you have to pass something to ...", "subpage_snippet": "", "source": "www.pythonforbeginners.com:443", "link": "https://www.pythonforbeginners.com:443/beautifulsoup/beautifulsoup-4-python", "content": "Beautiful Soup is a Python library for pulling data out of HTML and XML files. ... Python script, import the library Now you have to pass something to ..."}
+{"idx": 9, "title": "Web Scraping with BeautifulSoup - PythonForBeginners.com", "date": "", "ddg_snippet": "Beautiful Soup sits on top of popular Python parsers like lxml and html5lib, allowing you to try out different parsing strategies or trade speed for ...", "subpage_snippet": "", "source": "www.pythonforbeginners.com", "link": "https://www.pythonforbeginners.com/python-on-the-web/web-scraping-with-beautifulsoup", "content": "Beautiful Soup sits on top of popular Python parsers like lxml and html5lib, allowing you to try out different parsing strategies or trade speed for ..."}
diff --git a/data/sampled_jsons/Boltzmann-Aligned_Inverse_Folding_Model_BA-Cycle_method_Section_3.2_difference_previous_methods.jsonl b/data/sampled_jsons/Boltzmann-Aligned_Inverse_Folding_Model_BA-Cycle_method_Section_3.2_difference_previous_methods.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bc97b98e51a459ba87711e7cfba0b25e16908658
--- /dev/null
+++ b/data/sampled_jsons/Boltzmann-Aligned_Inverse_Folding_Model_BA-Cycle_method_Section_3.2_difference_previous_methods.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Boltzmann-Aligned Inverse Folding Model as a Predictor of ...", "date": "", "ddg_snippet": "12 Oct 2024 — Next, we present a method (Sec. 3.2 ) that integrates the inverse folding model into Boltzmann alignment . This method is named BA - Cycle and uses ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.09543v1", "content": "12 Oct 2024 — Next, we present a method (Sec. 3.2 ) that integrates the inverse folding model into Boltzmann alignment . This method is named BA - Cycle and uses ..."}
+{"idx": 1, "title": "Energy-Based Models for Predicting Mutational Effects on ...", "date": "", "ddg_snippet": "by P Soga · 2025 — Boltzmann - Aligned Inverse Folding Model as a Predictor of Mu- tational Effects on Protein-Protein Interactions. In The Thirteenth ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/pdf/10.1145/3711896.3736931", "content": "by P Soga · 2025 — Boltzmann - Aligned Inverse Folding Model as a Predictor of Mu- tational Effects on Protein-Protein Interactions. In The Thirteenth ..."}
+{"idx": 2, "title": "A general temperature-guided language model to design ...", "date": "", "ddg_snippet": "27 Nov 2024 — We introduce PRIME, a deep learning model , which can suggest protein mutants with improved stability and activity without any prior experimental mutagenesis ...", "subpage_snippet": "", "source": "www.science.org", "link": "https://www.science.org/doi/10.1126/sciadv.adr2641", "content": "27 Nov 2024 — We introduce PRIME, a deep learning model , which can suggest protein mutants with improved stability and activity without any prior experimental mutagenesis ..."}
+{"idx": 3, "title": "Book", "date": "", "ddg_snippet": "Boltzmann - Aligned Inverse Folding Model as a Predictor of Mutational Effects on Protein-Protein Interactions Xiaoran Jiao, Weian Mao, Wengong Jin, Peiyuan ...", "subpage_snippet": "", "source": "proceedings.iclr.cc", "link": "https://proceedings.iclr.cc/paper_files/paper/2025", "content": "Boltzmann - Aligned Inverse Folding Model as a Predictor of Mutational Effects on Protein-Protein Interactions Xiaoran Jiao, Weian Mao, Wengong Jin, Peiyuan ..."}
+{"idx": 4, "title": "Limitations and challenges in protein stability prediction ...", "date": "", "ddg_snippet": "by T Sanavia · 2020 · Cited by 155 — To model mutations, FoldX uses a rotamer approach , allowing conformational changes of sidechains and keeping the backbone fixed. Another group ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7397395/", "content": "by T Sanavia · 2020 · Cited by 155 — To model mutations, FoldX uses a rotamer approach , allowing conformational changes of sidechains and keeping the backbone fixed. Another group ..."}
+{"idx": 5, "title": "Protein representations: Encoding biological information ...", "date": "", "ddg_snippet": "by D Harding-Larsen · 2024 · Cited by 17 — In this review, we investigate the methodologies for protein representation utilizing the protein sequence, structure, or dynamics. We also analyse the ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0734975024001538", "content": "by D Harding-Larsen · 2024 · Cited by 17 — In this review, we investigate the methodologies for protein representation utilizing the protein sequence, structure, or dynamics. We also analyse the ..."}
+{"idx": 6, "title": "RNA Structural Dynamics As Captured by Molecular Simulations", "date": "", "ddg_snippet": "by J Šponer · 2018 · Cited by 567 — This procedure is analogous to the free-energy perturbation method (see section 3.2 .7) and is used to switch between two alternative force ...", "subpage_snippet": "", "source": "pubs.acs.org", "link": "https://pubs.acs.org/doi/10.1021/acs.chemrev.7b00427", "content": "by J Šponer · 2018 · Cited by 567 — This procedure is analogous to the free-energy perturbation method (see section 3.2 .7) and is used to switch between two alternative force ..."}
+{"idx": 7, "title": "UC Berkeley", "date": "", "ddg_snippet": "by O Zhang · 2025 — At present, few approaches exist for modeling IDRs within the context of folded regions. 167 pages", "subpage_snippet": "", "source": "escholarship.org", "link": "https://escholarship.org/content/qt7xz4z3c4/qt7xz4z3c4.pdf", "content": "by O Zhang · 2025 — At present, few approaches exist for modeling IDRs within the context of folded regions. 167 pages"}
+{"idx": 8, "title": "Statistical Computation for Problems in Dynamic ... - Harvard DASH", "date": "", "ddg_snippet": "loop modeling datasets and compare with other previous methods in Section 3.3. We conclude the paper with a brief discussion in Section 3.4. 3.2 Methods . Our ...", "subpage_snippet": "", "source": "dash.harvard.edu", "link": "https://dash.harvard.edu/bitstreams/891f2837-34ad-48b8-b60f-3a4ca88e179a/download", "content": "loop modeling datasets and compare with other previous methods in Section 3.3. We conclude the paper with a brief discussion in Section 3.4. 3.2 Methods . Our ..."}
+{"idx": 9, "title": "Machine learning meets physics: A two-way street - PMC", "date": "", "ddg_snippet": "by H Levine · 2024 · Cited by 20 — This article introduces a special issue on the interaction between the rapidly expanding field of machine learning and ongoing research in physics.", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11228530/", "content": "by H Levine · 2024 · Cited by 20 — This article introduces a special issue on the interaction between the rapidly expanding field of machine learning and ongoing research in physics."}
diff --git a/data/sampled_jsons/Buchholz_causal_representation_learning_Gaussian_distribution_assumptions.jsonl b/data/sampled_jsons/Buchholz_causal_representation_learning_Gaussian_distribution_assumptions.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6d21b68a9162b2cce01319ff66ba0429c0742de9
--- /dev/null
+++ b/data/sampled_jsons/Buchholz_causal_representation_learning_Gaussian_distribution_assumptions.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "CausalML", "date": "", "ddg_snippet": "At the moment, I m particularly interested in causal representation learning and applications in biology. ... Causal Representation Learning .", "subpage_snippet": "", "source": "www.juliusvonkugelgen.com", "link": "https://www.juliusvonkugelgen.com/", "content": "At the moment, I m particularly interested in causal representation learning and applications in biology. ... Causal Representation Learning ."}
+{"idx": 1, "title": "Identifying Weight-Variant Latent Causal Models", "date": "", "ddg_snippet": "Despite its advantages, causal representation learning is a notoriously hard problem—without certain assumptions , identifying the true latent ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2208.14153v6", "content": "Despite its advantages, causal representation learning is a notoriously hard problem—without certain assumptions , identifying the true latent ..."}
+{"idx": 2, "title": "On the Origins of Linear Representations in Large Language", "date": "", "ddg_snippet": "First, linear representation structure is not specific to the choice of model architecture, but a by-product of how the model learns the conditional ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2403.03867v1", "content": "First, linear representation structure is not specific to the choice of model architecture, but a by-product of how the model learns the conditional ..."}
+{"idx": 3, "title": "CausalML", "date": "", "ddg_snippet": "At the moment, I m particularly interested in causal representation learning and applications in biology. ... Causal Representation Learning .", "subpage_snippet": "", "source": "sites.google.com", "link": "https://sites.google.com/view/julius-von-kuegelgen/home/", "content": "At the moment, I m particularly interested in causal representation learning and applications in biology. ... Causal Representation Learning ."}
+{"idx": 4, "title": "Bryon Aragam // University of Chicago", "date": "", "ddg_snippet": "This is the goal of the emerging field of causal representation learning (CRL) that uses causality as a vector for building flexible, interpretable ...", "subpage_snippet": "", "source": "www.bryonaragam.com", "link": "http://www.bryonaragam.com/", "content": "This is the goal of the emerging field of causal representation learning (CRL) that uses causality as a vector for building flexible, interpretable ..."}
+{"idx": 5, "title": "Bryon Aragam // University of Chicago", "date": "", "ddg_snippet": "... causal discovery methods and then uses a variational autoencoder (VAE) that is constrained to abide by the Markov factorization of the distribution ...", "subpage_snippet": "", "source": "www.bryonaragam.com", "link": "https://www.bryonaragam.com/", "content": "... causal discovery methods and then uses a variational autoencoder (VAE) that is constrained to abide by the Markov factorization of the distribution ..."}
+{"idx": 6, "title": "Pradeep Ravikumar", "date": "", "ddg_snippet": "Abstract: We consider the linear causal representation learning setting where we observe a linear mixing of $d$ unknown latent factors, which follow ...", "subpage_snippet": "", "source": "www.catalyzex.com", "link": "https://www.catalyzex.com/author/Pradeep+Ravikumar", "content": "Abstract: We consider the linear causal representation learning setting where we observe a linear mixing of $d$ unknown latent factors, which follow ..."}
+{"idx": 7, "title": "Pradeep Ravikumar", "date": "", "ddg_snippet": "Learning Interpretable Concepts: Unifying Causal Representation Learning and Foundation Models. ... Learning Linear Causal Representations from ...", "subpage_snippet": "", "source": "www.csauthors.net", "link": "https://www.csauthors.net/pradeep-ravikumar/", "content": "Learning Interpretable Concepts: Unifying Causal Representation Learning and Foundation Models. ... Learning Linear Causal Representations from ..."}
+{"idx": 8, "title": "Stefan Bauer", "date": "", "ddg_snippet": "Exploring the Effectiveness of Object-Centric Representations in Visual Question Answering: Comparative Insights with Foundation Models.", "subpage_snippet": "", "source": "www.csauthors.net", "link": "https://www.csauthors.net/stefan-bauer/", "content": "Exploring the Effectiveness of Object-Centric Representations in Visual Question Answering: Comparative Insights with Foundation Models."}
+{"idx": 9, "title": "NeurIPS 2023 Orals", "date": "", "ddg_snippet": "... conditioning on discrete events.Our key tool is probability generating functions :they provide a compact closed-form representation of distributions ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2023/events/oral", "content": "... conditioning on discrete events.Our key tool is probability generating functions :they provide a compact closed-form representation of distributions ..."}
diff --git a/data/sampled_jsons/CRAB_benchmark_Xu_et_al_2024a_virtual_agent.jsonl b/data/sampled_jsons/CRAB_benchmark_Xu_et_al_2024a_virtual_agent.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3fbe6b58f6fdbbb553105b92c1bdc4344c0041bb
--- /dev/null
+++ b/data/sampled_jsons/CRAB_benchmark_Xu_et_al_2024a_virtual_agent.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2407.01511] CRAB: Cross-environment Agent Benchmark for Multimodal ...", "date": "", "ddg_snippet": "To overcome these limitations, we introduce Crab , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an efficient mechanism for task and evaluator construction.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2407.01511", "content": "To overcome these limitations, we introduce Crab , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an efficient mechanism for task and evaluator construction."}
+{"idx": 1, "title": "CRAB: Cross-platform Agent Benchmark for Multimodal ... - GitHub", "date": "", "ddg_snippet": "🌐 Cross-platform and Multi-environment Create build agent environments that support various deployment options including in-memory, Docker-hosted, virtual machines, or distributed physical machines, provided they are accessible via Python functions. Let the agent access all the environments in the same time through a unified interface. ⚙ ️Easy-to-use Configuration Add a new action by ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/camel-ai/crab", "content": "🌐 Cross-platform and Multi-environment Create build agent environments that support various deployment options including in-memory, Docker-hosted, virtual machines, or distributed physical machines, provided they are accessible via Python functions. Let the agent access all the environments in the same time through a unified interface. ⚙ ️Easy-to-use Configuration Add a new action by ..."}
+{"idx": 2, "title": "CRAB: Cross-platfrom agent benchmark for multi-modal embodied language ...", "date": "", "ddg_snippet": "CRAB : Cross-platfrom agent benchmark for multi-modal embodied language model agents Tianqi Xu · Linyao Chen · Dai-Jie Wu · Yanjun Chen · Zecheng Zhang · Xiang Yao · Zhiqiang Xie · Yongchao Chen · Shilong Liu · Bochen Qian · Philip Torr · Bernard Ghanem · Guohao Li", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2024/100913", "content": "CRAB : Cross-platfrom agent benchmark for multi-modal embodied language model agents Tianqi Xu · Linyao Chen · Dai-Jie Wu · Yanjun Chen · Zecheng Zhang · Xiang Yao · Zhiqiang Xie · Yongchao Chen · Shilong Liu · Bochen Qian · Philip Torr · Bernard Ghanem · Guohao Li"}
+{"idx": 3, "title": "CRAB: Cross-environment Agent Benchmark for Multimodal Language Model ...", "date": "", "ddg_snippet": "To overcome these limitations, we introduce CRAB , the first cross-environment agent benchmark framework, incorporating a graph-based fine-grained evaluation method and an efficient task generation method.", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.findings-acl.1113/", "content": "To overcome these limitations, we introduce CRAB , the first cross-environment agent benchmark framework, incorporating a graph-based fine-grained evaluation method and an efficient task generation method."}
+{"idx": 4, "title": "CRAB: Cross-environment Agent Benchmark for Multimodal Language Model ...", "date": "", "ddg_snippet": "CRAB aims to become a general-purpose agent benchmark framework for Multimodal Language Model (MLM) agents . CRAB provides an end-to-end while easy-to-use framework to build agents , operate environments, and create benchmarks to evaluate them, featuring three key components: cross-environment support, a graph evaluator, and task generation.", "subpage_snippet": "", "source": "crab.camel-ai.org", "link": "https://crab.camel-ai.org/", "content": "CRAB aims to become a general-purpose agent benchmark framework for Multimodal Language Model (MLM) agents . CRAB provides an end-to-end while easy-to-use framework to build agents , operate environments, and create benchmarks to evaluate them, featuring three key components: cross-environment support, a graph evaluator, and task generation."}
+{"idx": 5, "title": "(PDF) CRAB: Cross-environment Agent Benchmark for ... - ResearchGate", "date": "", "ddg_snippet": "To overcome these limitations, we introduce Crab , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/381910878_CRAB_Cross-environment_Agent_Benchmark_for_Multimodal_Language_Model_Agents", "content": "To overcome these limitations, we introduce Crab , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an ..."}
+{"idx": 6, "title": "Crab: Cross-platform Agent Benchmark for Multimodal Embodied Language ...", "date": "", "ddg_snippet": "To address these limitations, we introduce the Crab framework, a well-defined, highly structured, and unified benchmark framework, which enables agents to interact across multiple environments, incorporates a graph-based, fine-grained evaluation system, and supports automatic task generation.", "subpage_snippet": "", "source": "repository.kaust.edu.sa", "link": "https://repository.kaust.edu.sa/items/d4cba8f4-69ec-4499-a347-9f2908a2b93d", "content": "To address these limitations, we introduce the Crab framework, a well-defined, highly structured, and unified benchmark framework, which enables agents to interact across multiple environments, incorporates a graph-based, fine-grained evaluation system, and supports automatic task generation."}
+{"idx": 7, "title": "CRAB: Cross-environment Agent Benchmark for Multimodal Language Model ...", "date": "", "ddg_snippet": "We build an agent benchmark CRAB Benchmark -v0 featuring with cross-environment, graph evaluator, and task generation through CRAB framework. The environments consists of an Android smartphone emulator and a Ubuntu Linux desktop virtual machine.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2407.01511", "content": "We build an agent benchmark CRAB Benchmark -v0 featuring with cross-environment, graph evaluator, and task generation through CRAB framework. The environments consists of an Android smartphone emulator and a Ubuntu Linux desktop virtual machine."}
+{"idx": 8, "title": "CRAB: CROSS ENVIRONMENT AGENT BENCHMARK FOR MULTIMODAL ... - OpenReview", "date": "", "ddg_snippet": "Leveraging CRAB , we developed a cross-platform Crab Benchmark -v0 comprising 120 tasks in computer desktop and mobile phone envi- ronments. We evaluated 6 advanced MLMs using different single and multi- agent system configurations on this benchmark .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qqKJjwibsp", "content": "Leveraging CRAB , we developed a cross-platform Crab Benchmark -v0 comprising 120 tasks in computer desktop and mobile phone envi- ronments. We evaluated 6 advanced MLMs using different single and multi- agent system configurations on this benchmark ."}
+{"idx": 9, "title": "CRAB: cross-environment agent benchmark for multimodal language model ...", "date": "", "ddg_snippet": "To overcome these limitations, we introduce CRAB , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an efficient mechanism for task and evaluator construction.", "subpage_snippet": "", "source": "ora.ox.ac.uk", "link": "https://ora.ox.ac.uk/objects/uuid:53a31ad5-7aa7-46c7-aa52-5a5e2dc0e6cd", "content": "To overcome these limitations, we introduce CRAB , the first agent benchmark framework designed to support cross-environment tasks, incorporating a graph-based fine-grained evaluation method and an efficient mechanism for task and evaluator construction."}
diff --git a/data/sampled_jsons/CVE-Bench_A_Benchmark_for_AI_Agents_Ability_to_Exploit_Real-World_Web_Application_Vulnerabilities.jsonl b/data/sampled_jsons/CVE-Bench_A_Benchmark_for_AI_Agents_Ability_to_Exploit_Real-World_Web_Application_Vulnerabilities.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3da451c71ff5b2e645ee6bca3f58267021b45049
--- /dev/null
+++ b/data/sampled_jsons/CVE-Bench_A_Benchmark_for_AI_Agents_Ability_to_Exploit_Real-World_Web_Application_Vulnerabilities.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "by Y Zhu · 2025 · Cited by 14 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.17332", "content": "by Y Zhu · 2025 · Cited by 14 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions."}
+{"idx": 1, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "This repository contains data and code used in the CVE-Bench (paper, blog), which is for evaluating AI agents on real world web vulnerabilities and exploits.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/uiuc-kang-lab/cve-bench", "content": "This repository contains data and code used in the CVE-Bench (paper, blog), which is for evaluating AI agents on real world web vulnerabilities and exploits."}
+{"idx": 2, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "by Y Zhu · Cited by 12 — We introduce CVE-Bench, a benchmark built from real-world web applications . Each application contains a critical-security vulnerability that actually happened.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=3pk0p4NGmQ", "content": "by Y Zhu · Cited by 12 — We introduce CVE-Bench, a benchmark built from real-world web applications . Each application contains a critical-security vulnerability that actually happened."}
+{"idx": 3, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "by Y Zhu · 2025 · Cited by 12 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vul- nerable web applications in scenarios that mimic real - world ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.17332", "content": "by Y Zhu · 2025 · Cited by 12 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vul- nerable web applications in scenarios that mimic real - world ..."}
+{"idx": 4, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "by Y Zhu · 2025 · Cited by 14 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions.", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "https://ui.adsabs.harvard.edu/abs/2025arXiv250317332Z/abstract", "content": "by Y Zhu · 2025 · Cited by 14 — In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions."}
+{"idx": 5, "title": "[Literature Review] CVE-Bench: A Benchmark for AI Agents ...", "date": "", "ddg_snippet": "21 Mar 2025 — A novel framework called CVE-Bench designed to evaluate the capacities of AI agents, particularly large language models (LLMs), in conducting cyberattacks on ...", "subpage_snippet": "", "source": "www.themoonlight.io", "link": "https://www.themoonlight.io/en/review/cve-bench-a-benchmark-for-ai-agents-ability-to-exploit-real-world-web-application-vulnerabilities", "content": "21 Mar 2025 — A novel framework called CVE-Bench designed to evaluate the capacities of AI agents, particularly large language models (LLMs), in conducting cyberattacks on ..."}
+{"idx": 6, "title": "CVE-Bench: A Benchmark for AI Agents' Ability to Exploit ...", "date": "", "ddg_snippet": "In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions, while ...", "subpage_snippet": "", "source": "chatpaper.com", "link": "https://chatpaper.com/paper/165204", "content": "In CVE - Bench , we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real - world conditions, while ..."}
+{"idx": 7, "title": "CVE-Bench: A Real-World Cybersecurity Benchmark for ...", "date": "", "ddg_snippet": "23 Apr 2025 — Daniel Kang introduced \" CVE-Bench,\" the first benchmark to evaluate AI agents against real-world cybersecurity vulnerabilities. SESSION ...", "subpage_snippet": "", "source": "far.ai", "link": "https://far.ai/events/sessions/daniel-kang-cve-bench-a-real-world-cybersecurity-benchmark-for-ai-agents", "content": "23 Apr 2025 — Daniel Kang introduced \" CVE-Bench,\" the first benchmark to evaluate AI agents against real-world cybersecurity vulnerabilities. SESSION ..."}
+{"idx": 8, "title": "Chloé Messdaghi's Post - CVE-Bench", "date": "", "ddg_snippet": "Evaluate the capability of AI systems to leverage actual vulnerabilities by utilizing a benchmark that tests them against 40 key CVEs from the National ...", "subpage_snippet": "", "source": "www.linkedin.com", "link": "https://www.linkedin.com/posts/chloemessdaghi_cve-bench-a-benchmark-for-ai-agents-ability-activity-7325202007328194560-oJXW", "content": "Evaluate the capability of AI systems to leverage actual vulnerabilities by utilizing a benchmark that tests them against 40 key CVEs from the National ..."}
+{"idx": 9, "title": "Measuring AI Agents' Ability to Exploit Web Applications", "date": "", "ddg_snippet": "In this post, we introduce CVE-bench — the first benchmark built on real-world vulnerabilities, which contains: 40 real-world vulnerability- ...", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/@danieldkang/measuring-ai-agents-ability-to-exploit-web-applications-ba4225aa281f", "content": "In this post, we introduce CVE-bench — the first benchmark built on real-world vulnerabilities, which contains: 40 real-world vulnerability- ..."}
diff --git "a/data/sampled_jsons/Catoni_Contextual_Bandits_Theorem_3.1_Theorem_3.4_regret_bound_\317\203_t_\342\210\232(\316\243\317\203_t^2).jsonl" "b/data/sampled_jsons/Catoni_Contextual_Bandits_Theorem_3.1_Theorem_3.4_regret_bound_\317\203_t_\342\210\232(\316\243\317\203_t^2).jsonl"
new file mode 100644
index 0000000000000000000000000000000000000000..aec90f69034c5c64c618f47b5aeb6d2d76ed06b3
--- /dev/null
+++ "b/data/sampled_jsons/Catoni_Contextual_Bandits_Theorem_3.1_Theorem_3.4_regret_bound_\317\203_t_\342\210\232(\316\243\317\203_t^2).jsonl"
@@ -0,0 +1,9 @@
+{"idx": 0, "title": "Heavy-Tailed Linear Bandits: Huber Regression with One-Pass", "date": "", "ddg_snippet": "Table 1 : Comparisons of our regret bounds and computational complexity to previous best-known results for heavy-tailed linear bandits .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.00419v1", "content": "Table 1 : Comparisons of our regret bounds and computational complexity to previous best-known results for heavy-tailed linear bandits ."}
+{"idx": 1, "title": "ICML 2022 Papers", "date": "", "ddg_snippet": "SPECTRE: Spectral Conditioning Helps to Overcome the Expressivity Limits of One -shot Graph Generators ... Versatile Dueling Bandits : Best-of-both ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2022/papers.html", "content": "SPECTRE: Spectral Conditioning Helps to Overcome the Expressivity Limits of One -shot Graph Generators ... Versatile Dueling Bandits : Best-of-both ..."}
+{"idx": 2, "title": "Extended UCB Policies for Multi-Armed Bandit Problems", "date": "", "ddg_snippet": "... 4 p= 4 and q = 2 q=2 ) to arbitrarily chosen p > q > 1 p > q > 1 as long as the two moments have a known controlled relationship, while ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/1112.1768v5", "content": "... 4 p= 4 and q = 2 q=2 ) to arbitrarily chosen p > q > 1 p > q > 1 as long as the two moments have a known controlled relationship, while ..."}
+{"idx": 3, "title": "How Balanced Should Causal Covariates Be?", "date": "", "ddg_snippet": "A key assumption in causal inference is unconfoundedness [Rosenbaum and Rubin, 1983 ] , which posits that biases of the selection mechanism can be ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.03860v1", "content": "A key assumption in causal inference is unconfoundedness [Rosenbaum and Rubin, 1983 ] , which posits that biases of the selection mechanism can be ..."}
+{"idx": 4, "title": "Machine Learning Feb 2025", "date": "", "ddg_snippet": "Title: LoRA- One : One -Step Full Gradient Could Suffice for Fine-Tuning Large Language Models, Provably and Efficiently", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/list/stat.ML/2025-02", "content": "Title: LoRA- One : One -Step Full Gradient Could Suffice for Fine-Tuning Large Language Models, Provably and Efficiently"}
+{"idx": 5, "title": "The Stats Map · Basic Inequalities", "date": "", "ddg_snippet": "To prove this just apply expectations to the inequality 1 { x ≥ t } ≤ t x 1 { x ≥ t } which holds for all x and all t ≥ 0 .", "subpage_snippet": "", "source": "thestatsmap.com", "link": "https://thestatsmap.com/basic-inequalities", "content": "To prove this just apply expectations to the inequality 1 { x ≥ t } ≤ t x 1 { x ≥ t } which holds for all x and all t ≥ 0 ."}
+{"idx": 6, "title": "The Stats Map · Active Statistical Inference", "date": "", "ddg_snippet": "Suppose we have data ( X i , Y i ) i = 1 N , where Y i is unobserved and X i is observed. ... V ( θ ) = N 1 ( V ( Y ) + E ...", "subpage_snippet": "", "source": "thestatsmap.com", "link": "https://thestatsmap.com/active-statistical-inference", "content": "Suppose we have data ( X i , Y i ) i = 1 N , where Y i is unobserved and X i is observed. ... V ( θ ) = N 1 ( V ( Y ) + E ..."}
+{"idx": 7, "title": "Downloads", "date": "", "ddg_snippet": "3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Downloads/2022", "content": "3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation"}
+{"idx": 8, "title": "Statistics and Machine Learning Working Group: List of Past", "date": "", "ddg_snippet": "14 Feb 2025 Off-Policy Evaluations of Linear Functionals: From Nonadaptive Semiparametric Efficiency to Adaptive Nonasymptotic Optimality Ojash ...", "subpage_snippet": "", "source": "statml.cs.cmu.edu", "link": "http://statml.cs.cmu.edu/blog/", "content": "14 Feb 2025 Off-Policy Evaluations of Linear Functionals: From Nonadaptive Semiparametric Efficiency to Adaptive Nonasymptotic Optimality Ojash ..."}
diff --git a/data/sampled_jsons/Catoni_Contextual_Bandits_are_Robust_to_Heavy-tailed_Rewards_Assumption_4.1.jsonl b/data/sampled_jsons/Catoni_Contextual_Bandits_are_Robust_to_Heavy-tailed_Rewards_Assumption_4.1.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..24dc4b44e0397e45bb77ee00e83d631c579dc76a
--- /dev/null
+++ b/data/sampled_jsons/Catoni_Contextual_Bandits_are_Robust_to_Heavy-tailed_Rewards_Assumption_4.1.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Catoni Contextual Bandits are Robust to Heavy - tailed Rewards", "date": "", "ddg_snippet": "This work considers a dierent route for robustness to heavy - tailed rewards , building on the well-studied Catoni ’s mean estimator from the robust statistics literature. We design a contextual bandit (CB)...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2502.02486", "content": "This work considers a dierent route for robustness to heavy - tailed rewards , building on the well-studied Catoni ’s mean estimator from the robust statistics literature. We design a contextual bandit (CB)..."}
+{"idx": 1, "title": "ICML Poster Catoni Contextual Bandits are Robust to Heavy - tailed ...", "date": "", "ddg_snippet": "Abstract: Typical contextual bandit algorithms assume that the rewards at each round lie in some fixed range $[0, R]$, and their regret scales polynomially with this reward range $R$.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46438", "content": "Abstract: Typical contextual bandit algorithms assume that the rewards at each round lie in some fixed range $[0, R]$, and their regret scales polynomially with this reward range $R$."}
+{"idx": 2, "title": "Multi-Armed Bandits | Papers With Code", "date": "", "ddg_snippet": "Catoni Contextual Bandits are Robust to Heavy - tailed Rewards .Multi-agent Multi-armed Bandit with Fully Heavy - tailed Dynamics.", "subpage_snippet": "", "source": "paperswithcode.com", "link": "https://paperswithcode.com/task/multi-armed-bandits/codeless?page=4", "content": "Catoni Contextual Bandits are Robust to Heavy - tailed Rewards .Multi-agent Multi-armed Bandit with Fully Heavy - tailed Dynamics."}
+{"idx": 3, "title": "Cooperative Multi-Agent Bandits with Heavy Tails", "date": "", "ddg_snippet": "Cooperative Multiagent Bandits with Heavy Tails . tailed effects, which is the central theme of this paper. Catoni (2012) provides this as the optimal achievable rate under heavy tails , and Bubeck et al. (2013) demonstrate that the trimmed mean achieves this rate (see appendix).", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v119/dubey20a/dubey20a.pdf", "content": "Cooperative Multiagent Bandits with Heavy Tails . tailed effects, which is the central theme of this paper. Catoni (2012) provides this as the optimal achievable rate under heavy tails , and Bubeck et al. (2013) demonstrate that the trimmed mean achieves this rate (see appendix)."}
+{"idx": 4, "title": "Bandits Corrupted by Nature: Lower Bounds on Regret and Robust ...", "date": "", "ddg_snippet": "Heavy - tailed bandits . Bubeck et al.The choice of βi reflects the difference between heavy - tailed bandits √and corrupted bandits . When the data are heavy - tailed but not corrupted, Catoni (2012) shows that βi ≃ σi n is a good choice for. the scaling parameter.", "subpage_snippet": "", "source": "hal.science", "link": "https://hal.science/hal-04615733v1/document", "content": "Heavy - tailed bandits . Bubeck et al.The choice of βi reflects the difference between heavy - tailed bandits √and corrupted bandits . When the data are heavy - tailed but not corrupted, Catoni (2012) shows that βi ≃ σi n is a good choice for. the scaling parameter."}
+{"idx": 5, "title": "(PDF) Regret Minimization in Isotonic, Heavy - Tailed Contextual ...", "date": "", "ddg_snippet": "(ii) Bandits with heavy tails — In the multi-armed bandit literature, one typically assumes that the reward . distributions are bounded and sub-gaussian. This assumption is restrictive in applications with heavy", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/355496335_Regret_Minimization_in_Isotonic_Heavy-Tailed_Contextual_Bandits_via_Adaptive_Confidence_Bands", "content": "(ii) Bandits with heavy tails — In the multi-armed bandit literature, one typically assumes that the reward . distributions are bounded and sub-gaussian. This assumption is restrictive in applications with heavy"}
+{"idx": 6, "title": "Nearly optimal regret for stochastic linear bandits with heavy - tailed ...", "date": "", "ddg_snippet": "In linear stochastic bandits , it is commonly assumed that payoffs are with sub-Gaussian noises. In this paper, under a weaker assumption on noises, we study the problem of linear stochastic bandits with heavy - tailed payoffs (LinBET), where the ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3491440.3491846", "content": "In linear stochastic bandits , it is commonly assumed that payoffs are with sub-Gaussian noises. In this paper, under a weaker assumption on noises, we study the problem of linear stochastic bandits with heavy - tailed payoffs (LinBET), where the ..."}
+{"idx": 7, "title": "On Private and Robust Bandits", "date": "", "ddg_snippet": "We study private and robust multi-armed bandits (MABs), where the agent receives Huber’s contaminated heavy - tailed rewards and meanwhile needs to ensure dif-ferential privacy.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/6d13e085b79d454da5910e4ca82a3d9d-Paper-Conference.pdf", "content": "We study private and robust multi-armed bandits (MABs), where the agent receives Huber’s contaminated heavy - tailed rewards and meanwhile needs to ensure dif-ferential privacy."}
+{"idx": 8, "title": "Chenlu Ye - Google Akademik", "date": "", "ddg_snippet": "Corruption- Robust Offline Reinforcement Learning with General Function Approximation. Catoni Contextual Bandits are Robust to Heavy - tailed Rewards .", "subpage_snippet": "", "source": "scholar.google.com.tr", "link": "https://scholar.google.com.tr/citations?user=c8yK5XsAAAAJ&hl=tr", "content": "Corruption- Robust Offline Reinforcement Learning with General Function Approximation. Catoni Contextual Bandits are Robust to Heavy - tailed Rewards ."}
+{"idx": 9, "title": "Optimal Algorithms for Stochastic Multi-Armed", "date": "", "ddg_snippet": "Multi-Armed Bandits with Heavy - Tailed Rewards We compare APE2 with robust UCB [3] and DSEE [17].", "subpage_snippet": "", "source": "papers.nips.cc", "link": "https://papers.nips.cc/paper/2020/file/607bc9ebe4abfcd65181bfbef6252830-Paper.pdf", "content": "Multi-Armed Bandits with Heavy - Tailed Rewards We compare APE2 with robust UCB [3] and DSEE [17]."}
diff --git a/data/sampled_jsons/Causal_Representation_Learning_robust_stochastic_mixing_function_2024_arxiv.jsonl b/data/sampled_jsons/Causal_Representation_Learning_robust_stochastic_mixing_function_2024_arxiv.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..299b0d765d33753ad32007a052794979dba9710c
--- /dev/null
+++ b/data/sampled_jsons/Causal_Representation_Learning_robust_stochastic_mixing_function_2024_arxiv.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Sanity Checking Causal Representation Learning on a ...", "date": "", "ddg_snippet": "28 Apr 2025 — The underlying causal factors are transformed into observations through a nonlinear and deterministic mixing function (Buchholz et al., 2023, ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.20099v2", "content": "28 Apr 2025 — The underlying causal factors are transformed into observations through a nonlinear and deterministic mixing function (Buchholz et al., 2023, ..."}
+{"idx": 1, "title": "unifying causal representation learning with", "date": "", "ddg_snippet": "by D Yao · 2024 · Cited by 13 — Causal representation learning (CRL ) aims at recovering latent causal variables from high-dimensional observations to solve causal ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2409.02772", "content": "by D Yao · 2024 · Cited by 13 — Causal representation learning (CRL ) aims at recovering latent causal variables from high-dimensional observations to solve causal ..."}
+{"idx": 2, "title": "SPARTAN: A Sparse Transformer Learning Local Causation", "date": "", "ddg_snippet": "by A Lei · 2024 · Cited by 4 — We demonstrate experimentally that learning a sparse, local causal world model improves robustness and adaptation efficiency, and that our model ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2411.06890", "content": "by A Lei · 2024 · Cited by 4 — We demonstrate experimentally that learning a sparse, local causal world model improves robustness and adaptation efficiency, and that our model ..."}
+{"idx": 3, "title": "Causal Representation Learning Made Identifiable by ...", "date": "", "ddg_snippet": "7 Jun 2024 — Here, we show identifiability based on novel, weak constraints, which requires no temporal structure, intervention, nor weak supervision.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2310.15709v2", "content": "7 Jun 2024 — Here, we show identifiability based on novel, weak constraints, which requires no temporal structure, intervention, nor weak supervision."}
+{"idx": 4, "title": "From Causal to Concept-Based Representation Learning", "date": "", "ddg_snippet": "by G Rajendran · 2024 · Cited by 6 — A sparsity principle for partially observable causal representation learning. arXiv preprint. arXiv:2403.08335, 2024 . [134] M. Yang, F. Liu, Z. Chen, X. Shen ...", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2024/file/b76a9959151d377ddd2c77a275a97475-Paper-Conference.pdf", "content": "by G Rajendran · 2024 · Cited by 6 — A sparsity principle for partially observable causal representation learning. arXiv preprint. arXiv:2403.08335, 2024 . [134] M. Yang, F. Liu, Z. Chen, X. Shen ..."}
+{"idx": 5, "title": "Linear Causal Representation Learning from Unknown ...", "date": "", "ddg_snippet": "by B Varıcı · 2024 · Cited by 4 — In this paper, we address the open problem of using unknown multi-node (UMN) stochastic interven- tions to recover the latent causal variables Z ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2406.05937", "content": "by B Varıcı · 2024 · Cited by 4 — In this paper, we address the open problem of using unknown multi-node (UMN) stochastic interven- tions to recover the latent causal variables Z ..."}
+{"idx": 6, "title": "Unifying Causal Representation Learning and Foundation ...", "date": "", "ddg_snippet": "9 Dec 2024 — In this work, we relate these two approaches and study how to learn human-interpretable concepts from data.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2402.09236v2", "content": "9 Dec 2024 — In this work, we relate these two approaches and study how to learn human-interpretable concepts from data."}
+{"idx": 7, "title": "Disentangled Representation Learning in Non-Markovian ...", "date": "", "ddg_snippet": "by A Li · 2024 · Cited by 5 — Considering various data modalities, such as images, videos, and text, humans perform causal reasoning using high-level causal variables, as opposed to ...", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2024/file/bd7ffcbd088b5e12f3b9eecc9c498b27-Paper-Conference.pdf", "content": "by A Li · 2024 · Cited by 5 — Considering various data modalities, such as images, videos, and text, humans perform causal reasoning using high-level causal variables, as opposed to ..."}
+{"idx": 8, "title": "Causal Representation Learning from Multi-modal ...", "date": "", "ddg_snippet": "by Y Sun · 2025 · Cited by 6 — In this paper, we leverage causal principles to develop theoretically sound ML models for multimodal biomedical data, aiming to provide reliable and ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11952583/", "content": "by Y Sun · 2025 · Cited by 6 — In this paper, we leverage causal principles to develop theoretically sound ML models for multimodal biomedical data, aiming to provide reliable and ..."}
+{"idx": 9, "title": "arXiv:2306.01213v4 [cs.LG] 23 Aug 2024", "date": "", "ddg_snippet": "by A Komanduri · 2023 · Cited by 17 — We propose ICM-VAE, a framework for learning causally disentangled representations supervised by causally related ob- served labels. We model ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2306.01213", "content": "by A Komanduri · 2023 · Cited by 17 — We propose ICM-VAE, a framework for learning causally disentangled representations supervised by causally related ob- served labels. We model ..."}
diff --git a/data/sampled_jsons/Checks-and-Balances_Framework_Section_3.2_four-step_self-supervised_learning_process_mapping_behavio.jsonl b/data/sampled_jsons/Checks-and-Balances_Framework_Section_3.2_four-step_self-supervised_learning_process_mapping_behavio.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b872c6defc7516ac7f650db785364c1a19d28556
--- /dev/null
+++ b/data/sampled_jsons/Checks-and-Balances_Framework_Section_3.2_four-step_self-supervised_learning_process_mapping_behavio.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Order Personal Checks Online At Affordable Prices | Checks.com", "date": "", "ddg_snippet": "We print an exceptional collection of high-quality personal checks at cheap prices. Whether you are buying a checkbook for the first time or are reordering your favorite check design, you’ll always get the same low price - no discounts codes required.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/", "content": "We print an exceptional collection of high-quality personal checks at cheap prices. Whether you are buying a checkbook for the first time or are reordering your favorite check design, you’ll always get the same low price - no discounts codes required."}
+{"idx": 1, "title": "View Over 70 Personal Checks Designs at Low Prices", "date": "", "ddg_snippet": "Order your favorite personal checks online at low prices. Checks .com features a variety of over 70 personal check designs to fit your unique personality.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/c/129/personal-checks", "content": "Order your favorite personal checks online at low prices. Checks .com features a variety of over 70 personal check designs to fit your unique personality."}
+{"idx": 2, "title": "Value Checks - Order Our Best Value Checks Online", "date": "", "ddg_snippet": "Order value-priced personal checks online starting at just $8.20 per box at Checks .com! We have a variety of inexpensive check designs to choose from, so you can find the perfect one to match your style and budget.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/c/323/value-checks", "content": "Order value-priced personal checks online starting at just $8.20 per box at Checks .com! We have a variety of inexpensive check designs to choose from, so you can find the perfect one to match your style and budget."}
+{"idx": 3, "title": "How To Order Checks Online | Checks.com", "date": "", "ddg_snippet": "Order checks , address labels and checkbook covers online, secure, fast and easy with Checks .com. Customize and preview your checks before you order.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/how-to-order-checks", "content": "Order checks , address labels and checkbook covers online, secure, fast and easy with Checks .com. Customize and preview your checks before you order."}
+{"idx": 4, "title": "Traditional Checks - Order Affordable Personal Checks Online", "date": "", "ddg_snippet": "Checks .com's collection of traditional checks offers the widest variety of designs where you're sure to find a favorite or two. Also check out our line of Choice Checks for our most exclusive and most secure check designs, or browse our cheaper-priced checks for an incredibly low price per box.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/c/334/traditional-checks", "content": "Checks .com's collection of traditional checks offers the widest variety of designs where you're sure to find a favorite or two. Also check out our line of Choice Checks for our most exclusive and most secure check designs, or browse our cheaper-priced checks for an incredibly low price per box."}
+{"idx": 5, "title": "Reorder Discounted Personal Checks Online", "date": "", "ddg_snippet": "At Checks .com, whenever you reorder checks , you pay the same low price as intro customers. Reorder checks for discounted prices online today!", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/quickreorder", "content": "At Checks .com, whenever you reorder checks , you pay the same low price as intro customers. Reorder checks for discounted prices online today!"}
+{"idx": 6, "title": "Renaissance Checks - Order Discounted Personal Checks", "date": "", "ddg_snippet": "Artistic scrolls and elegance grace these renaissance checks in a rich four color rotation. Coordinating renaissance address labels and checkbook cover are also available.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/p/834/renaissance-checks", "content": "Artistic scrolls and elegance grace these renaissance checks in a rich four color rotation. Coordinating renaissance address labels and checkbook cover are also available."}
+{"idx": 7, "title": "Modern Checks - Order Cool & Contemporary Personal Checks", "date": "", "ddg_snippet": "Show your trendy style with contemporary and cool personal checks from Checks .com! Choose a stylish check design or pattern and order checks online today!", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/c/274/modern-checks", "content": "Show your trendy style with contemporary and cool personal checks from Checks .com! Choose a stylish check design or pattern and order checks online today!"}
+{"idx": 8, "title": "Pretty in Pink Checks - Order Discounted Personal Checks", "date": "", "ddg_snippet": "Discover the perfect blend of Parisian style and elegance with our Pretty in Pink checks . Designed for those who live life with flair, these top-tear pink personal checks are available in four vibrant designs featuring stripes, polka dots and diamond patterns.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/p/1799/pretty-in-pink-checks", "content": "Discover the perfect blend of Parisian style and elegance with our Pretty in Pink checks . Designed for those who live life with flair, these top-tear pink personal checks are available in four vibrant designs featuring stripes, polka dots and diamond patterns."}
+{"idx": 9, "title": "Horse Checks - Order Discounted Personal Checks", "date": "", "ddg_snippet": "Horse Play Checks Photographs of galloping horses adorn this equine rotation of stallions, mustangs and foals. Coordinating horse return address labels are available.", "subpage_snippet": "", "source": "www.checks.com", "link": "https://www.checks.com/p/811/horse-play-checks", "content": "Horse Play Checks Photographs of galloping horses adorn this equine rotation of stallions, mustangs and foals. Coordinating horse return address labels are available."}
diff --git a/data/sampled_jsons/Concept_Bottleneck_Models_Without_Predefined_Concepts_abstract.jsonl b/data/sampled_jsons/Concept_Bottleneck_Models_Without_Predefined_Concepts_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..791b9dd9e95882e4bdbf6371606295e5f11e4be9
--- /dev/null
+++ b/data/sampled_jsons/Concept_Bottleneck_Models_Without_Predefined_Concepts_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Concept Bottleneck Models Without Predefined Concepts", "date": "", "ddg_snippet": "by S Schrodi · 2024 · Cited by 8 — Abstract:There has been considerable recent interest in interpretable concept-based models such as Concept Bottleneck Models (CBMs), ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2407.03921", "content": "by S Schrodi · 2024 · Cited by 8 — Abstract:There has been considerable recent interest in interpretable concept-based models such as Concept Bottleneck Models (CBMs), ..."}
+{"idx": 1, "title": "Selective Concept Bottleneck Models Without Predefined ...", "date": "", "ddg_snippet": "by S Schrodi — We present a method to convert black-box models into concept bottleneck models without predefined concepts and introduce an input-dependent concept selection ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=uuvujfQXZy", "content": "by S Schrodi — We present a method to convert black-box models into concept bottleneck models without predefined concepts and introduce an input-dependent concept selection ..."}
+{"idx": 2, "title": "Concept Bottleneck Models Without Predefined Concepts", "date": "", "ddg_snippet": "4 Jul 2024 — One popular framework are Concept Bottleneck Models (CBMs ) [1] , i.e., models that first predicts human-interpretable concepts and then use ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2407.03921v1", "content": "4 Jul 2024 — One popular framework are Concept Bottleneck Models (CBMs ) [1] , i.e., models that first predicts human-interpretable concepts and then use ..."}
+{"idx": 3, "title": "Selective Concept Bottleneck Models Without Predefined ...", "date": "", "ddg_snippet": "In this work, we discover the concepts encoded by the model through unsupervised concept discovery techniques instead. We further leverage a simple input- ...", "subpage_snippet": "", "source": "tmlr.infinite-conf.org", "link": "https://tmlr.infinite-conf.org/paper_pages/PMO30TLI4l.html", "content": "In this work, we discover the concepts encoded by the model through unsupervised concept discovery techniques instead. We further leverage a simple input- ..."}
+{"idx": 4, "title": "Hierarchical concept Bottleneck models for vision and their ...", "date": "", "ddg_snippet": "by F Pittino · 2023 · Cited by 12 — In this work we develop and discuss a Concept Bottleneck model for images segmentation, objects fine classification and tracking, and compare it to more ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0952197622006649", "content": "by F Pittino · 2023 · Cited by 12 — In this work we develop and discuss a Concept Bottleneck model for images segmentation, objects fine classification and tracking, and compare it to more ..."}
+{"idx": 5, "title": "Cross-Modal Conceptualization in Bottleneck Models", "date": "", "ddg_snippet": "by D Alukaev · 2023 · Cited by 7 — The objective of this study is to enhance the prac- ticality of Concept Bottleneck Models by elimi- nating the need for predefined concepts ...", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2023.emnlp-main.318.pdf", "content": "by D Alukaev · 2023 · Cited by 7 — The objective of this study is to enhance the prac- ticality of Concept Bottleneck Models by elimi- nating the need for predefined concepts ..."}
+{"idx": 6, "title": "improving the predictability of concept bottleneck models", "date": "", "ddg_snippet": "We establish theoretical results for concept - bottleneck models (CBMs), revealing how these properties guide the design of concept sets that optimize model ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2024/poster/93623", "content": "We establish theoretical results for concept - bottleneck models (CBMs), revealing how these properties guide the design of concept sets that optimize model ..."}
+{"idx": 7, "title": "Learning Bottleneck Concepts in Image Classification", "date": "", "ddg_snippet": "by B Wang · 2023 · Cited by 88 — This paper proposes Bot- tleneck Concept Learner (BotCL), which represents an im- age solely by the presence/absence of concepts learned through training over ...", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2023/papers/Wang_Learning_Bottleneck_Concepts_in_Image_Classification_CVPR_2023_paper.pdf", "content": "by B Wang · 2023 · Cited by 88 — This paper proposes Bot- tleneck Concept Learner (BotCL), which represents an im- age solely by the presence/absence of concepts learned through training over ..."}
+{"idx": 8, "title": "LABEL-FREE CONCEPT BOTTLENECK MODELS", "date": "", "ddg_snippet": "Given a neural network backbone, Label-free CBM transforms the backbone into an interpretable CBM without the need of concept labels with the following 4 steps ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/f6e0ef2be0578f171a67040f6fc6873a70f1ac9c.pdf", "content": "Given a neural network backbone, Label-free CBM transforms the backbone into an interpretable CBM without the need of concept labels with the following 4 steps ..."}
+{"idx": 9, "title": "Label-Free Concept Bottleneck Models", "date": "", "ddg_snippet": "by T Oikarinen · 2023 · Cited by 264 — A novel framework to transform any neural network into an interpretable CBM without labeled concept data, while retaining a high accuracy.", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "https://ui.adsabs.harvard.edu/abs/2023arXiv230406129O/abstract", "content": "by T Oikarinen · 2023 · Cited by 264 — A novel framework to transform any neural network into an interpretable CBM without labeled concept data, while retaining a high accuracy."}
diff --git a/data/sampled_jsons/Coordinate-wise_Private_Median_and_the_l_infinity_Guarantee_challenge_regression_solution_vector.jsonl b/data/sampled_jsons/Coordinate-wise_Private_Median_and_the_l_infinity_Guarantee_challenge_regression_solution_vector.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e96abaa26082bf6898aadab693e7c76709ed2cc2
--- /dev/null
+++ b/data/sampled_jsons/Coordinate-wise_Private_Median_and_the_l_infinity_Guarantee_challenge_regression_solution_vector.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Approximation guarantees of Median Mechanism in $\\\\mathbb{R}^d$ A Nearly-Optimal Bound for Fast Regression with $\\\\ell ... Coordinate-Wise Median in Byzantine Federated Learning Approximation Guarantees of Median Mechanism in Rd arXiv:2502.08578v2 [cs.GT] 13 Feb 2025 Communication-efficient and Byzantine-robust distributed ... Optimality of the coordinate-wise median mechanism for ...", "date": "", "ddg_snippet": "Feb 12, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d -dimensional spaces. Apr 24, 2023 · Moreover, we develop a novel analytical framework for $\\ell_\\infty$ guarantee regression that utilizes the Oblivious Coordinate-wise Embedding (OCE) property introduced in [Song and Yu, ICML'21]. Our analysis is much simpler and more general than that of [Price, Song and Woodruff, ICALP'17]. Abstract Federated learning enables clients to train local models on private data while exchanging model updates only. A key step in this pro-cess is gradient aggregation. We investigate the coordinate-wise median as an aggregation rule in both centralized and decentralized federated learning under Byzantine failures. In order to lower the communication cost in the decentralized setting, we ... The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approx-imation ratio in -dimensional spaces. Feb 13, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d-dimensional spaces. √ The best known approximation guarantee in May 1, 2023 · One algorithm is based on communication-efficient surrogate likelihood framework, another algorithm integrates approximate proximal point technique into the first algorithm. The two algorithms both explore two robust local gradient aggregation schemes: coordinate-wise median and coordinate-wise trimmed mean. We show that the trimmed-mean-based We refer to the generalized coordinate-wise median mechanism with 0 constant points and the standard coordinate -system as the coordinate-wise median mechanism and denote it by c(x).", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2502.08578", "content": "Feb 12, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d -dimensional spaces. Apr 24, 2023 · Moreover, we develop a novel analytical framework for $\\ell_\\infty$ guarantee regression that utilizes the Oblivious Coordinate-wise Embedding (OCE) property introduced in [Song and Yu, ICML'21]. Our analysis is much simpler and more general than that of [Price, Song and Woodruff, ICALP'17]. Abstract Federated learning enables clients to train local models on private data while exchanging model updates only. A key step in this pro-cess is gradient aggregation. We investigate the coordinate-wise median as an aggregation rule in both centralized and decentralized federated learning under Byzantine failures. In order to lower the communication cost in the decentralized setting, we ... The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approx-imation ratio in -dimensional spaces. Feb 13, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d-dimensional spaces. √ The best known approximation guarantee in May 1, 2023 · One algorithm is based on communication-efficient surrogate likelihood framework, another algorithm integrates approximate proximal point technique into the first algorithm. The two algorithms both explore two robust local gradient aggregation schemes: coordinate-wise median and coordinate-wise trimmed mean. We show that the trimmed-mean-based We refer to the generalized coordinate-wise median mechanism with 0 constant points and the standard coordinate -system as the coordinate-wise median mechanism and denote it by c(x)."}
+{"idx": 1, "title": "A Nearly-Optimal Bound for Fast Regression with $\\\\ell ...", "date": "", "ddg_snippet": "Apr 24, 2023 · Moreover, we develop a novel analytical framework for $\\ell_\\infty$ guarantee regression that utilizes the Oblivious Coordinate-wise Embedding (OCE) property introduced in [Song and Yu, ICML'21]. Our analysis is much simpler and more general than that of [Price, Song and Woodruff, ICALP'17].", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=cJh37mrFms", "content": "Apr 24, 2023 · Moreover, we develop a novel analytical framework for $\\ell_\\infty$ guarantee regression that utilizes the Oblivious Coordinate-wise Embedding (OCE) property introduced in [Song and Yu, ICML'21]. Our analysis is much simpler and more general than that of [Price, Song and Woodruff, ICALP'17]."}
+{"idx": 2, "title": "Coordinate-Wise Median in Byzantine Federated Learning", "date": "", "ddg_snippet": "Abstract Federated learning enables clients to train local models on private data while exchanging model updates only. A key step in this pro-cess is gradient aggregation. We investigate the coordinate-wise median as an aggregation rule in both centralized and decentralized federated learning under Byzantine failures. In order to lower the communication cost in the decentralized setting, we ...", "subpage_snippet": "", "source": "schmiste.github.io", "link": "https://schmiste.github.io/asiaccs25.pdf", "content": "Abstract Federated learning enables clients to train local models on private data while exchanging model updates only. A key step in this pro-cess is gradient aggregation. We investigate the coordinate-wise median as an aggregation rule in both centralized and decentralized federated learning under Byzantine failures. In order to lower the communication cost in the decentralized setting, we ..."}
+{"idx": 3, "title": "Approximation Guarantees of Median Mechanism in Rd", "date": "", "ddg_snippet": "The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approx-imation ratio in -dimensional spaces.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/epdf/10.1145/3717823.3718156", "content": "The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approx-imation ratio in -dimensional spaces."}
+{"idx": 4, "title": "Communication-efficient and Byzantine-robust distributed ...", "date": "", "ddg_snippet": "May 1, 2023 · One algorithm is based on communication-efficient surrogate likelihood framework, another algorithm integrates approximate proximal point technique into the first algorithm. The two algorithms both explore two robust local gradient aggregation schemes: coordinate-wise median and coordinate-wise trimmed mean. We show that the trimmed-mean-based", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0031320323000134", "content": "May 1, 2023 · One algorithm is based on communication-efficient surrogate likelihood framework, another algorithm integrates approximate proximal point technique into the first algorithm. The two algorithms both explore two robust local gradient aggregation schemes: coordinate-wise median and coordinate-wise trimmed mean. We show that the trimmed-mean-based"}
+{"idx": 5, "title": "Optimality of the coordinate-wise median mechanism for ...", "date": "", "ddg_snippet": "We refer to the generalized coordinate-wise median mechanism with 0 constant points and the standard coordinate -system as the coordinate-wise median mechanism and denote it by c(x).", "subpage_snippet": "", "source": "whanncar.github.io", "link": "https://whanncar.github.io/papers/median.pdf", "content": "We refer to the generalized coordinate-wise median mechanism with 0 constant points and the standard coordinate -system as the coordinate-wise median mechanism and denote it by c(x)."}
+{"idx": 6, "title": "Efficient Sparse Least Absolute Deviation Regression with", "date": "", "ddg_snippet": "... by reformulating the sparse LAD problem as a penalized least square estimation problem and adopts a three-stage noise injection to guarantee the ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2401.01294v1", "content": "... by reformulating the sparse LAD problem as a penalized least square estimation problem and adopts a three-stage noise injection to guarantee the ..."}
+{"idx": 7, "title": "Real-time Bus Travel Time Prediction and Reliability", "date": "", "ddg_snippet": "However, this remains a challenging task due to the limitations of existing models and data sources. ... the parameters of the distributions were ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.05907v1", "content": "However, this remains a challenging task due to the limitations of existing models and data sources. ... the parameters of the distributions were ..."}
+{"idx": 8, "title": "Approximation, Randomization, and Combinatorial Optimization.", "date": "", "ddg_snippet": "... them can a carpenter cut out using a panel saw making guillotine cuts, i.e., only moving forward through the material along a straight line until it ...", "subpage_snippet": "", "source": "drops.dagstuhl.de", "link": "https://drops.dagstuhl.de/entities/volume/LIPIcs-volume-40", "content": "... them can a carpenter cut out using a panel saw making guillotine cuts, i.e., only moving forward through the material along a straight line until it ..."}
+{"idx": 9, "title": "arXiv:2502.08578v2 [cs.GT] 13 Feb 2025", "date": "", "ddg_snippet": "Feb 13, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d-dimensional spaces. √ The best known approximation guarantee in", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2502.08578", "content": "Feb 13, 2025 · The coordinate-wise median is a classic and most well-studied strategy-proof mechanism in social choice and facility location scenarios. Surprisingly, there is no systematic study of its approximation ratio in d-dimensional spaces. √ The best known approximation guarantee in"}
diff --git a/data/sampled_jsons/Crocker_stacks_persistent_homology.jsonl b/data/sampled_jsons/Crocker_stacks_persistent_homology.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cd00efaf2daf8fc3038642f512f5f0f5d29f66e5
--- /dev/null
+++ b/data/sampled_jsons/Crocker_stacks_persistent_homology.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Persistent homology approach for uncovering transitions to Chaos", "date": "", "ddg_snippet": "Conclusion This study demonstrates the effectiveness of using persistent homology to uncover topological insights in the time series of dynamic systems. By varying the control parameter, we observed different transitions in the behavior of the system, which are reflected in the time series.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0960077925000670", "content": "Conclusion This study demonstrates the effectiveness of using persistent homology to uncover topological insights in the time series of dynamic systems. By varying the control parameter, we observed different transitions in the behavior of the system, which are reflected in the time series."}
+{"idx": 1, "title": "Persistent Topological Features in Large Language Models", "date": "", "ddg_snippet": "Differently than standard persistent homology , short- and long-lived features represent how the model dynamically evolves. Short-lived features indicate a high rate of rearrangement of the points x i between adjacent layers, while long-lived features suggest a phase of retention of (relative) positions in the model.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.11042v1", "content": "Differently than standard persistent homology , short- and long-lived features represent how the model dynamically evolves. Short-lived features indicate a high rate of rearrangement of the points x i between adjacent layers, while long-lived features suggest a phase of retention of (relative) positions in the model."}
+{"idx": 2, "title": "A roadmap for the computation of persistent homology - PMC", "date": "", "ddg_snippet": "Abstract Persistent homology (PH) is a method used in topological data analysis (TDA) to study qualitative features of data that persist across multiple scales. It is robust to perturbations of input data, independent of dimensions and coordinates, and provides a compact representation of the qualitative features of the input. The computation of PH is an open area with numerous important and ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6979512/", "content": "Abstract Persistent homology (PH) is a method used in topological data analysis (TDA) to study qualitative features of data that persist across multiple scales. It is robust to perturbations of input data, independent of dimensions and coordinates, and provides a compact representation of the qualitative features of the input. The computation of PH is an open area with numerous important and ..."}
+{"idx": 3, "title": "A Case Study on Identifying Bifurcation and Chaos with CROCKER Plots", "date": "", "ddg_snippet": "A proposed persistent homology point summary, conformance, was able to identify models with successful dynamics learning and detect discrepancies in the manifold density distribution.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/359937143_A_Case_Study_on_Identifying_Bifurcation_and_Chaos_with_CROCKER_Plots", "content": "A proposed persistent homology point summary, conformance, was able to identify models with successful dynamics learning and detect discrepancies in the manifold density distribution."}
+{"idx": 4, "title": "TDA: Applications and Deep Dive into Persistent Homology", "date": "", "ddg_snippet": "This article aims to introduce readers to the basics of Topological Data Analysis (TDA), focusing on persistent homology and its applications. This concept is crucial because it serves as the ...", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/@kennywang2003/topological-data-analysis-persistent-homology-22636f4e34bf", "content": "This article aims to introduce readers to the basics of Topological Data Analysis (TDA), focusing on persistent homology and its applications. This concept is crucial because it serves as the ..."}
+{"idx": 5, "title": "Persistent homology approach for uncovering transitions to Chaos", "date": "", "ddg_snippet": "The CROCKER plot is an interactive visualization technique used to track changes in the homology of data over varying parameters. CROCKER stands for \"Contour Realization Of Computed k -dimensional hole Evolution in the VR complex\" [44] and provides a comprehensive visual representation of PH.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0960077925000670", "content": "The CROCKER plot is an interactive visualization technique used to track changes in the homology of data over varying parameters. CROCKER stands for \"Contour Realization Of Computed k -dimensional hole Evolution in the VR complex\" [44] and provides a comprehensive visual representation of PH."}
+{"idx": 6, "title": "Capturing dynamics of time-varying data via topology", "date": "", "ddg_snippet": "Crocker stacks are convenient for visualization, amenable to machine learning, and satisfy a desirable continuity property which we prove. We demonstrate the utility of crocker stacks for a parameter identification task involving an influential model of biological aggregations [57].", "subpage_snippet": "", "source": "www.aimsciences.org", "link": "https://www.aimsciences.org/article/id/2acaee54-6688-46a4-b35d-447f84c4c691", "content": "Crocker stacks are convenient for visualization, amenable to machine learning, and satisfy a desirable continuity property which we prove. We demonstrate the utility of crocker stacks for a parameter identification task involving an influential model of biological aggregations [57]."}
+{"idx": 7, "title": "Lu Xian Henry Adams Chad M. Topaz Lori Ziegelmeier", "date": "", "ddg_snippet": "these are the same components as scale ε varies. However, as we will show in Section 6, since the crocker stack contains enough information to recover the persistent homology barcodes, one can confirm these are the same connected components as ε varies by considerin", "subpage_snippet": "", "source": "par.nsf.gov", "link": "https://par.nsf.gov/servlets/purl/10388958", "content": "these are the same components as scale ε varies. However, as we will show in Section 6, since the crocker stack contains enough information to recover the persistent homology barcodes, one can confirm these are the same connected components as ε varies by considerin"}
+{"idx": 8, "title": "PDF Neural Persistence Dynamics - proceedings.neurips.cc", "date": "", "ddg_snippet": "In the following, we present a runtime breakdown of the pre-processing steps (i.e., Vietoris-Rips persistent homology (PH) computation, and the vectorization of persistence diagrams), as well as a runtime comparison to prior work (PSK and crocker stacks ) and our baseline approach.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2024/file/3a509449a73fd0aab8c0cf5705827036-Paper-Conference.pdf", "content": "In the following, we present a runtime breakdown of the pre-processing steps (i.e., Vietoris-Rips persistent homology (PH) computation, and the vectorization of persistence diagrams), as well as a runtime comparison to prior work (PSK and crocker stacks ) and our baseline approach."}
+{"idx": 9, "title": "PDF Z-GCNETs: Time Zigzags at Graph Convolutional Networks for Time Series ...", "date": "", "ddg_snippet": "An alternative to zigzag but a closely related approach to assess properties of time-varying data with per-sistent homology , namely, crocker stacks , has been recently suggested by Xian et al. (2020), though the crocker stacks representations are not learnable in DL models.", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v139/chen21o/chen21o.pdf", "content": "An alternative to zigzag but a closely related approach to assess properties of time-varying data with per-sistent homology , namely, crocker stacks , has been recently suggested by Xian et al. (2020), though the crocker stacks representations are not learnable in DL models."}
diff --git a/data/sampled_jsons/Curriculum_Learning_Physics-Informed_Neural_Networks_year_2023.jsonl b/data/sampled_jsons/Curriculum_Learning_Physics-Informed_Neural_Networks_year_2023.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..db59213bb3a015c44ba4b24c39e2e6e78a8626d0
--- /dev/null
+++ b/data/sampled_jsons/Curriculum_Learning_Physics-Informed_Neural_Networks_year_2023.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Meta-learning (computer science) - Wikipedia", "date": "", "ddg_snippet": "A Memory-Augmented Neural Network , or MANN for short, is claimed to be able to encode new information quickly and thus to adapt to new tasks after ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Meta-learning_(computer_science)", "content": "A Memory-Augmented Neural Network , or MANN for short, is claimed to be able to encode new information quickly and thus to adapt to new tasks after ..."}
+{"idx": 1, "title": "Machine learning - Wikipedia", "date": "", "ddg_snippet": "1 Within a subdiscipline in machine learning , advances in the field of deep learning have allowed neural networks , a class of statistical algorithms ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Machine_learning", "content": "1 Within a subdiscipline in machine learning , advances in the field of deep learning have allowed neural networks , a class of statistical algorithms ..."}
+{"idx": 2, "title": "Self-supervised learning - Wikipedia", "date": "", "ddg_snippet": "In the context of neural networks , self-supervised learning aims to leverage inherent structures or relationships within the input data to create ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Self-supervised_learning", "content": "In the context of neural networks , self-supervised learning aims to leverage inherent structures or relationships within the input data to create ..."}
+{"idx": 3, "title": "Feature learning - Wikipedia", "date": "", "ddg_snippet": "Multilayer neural networks can be used to perform feature learning , since they learn a representation of their input at the hidden layer(s) which is ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Feature_learning", "content": "Multilayer neural networks can be used to perform feature learning , since they learn a representation of their input at the hidden layer(s) which is ..."}
+{"idx": 4, "title": "Conference on Neural Information Processing Systems - Wikipedia", "date": "", "ddg_snippet": "The first proceedings was published in book form by the American Institute of Physics in 1987, and was entitled Neural Information Processing Systems ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Conference_on_Neural_Information_Processing_Systems", "content": "The first proceedings was published in book form by the American Institute of Physics in 1987, and was entitled Neural Information Processing Systems ..."}
+{"idx": 5, "title": "DataPro | 0 articles | Packt Learning Hub", "date": "", "ddg_snippet": "Its architecture includes novel training strategies like GRPO, curriculum learning , and expert initialization, making it a new benchmark in ...", "subpage_snippet": "", "source": "www.packtpub.com", "link": "https://www.packtpub.com/en-ic/learning/datapro", "content": "Its architecture includes novel training strategies like GRPO, curriculum learning , and expert initialization, making it a new benchmark in ..."}
+{"idx": 6, "title": "Chen Xu | DeepAI", "date": "", "ddg_snippet": "Transfer learning based physics - informed neural networks for solving inverse problems in tunneling ... learning methods called physics - informed ne ...", "subpage_snippet": "", "source": "deepai.org", "link": "https://deepai.org/profile/chen-xu-6525", "content": "Transfer learning based physics - informed neural networks for solving inverse problems in tunneling ... learning methods called physics - informed ne ..."}
+{"idx": 7, "title": "Shiji Song | DeepAI", "date": "", "ddg_snippet": "Learning Specialized Activation Functions for Physics - informed Neural ... Physics - informed neural networks (PINNs) are known to suffer from optimi...", "subpage_snippet": "", "source": "deepai.org", "link": "https://deepai.org/profile/shiji-song", "content": "Learning Specialized Activation Functions for Physics - informed Neural ... Physics - informed neural networks (PINNs) are known to suffer from optimi..."}
+{"idx": 8, "title": "ICLR 2024 Schedule", "date": "", "ddg_snippet": "Going Beyond Neural Network Feature Similarity: The Network Feature Complexity and Its Interpretation Using Category Theory", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2024/calendar", "content": "Going Beyond Neural Network Feature Similarity: The Network Feature Complexity and Its Interpretation Using Category Theory"}
+{"idx": 9, "title": "Downloads", "date": "", "ddg_snippet": "... for Multi-Objective Reinforcement Learning ... Adversarial Robustness without Adversarial Training: A Teacher-Guided Curriculum Learning Approach", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/Downloads/2021", "content": "... for Multi-Objective Reinforcement Learning ... Adversarial Robustness without Adversarial Training: A Teacher-Guided Curriculum Learning Approach"}
diff --git a/data/sampled_jsons/Cybench_A_Framework_for_Evaluating_Cybersecurity_Capabilities_abstract.jsonl b/data/sampled_jsons/Cybench_A_Framework_for_Evaluating_Cybersecurity_Capabilities_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..65643d537ec10181ca08fb3ee75627b381d463d5
--- /dev/null
+++ b/data/sampled_jsons/Cybench_A_Framework_for_Evaluating_Cybersecurity_Capabilities_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2408.08926] Cybench : A Framework for Evaluating Cybersecurity ...", "date": "", "ddg_snippet": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2408.08926", "content": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful..."}
+{"idx": 1, "title": "(PDF) Cybench : A Framework for Evaluating Cybersecurity ...", "date": "", "ddg_snippet": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/383236134_Cybench_A_Framework_for_Evaluating_Cybersecurity_Capabilities_and_Risk_of_Language_Models", "content": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful..."}
+{"idx": 2, "title": "Cybench : A Framework for Evaluating Cybersecurity Capabilities ...", "date": "", "ddg_snippet": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks.1 We include 40 professionallevel Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful, and spanning a wide range...", "subpage_snippet": "", "source": "ee.stanford.edu", "link": "https://ee.stanford.edu/cybench-framework-evaluating-cybersecurity-capabilities-and-risks-language-models", "content": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks.1 We include 40 professionallevel Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful, and spanning a wide range..."}
+{"idx": 3, "title": "Cybench : A Framework for Evaluating Cybersecurity Capabilities ...", "date": "", "ddg_snippet": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=tc90LV0yRL", "content": "Toward that end, we introduce Cybench , a framework for specifying cybersecurity tasks and evaluating agents on those tasks. We include 40 professional-level Capture the Flag (CTF) tasks from 4 distinct CTF competitions, chosen to be recent, meaningful..."}
+{"idx": 4, "title": "GitHub - andyzorigin/ cybench", "date": "", "ddg_snippet": "Cybench : A Framework for Evaluating Cybersecurity Capabilities and Risk. This repository contains code used in the Cybench project (website, paper), which serves to evaluate agent capabilities and risks on cybersecurity tasks. Cybench includes the following features", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/andyzorigin/cybench", "content": "Cybench : A Framework for Evaluating Cybersecurity Capabilities and Risk. This repository contains code used in the Cybench project (website, paper), which serves to evaluate agent capabilities and risks on cybersecurity tasks. Cybench includes the following features"}
+{"idx": 5, "title": "Cybench : A Framework for Evaluating Cybersecurity Capabilities ...", "date": "", "ddg_snippet": "This paper presents Cybench , a framework designed to evaluate the cybersecurity capabilities and risks of language models (LMs) in identifying vulnerabilities and executing exploits.", "subpage_snippet": "", "source": "ai-search.io", "link": "https://ai-search.io/papers/cybench-a-framework-for-evaluating-cybersecurity-capabilities-and-risk-of-language-models", "content": "This paper presents Cybench , a framework designed to evaluate the cybersecurity capabilities and risks of language models (LMs) in identifying vulnerabilities and executing exploits."}
+{"idx": 6, "title": "Can LLM Agents Compete in CTFs?. Evaluating LLM... | Medium", "date": "", "ddg_snippet": "Cybench is a framework designed to evaluate language models' (LMs) ability to solve cybersecurity tasks like Capture the Flag (CTF) challenges. It allows AI agents to identify vulnerabilities, execute exploits, and solve tasks in a structured environment.", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/@arohablue/can-ai-compete-in-ctf-b6d9016ff254", "content": "Cybench is a framework designed to evaluate language models' (LMs) ability to solve cybersecurity tasks like Capture the Flag (CTF) challenges. It allows AI agents to identify vulnerabilities, execute exploits, and solve tasks in a structured environment."}
+{"idx": 7, "title": "Google News - Cybersecurity threats to language models - Overview", "date": "", "ddg_snippet": "Cybench : A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models.", "subpage_snippet": "", "source": "news.google.com", "link": "https://news.google.com/stories/CAAqNggKIjBDQklTSGpvSmMzUnZjbmt0TXpZd1NoRUtEd2kxcllQNERSSGxSZ0JldXB4UE1TZ0FQAQ?hl=en-US&gl=US&ceid=US:en", "content": "Cybench : A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models."}
+{"idx": 8, "title": "AI Task Length Horizons in Offensive Cybersecurity — LessWrong", "date": "", "ddg_snippet": "I asked the same question for offensive cybersecurity , a domain with distinct skills and unique AI-safety implications. Using METR’s methodology on five cyber benchmarks, with tasks ranging from 0.5s to 25h in human-expert estimated times...", "subpage_snippet": "", "source": "www.lesswrong.com", "link": "https://www.lesswrong.com/posts/fjgYkTWKAXSxsxdsj/preload", "content": "I asked the same question for offensive cybersecurity , a domain with distinct skills and unique AI-safety implications. Using METR’s methodology on five cyber benchmarks, with tasks ranging from 0.5s to 25h in human-expert estimated times..."}
+{"idx": 9, "title": "2025.02.20 xAI Risk Management Framework Draft - Google Docs", "date": "", "ddg_snippet": "capabilities for benign or beneficial purposes, such as scientifically investigating Grok’s capabilities for risk assessment purposes, or if such requests cover information that is already readily and easily available, including by an internet search.", "subpage_snippet": "", "source": "x.ai", "link": "https://x.ai/documents/2025.02.20-RMF-Draft.pdf", "content": "capabilities for benign or beneficial purposes, such as scientifically investigating Grok’s capabilities for risk assessment purposes, or if such requests cover information that is already readily and easily available, including by an internet search."}
diff --git a/data/sampled_jsons/DCBM_GitHub_min_area_max_area_area_threshold.jsonl b/data/sampled_jsons/DCBM_GitHub_min_area_max_area_area_threshold.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cb6ec15d46167eb5e5f8e4168ea6a41cb0c55802
--- /dev/null
+++ b/data/sampled_jsons/DCBM_GitHub_min_area_max_area_area_threshold.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "GitHub - deepopo/DCBM", "date": "", "ddg_snippet": "This is an implementation of the IEEE TPAMI paper The Decoupling Concept Bottleneck Model ( DCBM ). The vision-language-model (VLM) part is being refined and will be available soon. Figure 1: DCBM Pipeline. (A) DCBM for prediction and interpretation. (B) DCBM for human-machine interaction, including forward intervention and backward rectification.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/deepopo/DCBM", "content": "This is an implementation of the IEEE TPAMI paper The Decoupling Concept Bottleneck Model ( DCBM ). The vision-language-model (VLM) part is being refined and will be available soon. Figure 1: DCBM Pipeline. (A) DCBM for prediction and interpretation. (B) DCBM for human-machine interaction, including forward intervention and backward rectification."}
+{"idx": 1, "title": "DCBM: Data-Efficient Visual Concept Bottleneck Models", "date": "", "ddg_snippet": "Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts. However, current CBMs typically rely on concept sets extracted from large language models or extensive image corpora, limiting their effectiveness in data-sparse scenarios. We propose Data-efficient CBMs ( DCBMs ), which reduce the need for large sample sizes ...", "subpage_snippet": "", "source": "kathpra.github.io", "link": "https://kathpra.github.io/DCBM/", "content": "Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts. However, current CBMs typically rely on concept sets extracted from large language models or extensive image corpora, limiting their effectiveness in data-sparse scenarios. We propose Data-efficient CBMs ( DCBMs ), which reduce the need for large sample sizes ..."}
+{"idx": 2, "title": "DCBM/main.py at main · deepopo/DCBM · GitHub", "date": "", "ddg_snippet": "Contribute to deepopo/ DCBM development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/deepopo/DCBM/blob/main/main.py", "content": "Contribute to deepopo/ DCBM development by creating an account on GitHub ."}
+{"idx": 3, "title": "[Support]: Min Max Ratio help · Issue #8350 - GitHub", "date": "", "ddg_snippet": "person: min_area : 18000 # max_area : 100000 # min_ratio: 0.5 how is this worked out ? # max_ratio: 2.0 how is this worked out ? # min_score: 0.5 threshold: 0.75 Also on a separate question can I run a Google coral on top of using my Nvidia gpu or is it one or the other? Version 12.1 Frigate config file ui:", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/blakeblackshear/frigate/issues/8350", "content": "person: min_area : 18000 # max_area : 100000 # min_ratio: 0.5 how is this worked out ? # max_ratio: 2.0 how is this worked out ? # min_score: 0.5 threshold: 0.75 Also on a separate question can I run a Google coral on top of using my Nvidia gpu or is it one or the other? Version 12.1 Frigate config file ui:"}
+{"idx": 4, "title": "Releases · KathPra/DCBM · GitHub", "date": "", "ddg_snippet": "Official repo for ICML25 paper: DCBM : Data-Efficient Visual Concept Bottleneck Models - Releases · KathPra/ DCBM", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/KathPra/DCBM/releases", "content": "Official repo for ICML25 paper: DCBM : Data-Efficient Visual Concept Bottleneck Models - Releases · KathPra/ DCBM"}
+{"idx": 5, "title": "DCBM: Data-Efficient Visual Concept Bottleneck Models", "date": "", "ddg_snippet": "Abstract Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts. However, current CBMs typically rely on con-cept sets extracted from large language models or extensive image corpora, limiting their effec-tiveness in data-sparse scenarios. We propose Data-eficient CBMs ( DCBMs ), which reduce the need for large ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2412.11576v3", "content": "Abstract Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts. However, current CBMs typically rely on con-cept sets extracted from large language models or extensive image corpora, limiting their effec-tiveness in data-sparse scenarios. We propose Data-eficient CBMs ( DCBMs ), which reduce the need for large ..."}
+{"idx": 6, "title": "GitHub - fraspass/dcsbm: Spectral clustering on spherical coordinates ...", "date": "", "ddg_snippet": "The code in this repository can be used to reproduce the results and simulations in Sanna Passino, F., Heard, N. A., and Rubin-Delanchy, P. (2021) \"Spectral clustering on spherical coordinates under the degree-corrected stochastic blockmodel\", Technometrics (to appear, link to the arXiv publication).", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fraspass/dcsbm", "content": "The code in this repository can be used to reproduce the results and simulations in Sanna Passino, F., Heard, N. A., and Rubin-Delanchy, P. (2021) \"Spectral clustering on spherical coordinates under the degree-corrected stochastic blockmodel\", Technometrics (to appear, link to the arXiv publication)."}
+{"idx": 7, "title": "openmv学习之旅②之色块追踪算法的改善 - CSDN博客", "date": "", "ddg_snippet": "area_threshold 面积阈值，如果色块被框起来的面积小于这个值，会被过滤掉。 pixels_threshold 像素个数阈值，如果色块像素数量小于这个值，会被过滤掉 merge 合并，如果设置为True，那么合并所有重叠的blob为一个。 注意：这会合并所有的blob，无论是什么颜色的。", "subpage_snippet": "", "source": "blog.csdn.net", "link": "https://blog.csdn.net/jiejiemcu/article/details/80579893", "content": "area_threshold 面积阈值，如果色块被框起来的面积小于这个值，会被过滤掉。 pixels_threshold 像素个数阈值，如果色块像素数量小于这个值，会被过滤掉 merge 合并，如果设置为True，那么合并所有重叠的blob为一个。 注意：这会合并所有的blob，无论是什么颜色的。"}
+{"idx": 8, "title": "cvp_detailed.ipynb - Colab", "date": "", "ddg_snippet": "The min_area and max_area parameters here can help to look at a different set of plaques within the size range of the default value of these parameters can be used.", "subpage_snippet": "", "source": "colab.research.google.com", "link": "https://colab.research.google.com/github/plaque2/PyPlaque/blob/master/notebooks/cvp_detailed.ipynb", "content": "The min_area and max_area parameters here can help to look at a different set of plaques within the size range of the default value of these parameters can be used."}
+{"idx": 9, "title": "Exact Meaning of the parameters given to initialize MSER in OpenCV 2.4 ...", "date": "", "ddg_snippet": "OpenCV reference manual (2.4.x) states that the constructor that initializes MSER requires the following parameters: delta, min_area , max_area , max_variation, min_diversity, max_evolution, area_threshold , min_margin, edge_blur_size. I am dealing with grayscale images. What is the use of the parameters \"delta\", \"max_variation\" and \"min_diversity\"? What property of an MSER do these parameters ...", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/17647500/exact-meaning-of-the-parameters-given-to-initialize-mser-in-opencv-2-4-x", "content": "OpenCV reference manual (2.4.x) states that the constructor that initializes MSER requires the following parameters: delta, min_area , max_area , max_variation, min_diversity, max_evolution, area_threshold , min_margin, edge_blur_size. I am dealing with grayscale images. What is the use of the parameters \"delta\", \"max_variation\" and \"min_diversity\"? What property of an MSER do these parameters ..."}
diff --git a/data/sampled_jsons/Descriptor-In-Pixel_SCAMP-7_Computation_Time_Breakdown_Table_1_exact_values_numbers_percentages.jsonl b/data/sampled_jsons/Descriptor-In-Pixel_SCAMP-7_Computation_Time_Breakdown_Table_1_exact_values_numbers_percentages.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e947d53580a8bcc4ebd5399e005797dff7f7fac7
--- /dev/null
+++ b/data/sampled_jsons/Descriptor-In-Pixel_SCAMP-7_Computation_Time_Breakdown_Table_1_exact_values_numbers_percentages.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "PDF Descriptor-In-Pixel : Point-Feature Tracking For Pixel Processor Arrays", "date": "", "ddg_snippet": "A breakdown of the total computation time per frame is given in Table 1 . To put these numbers into perspective, the time taken to out-put a full uncompressed image from the sensor is well over 20000 μs, and even a 4-bit image requires 2700 μs.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Bose_Descriptor-In-Pixel__Point-Feature_Tracking_For_Pixel_Processor_Arrays_CVPR_2025_paper.pdf", "content": "A breakdown of the total computation time per frame is given in Table 1 . To put these numbers into perspective, the time taken to out-put a full uncompressed image from the sensor is well over 20000 μs, and even a 4-bit image requires 2700 μs."}
+{"idx": 1, "title": "PDF Live Demonstration: SCAMP-7 - GitHub Pages", "date": "", "ddg_snippet": "The program development environment will be demonstrated alongside discussion of specific programming techniques for pixel parallel processor arrays and the new features of the SCAMP-7 device. Figure 1 . SCAMP-5 system vs SCAMP-7 system.", "subpage_snippet": "", "source": "tub-rip.github.io", "link": "https://tub-rip.github.io/eventvision2023/papers/2023CVPRW_Live_Demonstration_Scamp-7.pdf", "content": "The program development environment will be demonstrated alongside discussion of specific programming techniques for pixel parallel processor arrays and the new features of the SCAMP-7 device. Figure 1 . SCAMP-5 system vs SCAMP-7 system."}
+{"idx": 2, "title": "SCAMP Vision Sensor - University of Manchester", "date": "", "ddg_snippet": "The SCAMP Vision Sensor integrates a massively parallel SIMD processor array into the pixels of the image sensor device. Unlike a conventional image sensor, it does not output raw images, but rather the results of on-sensor computations , for instance a feature map, optic flow map and/or address-events describing locations of pixels of interest. The device is fully programmable, to execute a ...", "subpage_snippet": "", "source": "personalpages.manchester.ac.uk", "link": "https://personalpages.manchester.ac.uk/staff/p.dudek/scamp/", "content": "The SCAMP Vision Sensor integrates a massively parallel SIMD processor array into the pixels of the image sensor device. Unlike a conventional image sensor, it does not output raw images, but rather the results of on-sensor computations , for instance a feature map, optic flow map and/or address-events describing locations of pixels of interest. The device is fully programmable, to execute a ..."}
+{"idx": 3, "title": "Live Demonstration: SCAMP-7 - IEEE Xplore", "date": "", "ddg_snippet": "We propose to demonstrate the next generation of the SCAMP vision sensor, SCAMP-7 , whose architecture embeds programmable processing cores into each pixel . This IC allows various computer vision tasks to be performed \"on-sensor\" at high speeds while avoiding the slow and power hungry transfer of image data off the device. Through in-pixel computations over the image array, visual information ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/10208514", "content": "We propose to demonstrate the next generation of the SCAMP vision sensor, SCAMP-7 , whose architecture embeds programmable processing cores into each pixel . This IC allows various computer vision tasks to be performed \"on-sensor\" at high speeds while avoiding the slow and power hungry transfer of image data off the device. Through in-pixel computations over the image array, visual information ..."}
+{"idx": 4, "title": "Bose Descriptor-In-Pixel Point-Feature Tracking For Pixel Processor ...", "date": "", "ddg_snippet": "This paper introduces a novel method for point-feature detection and tracking using Pixel Processor Array (PPA) vision sensors, which enables in-pixel computation and significantly reduces data transfer requirements. The proposed Descriptor-In-Pixel paradigm allows for efficient processing at over 3000 FPS, making it suitable for high-speed applications while maintaining low latency. By ...", "subpage_snippet": "", "source": "www.scribd.com", "link": "https://www.scribd.com/document/890539401/Bose-Descriptor-In-Pixel-Point-Feature-Tracking-for-Pixel-Processor-Arrays-CVPR-2025-Paper", "content": "This paper introduces a novel method for point-feature detection and tracking using Pixel Processor Array (PPA) vision sensors, which enables in-pixel computation and significantly reduces data transfer requirements. The proposed Descriptor-In-Pixel paradigm allows for efficient processing at over 3000 FPS, making it suitable for high-speed applications while maintaining low latency. By ..."}
+{"idx": 5, "title": "Live Demonstration: SCAMP-7 - computer.org", "date": "", "ddg_snippet": "We propose to demonstrate the next generation of the SCAMP vision sensor, SCAMP-7 , whose architecture embeds programmable processing cores into each pixel . This IC allows various computer vision tasks to be performed", "subpage_snippet": "", "source": "www.computer.org", "link": "https://www.computer.org/csdl/proceedings-article/cvprw/2023/024900d995/1PBy0sv9HtC", "content": "We propose to demonstrate the next generation of the SCAMP vision sensor, SCAMP-7 , whose architecture embeds programmable processing cores into each pixel . This IC allows various computer vision tasks to be performed"}
+{"idx": 6, "title": "SCAMP Vision Sensor", "date": "", "ddg_snippet": "The SCAMP vision sensor integrates a massively parallel SIMD processor array into the pixels of the image sensor device. It does not output raw images, like a conventional camera, but rather the results of on-sensor computations , for instance a feature map, optic flow map and/or address-events describing locations of pixels of interest. As costly early vision computations are done entirely on ...", "subpage_snippet": "", "source": "www.scamp-vision-chip.org", "link": "https://www.scamp-vision-chip.org/", "content": "The SCAMP vision sensor integrates a massively parallel SIMD processor array into the pixels of the image sensor device. It does not output raw images, like a conventional camera, but rather the results of on-sensor computations , for instance a feature map, optic flow map and/or address-events describing locations of pixels of interest. As costly early vision computations are done entirely on ..."}
+{"idx": 7, "title": "Descriptor In Pixel : Point Feature Tracking for Pixel ... - YouTube", "date": "", "ddg_snippet": "Instead of standard pixels , PPA sensors consists of thousands of \" pixel -processors\", enabling massive parallel computation of visual data at the point of light capture.", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=HIdQtf6mFSs", "content": "Instead of standard pixels , PPA sensors consists of thousands of \" pixel -processors\", enabling massive parallel computation of visual data at the point of light capture."}
+{"idx": 8, "title": "Live Demonstration: CNN Inference on the Focal Plane with a Pixel ...", "date": "", "ddg_snippet": "We present a novel method of CNN inference on a pixel processor array device, demonstrating it using a handwritten digit (digits 0-9) classification task, with all steps of the neural network computation performed on the focal plane. The vision chip that we deploy ( SCAMP-7 ) has a 256×256 array of processor elements (PE) integrated within the image sensor. The algorithm runs at over 3000 ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/9180959", "content": "We present a novel method of CNN inference on a pixel processor array device, demonstrating it using a handwritten digit (digits 0-9) classification task, with all steps of the neural network computation performed on the focal plane. The vision chip that we deploy ( SCAMP-7 ) has a 256×256 array of processor elements (PE) integrated within the image sensor. The algorithm runs at over 3000 ..."}
+{"idx": 9, "title": "Descriptor-In-Pixel : Point-Feature Tracking For Pixel Processor Arrays", "date": "", "ddg_snippet": "To achieve this we store feature descriptors inside every pixel -processor, adjusting the layout of these descriptors every frame. The PPA's architecture enables us to compute the response of every stored descriptor in parallel. This \"response map\" is utilized for both detection and tracking of point-features across the pixel -processor array.", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/poster/32867", "content": "To achieve this we store feature descriptors inside every pixel -processor, adjusting the layout of these descriptors every frame. The PPA's architecture enables us to compute the response of every stored descriptor in parallel. This \"response map\" is utilized for both detection and tracking of point-features across the pixel -processor array."}
diff --git a/data/sampled_jsons/DnCNN_GitHub_official_year_2017.jsonl b/data/sampled_jsons/DnCNN_GitHub_official_year_2017.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..56b6229cd4db19cc1f6a908496104c174f555257
--- /dev/null
+++ b/data/sampled_jsons/DnCNN_GitHub_official_year_2017.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "GitHub - cszn/DnCNN: Beyond a Gaussian Denoiser: Residual Learning of ...", "date": "", "ddg_snippet": "The parameters in DnCNN are mainly representing the image priors (task-independent), thus it is possible to learn a single model for different tasks, such as image denoising, image super-resolution and JPEG image deblocking. The left is the input image corrupted by different degradations, the right is the restored image by DnCNN-3.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/cszn/DnCNN", "content": "The parameters in DnCNN are mainly representing the image priors (task-independent), thus it is possible to learn a single model for different tasks, such as image denoising, image super-resolution and JPEG image deblocking. The left is the input image corrupted by different degradations, the right is the restored image by DnCNN-3."}
+{"idx": 1, "title": "DnCNN/README.md at master · cszn/DnCNN · GitHub", "date": "", "ddg_snippet": "The parameters in DnCNN are mainly representing the image priors (task-independent), thus it is possible to learn a single model for different tasks, such as image denoising, image super-resolution and JPEG image deblocking. The left is the input image corrupted by different degradations, the right is the restored image by DnCNN-3.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/cszn/DnCNN/blob/master/README.md", "content": "The parameters in DnCNN are mainly representing the image priors (task-independent), thus it is possible to learn a single model for different tasks, such as image denoising, image super-resolution and JPEG image deblocking. The left is the input image corrupted by different degradations, the right is the restored image by DnCNN-3."}
+{"idx": 2, "title": "GitHub - SaoYan/DnCNN-PyTorch: PyTorch implementation of the TIP2017 ...", "date": "", "ddg_snippet": "PyTorch implementation of the TIP2017 paper \"Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising\" - SaoYan/ DnCNN -PyTorch", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/SaoYan/DnCNN-PyTorch", "content": "PyTorch implementation of the TIP2017 paper \"Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising\" - SaoYan/ DnCNN -PyTorch"}
+{"idx": 3, "title": "GitHub - PuffannXu/DnCNN", "date": "", "ddg_snippet": "Contribute to PuffannXu/ DnCNN development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/PuffannXu/DnCNN", "content": "Contribute to PuffannXu/ DnCNN development by creating an account on GitHub ."}
+{"idx": 4, "title": "dncnn · GitHub Topics · GitHub", "date": "", "ddg_snippet": "GitHub is where people build software. More than 100 million people use GitHub to discover, fork, and contribute to over 330 million projects.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/topics/dncnn?o=desc&s=updated", "content": "GitHub is where people build software. More than 100 million people use GitHub to discover, fork, and contribute to over 330 million projects."}
+{"idx": 5, "title": "build minimal library · Tencent/ncnn Wiki · GitHub", "date": "", "ddg_snippet": "For some reason, if you're not happy with the binary size of the ncnn library, then here is the cheatsheet that helps you to build a minimal ncnn :P cmake -DNCNN_DISABLE_RTTI=ON -DNCNN_DISABLE_EXCEPTION=ON .. Cannot use RTTI and Exceptions when ncnn functions are called. cmake -DNCNN_PIXEL_DRAWING ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/Tencent/ncnn/wiki/build-minimal-library", "content": "For some reason, if you're not happy with the binary size of the ncnn library, then here is the cheatsheet that helps you to build a minimal ncnn :P cmake -DNCNN_DISABLE_RTTI=ON -DNCNN_DISABLE_EXCEPTION=ON .. Cannot use RTTI and Exceptions when ncnn functions are called. cmake -DNCNN_PIXEL_DRAWING ..."}
+{"idx": 6, "title": "DnCNN - Aditya Rastogi, PhD", "date": "", "ddg_snippet": "DnCNN This repository contains the codes for the implementation of the paper - Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising. This was done as part of a course project. Introduction During the past decade, convolutional neural networks have shown great success in handling various low-level vision tasks. Image denoising is one such long-standing problem in ...", "subpage_snippet": "", "source": "adityarastogi2k12.github.io", "link": "https://adityarastogi2k12.github.io/Projects/NNLS/", "content": "DnCNN This repository contains the codes for the implementation of the paper - Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising. This was done as part of a course project. Introduction During the past decade, convolutional neural networks have shown great success in handling various low-level vision tasks. Image denoising is one such long-standing problem in ..."}
+{"idx": 7, "title": "DnCNN - Image noise remover - atikul-islam-sajib.github.io", "date": "", "ddg_snippet": "DnCNN - Residual Learning of Deep CNN for Image Denoising The DnCNN (Denoising Convolutional Neural Network) is a deep learning model designed for image denoising tasks, offering state-of-the-art performance by learning end-to-end mappings from noisy to clean images. It employs a deep architecture consisting of several convolutional layers without any fully connected layers, enabling it to ...", "subpage_snippet": "", "source": "atikul-islam-sajib.github.io", "link": "https://atikul-islam-sajib.github.io/DnCNN-deploy/", "content": "DnCNN - Residual Learning of Deep CNN for Image Denoising The DnCNN (Denoising Convolutional Neural Network) is a deep learning model designed for image denoising tasks, offering state-of-the-art performance by learning end-to-end mappings from noisy to clean images. It employs a deep architecture consisting of several convolutional layers without any fully connected layers, enabling it to ..."}
+{"idx": 8, "title": "Charts - DnCNN - Image noise remover", "date": "", "ddg_snippet": "A class used to represent and perform operations for model visualization, including selecting the best model based on loss, loading data, normalizing images, creating GIFs from training images, and plotting comparison images (noisy, clean, predicted) using Matplotlib.", "subpage_snippet": "", "source": "atikul-islam-sajib.github.io", "link": "https://atikul-islam-sajib.github.io/DnCNN-deploy/charts/", "content": "A class used to represent and perform operations for model visualization, including selecting the best model based on loss, loading data, normalizing images, creating GIFs from training images, and plotting comparison images (noisy, clean, predicted) using Matplotlib."}
+{"idx": 9, "title": "Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image ...", "date": "", "ddg_snippet": "With the residual learning strategy, DnCNN implicitly removes the latent clean image in the hidden layers. This property motivates us to train a single DnCNN model to tackle with several general image denoising tasks such as Gaussian denoising, single image super-resolution and JPEG image deblocking.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1608.03981", "content": "With the residual learning strategy, DnCNN implicitly removes the latent clean image in the hidden layers. This property motivates us to train a single DnCNN model to tackle with several general image denoising tasks such as Gaussian denoising, single image super-resolution and JPEG image deblocking."}
diff --git a/data/sampled_jsons/Do_Not_Trust_What_They_Tell_Exposing_Malicious_Accomplices_in_Tor_Section_4.3_Equation_8_conn(a,b)_a.jsonl b/data/sampled_jsons/Do_Not_Trust_What_They_Tell_Exposing_Malicious_Accomplices_in_Tor_Section_4.3_Equation_8_conn(a,b)_a.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d6e38ddcb02d7e17df588f65aaf535b7589a2c3a
--- /dev/null
+++ b/data/sampled_jsons/Do_Not_Trust_What_They_Tell_Exposing_Malicious_Accomplices_in_Tor_Section_4.3_Equation_8_conn(a,b)_a.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via ...", "date": "", "ddg_snippet": "This paper presents a novel approach for detecting anomalous circuits in the Tor network, and for the first time provides a more comprehensive identification of potential malicious accomplice nodes in Tor by taking roles of nodes in anomalous circuits into consideration.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1145/3696410.3714767", "content": "This paper presents a novel approach for detecting anomalous circuits in the Tor network, and for the first time provides a more comprehensive identification of potential malicious accomplice nodes in Tor by taking roles of nodes in anomalous circuits into consideration."}
+{"idx": 1, "title": "PDF Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via ...", "date": "", "ddg_snippet": "Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection. In Proceedings of the ACM Web Conference 2025 (WWW '25), April 28-May 2, 2025, Sydney, NSW, Australia.", "subpage_snippet": "", "source": "cse.seu.edu.cn", "link": "https://cse.seu.edu.cn/_upload/article/files/82/36/84188a1a47a79fe2b6dc589a1c39/a6db7ad8-3fef-4518-8046-480a6ac64280.pdf", "content": "Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection. In Proceedings of the ACM Web Conference 2025 (WWW '25), April 28-May 2, 2025, Sydney, NSW, Australia."}
+{"idx": 2, "title": "通过异常电路检测揭露Tor中的恶意同伙 - 安全内参 | 决策者的网络安全知识库", "date": "", "ddg_snippet": "提出了一种方法来检测Tor网络中的异常电路，通过考虑节点在异常电路中的角色，首次提供了一个更全面的方法识别tor中的 ...", "subpage_snippet": "", "source": "www.secrss.com", "link": "https://www.secrss.com/articles/76608", "content": "提出了一种方法来检测Tor网络中的异常电路，通过考虑节点在异常电路中的角色，首次提供了一个更全面的方法识别tor中的 ..."}
+{"idx": 3, "title": "PDF Defending Against Malicious Cyber Activity Originating from Tor - CISA", "date": "", "ddg_snippet": "Defending Against Malicious Cyber Activity Originating from Tor This advisory—written by the Cybersecurity Security and Infrastructure Security Agency (CISA) with contributions from the Federal Bureau of Investigation (FBI)—highlights risks associated with Tor , along with technical details and recommendations for mitigation. Cyber threat actors can use Tor software and network ...", "subpage_snippet": "", "source": "www.cisa.gov", "link": "https://www.cisa.gov/sites/default/files/publications/AA20-183A_Defending_Against_Malicious_Cyber_Activity_Originating_from_Tor_S508C.pdf", "content": "Defending Against Malicious Cyber Activity Originating from Tor This advisory—written by the Cybersecurity Security and Infrastructure Security Agency (CISA) with contributions from the Federal Bureau of Investigation (FBI)—highlights risks associated with Tor , along with technical details and recommendations for mitigation. Cyber threat actors can use Tor software and network ..."}
+{"idx": 4, "title": "通过异常电路检测揭露tor中的恶意同伙", "date": "", "ddg_snippet": "原文标题： Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection 原文作者：Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan-Gu, Chunmian Wang", "subpage_snippet": "", "source": "sechub.in", "link": "https://sechub.in/view/3027398", "content": "原文标题： Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection 原文作者：Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan-Gu, Chunmian Wang"}
+{"idx": 5, "title": "Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via ...", "date": "", "ddg_snippet": "This feature describes the rela-tionship of two nodes in term of their abnormal behavioral patterns, which can be identified through the method in Section 4.2. We intro-(, ) duce a connectivity metric in Equation ( 8 ) to quantify this relationship between nodes and .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qcnePVejeV", "content": "This feature describes the rela-tionship of two nodes in term of their abnormal behavioral patterns, which can be identified through the method in Section 4.2. We intro-(, ) duce a connectivity metric in Equation ( 8 ) to quantify this relationship between nodes and ."}
+{"idx": 6, "title": "Spoiled Onions: Exposing Malicious Tor Exit Relays", "date": "", "ddg_snippet": "Tor exit relays are operated by volunteers and together push more than 1 GiB/s of network traffic. By design, these volunteers are able to inspect and modify the anonymized network traffic. In this paper, we seek to expose such malicious exit relays and document...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-319-08506-7_16", "content": "Tor exit relays are operated by volunteers and together push more than 1 GiB/s of network traffic. By design, these volunteers are able to inspect and modify the anonymized network traffic. In this paper, we seek to expose such malicious exit relays and document..."}
+{"idx": 7, "title": "Submissions | OpenReview", "date": "", "ddg_snippet": "Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan-Gu, Chunmian Wang", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/submissions?venue=ACM.org/TheWebConf/2025/Conference", "content": "Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan-Gu, Chunmian Wang"}
+{"idx": 8, "title": "PDF DEF CON 24 Hacking Conference", "date": "", "ddg_snippet": "Abstract—Tor is a widely used anonymity network that protects users' privacy and identity from corporations, agencies and governments. However, Tor remains a practical system with a variety of limitations which can be subverted [1]. In particular, Tor's security relies on the fact that a substantial number of its nodes do not misbehave.", "subpage_snippet": "", "source": "media.defcon.org", "link": "https://media.defcon.org/DEF+CON+24/DEF+CON+24+presentations/DEF+CON+24+-+Guevara-Noubir-Amirali-Sanatinia-Honey-Onions-Exposing-Snooping-Tor-Hsdir-Relays-WP.pdf", "content": "Abstract—Tor is a widely used anonymity network that protects users' privacy and identity from corporations, agencies and governments. However, Tor remains a practical system with a variety of limitations which can be subverted [1]. In particular, Tor's security relies on the fact that a substantial number of its nodes do not misbehave."}
+{"idx": 9, "title": "Xiaodan Gu - dblp", "date": "", "ddg_snippet": "Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan Gu, Chunmian Wang: Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection. WWW 2025: 2959-2968", "subpage_snippet": "", "source": "dblp.org", "link": "https://dblp.org/pid/148/7260", "content": "Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong, Xiaodan Gu, Chunmian Wang: Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection. WWW 2025: 2959-2968"}
diff --git a/data/sampled_jsons/ETHICS_dataset_reasons_context-aware_ethical_alignment_empirical_evaluation.jsonl b/data/sampled_jsons/ETHICS_dataset_reasons_context-aware_ethical_alignment_empirical_evaluation.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bd9cc2ba573f58b92043997205fda46ae7bfbf56
--- /dev/null
+++ b/data/sampled_jsons/ETHICS_dataset_reasons_context-aware_ethical_alignment_empirical_evaluation.jsonl
@@ -0,0 +1,6 @@
+{"idx": 0, "title": "Daily Papers - Hugging Face", "date": "", "ddg_snippet": "Ethical Reasoning over Moral Alignment : A Case and Framework for In- Context Ethical Policies in LLMs.With the ETHICS dataset , we find that current language models have a promising but incomplete ability to predict basic human ethical judgements.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers?q=ethical+verification+framework", "content": "Ethical Reasoning over Moral Alignment : A Case and Framework for In- Context Ethical Policies in LLMs.With the ETHICS dataset , we find that current language models have a promising but incomplete ability to predict basic human ethical judgements."}
+{"idx": 1, "title": "(PDF) On the Trustworthiness of Generative Foundation Models...", "date": "", "ddg_snippet": "In constructing the dynamic dataset for testing LLM ethics , the following ethical considerations. and procedures are observed: (a) Initially, the metadata curator utilizes a dataset pool derived from several key.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/389176427_On_the_Trustworthiness_of_Generative_Foundation_Models_Guideline_Assessment_and_Perspective", "content": "In constructing the dynamic dataset for testing LLM ethics , the following ethical considerations. and procedures are observed: (a) Initially, the metadata curator utilizes a dataset pool derived from several key."}
+{"idx": 2, "title": "Daily Papers - Hugging Face", "date": "", "ddg_snippet": "To address the need for ethical judgment in CIR, we present the QA- ETHICS dataset , adapted from the ETHICS benchmark, which serves as an evaluation tool by unifying scenarios and label meanings.", "subpage_snippet": "", "source": "hf.qhduan.com", "link": "https://hf.qhduan.com/papers?q=code+retrieval", "content": "To address the need for ethical judgment in CIR, we present the QA- ETHICS dataset , adapted from the ETHICS benchmark, which serves as an evaluation tool by unifying scenarios and label meanings."}
+{"idx": 3, "title": "Daily Papers", "date": "", "ddg_snippet": "6 days ago — Our empirical evaluation demonstrates that our approach leads to a ... To this end, we introduce a workflow that integrates ethical alignment , ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers?q=objection+detection", "content": "6 days ago — Our empirical evaluation demonstrates that our approach leads to a ... To this end, we introduce a workflow that integrates ethical alignment , ..."}
+{"idx": 4, "title": "Daily Papers", "date": "", "ddg_snippet": "A comprehensive empirical evaluation exhibits ... ETHICS dataset , adapted from the ETHICS benchmark ... Our research provides a practical method for introducing ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers?q=Conversational+QA", "content": "A comprehensive empirical evaluation exhibits ... ETHICS dataset , adapted from the ETHICS benchmark ... Our research provides a practical method for introducing ..."}
+{"idx": 5, "title": "On the Trustworthiness of Generative Foundation Models", "date": "", "ddg_snippet": "7 Nov 2024 — ethical alignment ... detailed logging of decision pathways [1788], and context-aware explanations [1789] can illuminate where and why . 231 pages", "subpage_snippet": "", "source": "howiehwong.github.io", "link": "https://howiehwong.github.io/TrustGen.pdf", "content": "7 Nov 2024 — ethical alignment ... detailed logging of decision pathways [1788], and context-aware explanations [1789] can illuminate where and why . 231 pages"}
diff --git a/data/sampled_jsons/EntityErasure_arxiv_IoS_intersection_over_segmentation_equation_threshold.jsonl b/data/sampled_jsons/EntityErasure_arxiv_IoS_intersection_over_segmentation_equation_threshold.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..176b57c1f3bd0a9756e16d55d17b925629a20c00
--- /dev/null
+++ b/data/sampled_jsons/EntityErasure_arxiv_IoS_intersection_over_segmentation_equation_threshold.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Jaccard index - Wikipedia", "date": "", "ddg_snippet": "The Jaccard index is a statistic used for gauging the similarity and diversity of sample sets. It is defined in general taking the ratio of two sizes, the intersection size divided by the union size, also called intersection over union.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Jaccard_index", "content": "The Jaccard index is a statistic used for gauging the similarity and diversity of sample sets. It is defined in general taking the ratio of two sizes, the intersection size divided by the union size, also called intersection over union."}
+{"idx": 1, "title": "EntityErasure: Erasing Entity Cleanly via Amodal Entity ...", "date": "", "ddg_snippet": "As shown Fig. 4, we use the pre-trained state-of-the-art open-world segmentation model EntitySeg [30] to predict the en-tity segmentation of the generated result, and then calculate the Intersection over Self ( IoS ) between each entity with in-painting mask by:", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Zhu_EntityErasure_Erasing_Entity_Cleanly_via_Amodal_Entity_Segmentation_and_Completion_CVPR_2025_paper.pdf", "content": "As shown Fig. 4, we use the pre-trained state-of-the-art open-world segmentation model EntitySeg [30] to predict the en-tity segmentation of the generated result, and then calculate the Intersection over Self ( IoS ) between each entity with in-painting mask by:"}
+{"idx": 2, "title": "EntityErasure: Erasing Entity Cleanly via Amodal Entity ...", "date": "", "ddg_snippet": "This paper presents EntityErasure , a novel diffusion-based inpainting method that can effectively erase entities without inducing unwanted sundries. To this end, we propose to address this problem by dividing it into amodal entity segmentation and completion, such that the region to inpaint takes only entities in the non-inpainting area as reference, avoiding the possibility to generate ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/11094156", "content": "This paper presents EntityErasure , a novel diffusion-based inpainting method that can effectively erase entities without inducing unwanted sundries. To this end, we propose to address this problem by dividing it into amodal entity segmentation and completion, such that the region to inpaint takes only entities in the non-inpainting area as reference, avoiding the possibility to generate ..."}
+{"idx": 3, "title": "arXiv:2103.16562v1 [cs.CV] 30 Mar 2021", "date": "", "ddg_snippet": "1. Introduction The Common Task Framework [27], in which standard-ized tasks, datasets, and evaluation metrics are used to track research progress, yields impressive results. For exam-ple, researchers working on the instance segmentation task, which requires an algorithm to delineate objects with pixel-level binary masks, have improved the standard Average Precision (AP) metric on COCO [28] by ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2103.16562", "content": "1. Introduction The Common Task Framework [27], in which standard-ized tasks, datasets, and evaluation metrics are used to track research progress, yields impressive results. For exam-ple, researchers working on the instance segmentation task, which requires an algorithm to delineate objects with pixel-level binary masks, have improved the standard Average Precision (AP) metric on COCO [28] by ..."}
+{"idx": 4, "title": "Optimizing Intersection-Over-Union in Deep Neural Networks ...", "date": "", "ddg_snippet": "Dec 10, 2016 · The intersection - over -union (IoU) is usually used to measure the performance of any object category segmentation method. In this paper, we propose an approach for directly optimizing this IoU measure in deep neural networks.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-319-50835-1_22", "content": "Dec 10, 2016 · The intersection - over -union (IoU) is usually used to measure the performance of any object category segmentation method. In this paper, we propose an approach for directly optimizing this IoU measure in deep neural networks."}
+{"idx": 5, "title": "Intersection over Union (IoU): Definition, Calculation, Code", "date": "", "ddg_snippet": "Intersection over Union is a popular metric to measure localization accuracy and compute localization errors in object detection models. It calculates the amount of overlapping between two bounding boxes—a predicted bounding box and a ground truth bounding box.", "subpage_snippet": "", "source": "www.v7labs.com", "link": "https://www.v7labs.com/blog/intersection-over-union-guide", "content": "Intersection over Union is a popular metric to measure localization accuracy and compute localization errors in object detection models. It calculates the amount of overlapping between two bounding boxes—a predicted bounding box and a ground truth bounding box."}
+{"idx": 6, "title": "Intersection Over Union IoU in Object Detection Segmentation", "date": "", "ddg_snippet": "Intersection Over Union (IoU) is a helper metric for evaluating object detection and segmentation model. Learn from the basics to implementing IoU from scratch.", "subpage_snippet": "", "source": "learnopencv.com", "link": "https://learnopencv.com/intersection-over-union-iou-in-object-detection-and-segmentation/", "content": "Intersection Over Union (IoU) is a helper metric for evaluating object detection and segmentation model. Learn from the basics to implementing IoU from scratch."}
+{"idx": 7, "title": "Understanding Intersection over Union for Model Accuracy", "date": "", "ddg_snippet": "The intersection - over -union (IoU) threshold acts as a gatekeeper, classifying predicted bounding boxes as true positives if they pass the threshold and false positives if they fall below it. By adjusting the threshold , we can control the trade-off between precision and recall.", "subpage_snippet": "", "source": "viso.ai", "link": "https://viso.ai/computer-vision/intersection-over-union-iou/", "content": "The intersection - over -union (IoU) threshold acts as a gatekeeper, classifying predicted bounding boxes as true positives if they pass the threshold and false positives if they fall below it. By adjusting the threshold , we can control the trade-off between precision and recall."}
+{"idx": 8, "title": "arXiv .org e-Print archive", "date": "", "ddg_snippet": "arXiv is working with academic researchers to investigate ways to improve the service for our users. This requires a rich dataset to better inform the research. Users who wish to contribute to the future of arXiv may opt in to allow arXiv to use their reading data for research purposes.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/", "content": "arXiv is working with academic researchers to investigate ways to improve the service for our users. This requires a rich dataset to better inform the research. Users who wish to contribute to the future of arXiv may opt in to allow arXiv to use their reading data for research purposes."}
+{"idx": 9, "title": "Intersection over Union (IoU) for object detection | SuperAnnotate", "date": "", "ddg_snippet": "Intersection over Union (IoU) is a measure that shows how well the prediction bounding box aligns with the ground truth box. It's one of the main metrics for evaluating the accuracy of object detection algorithms and helps distinguish between \"correct detection\" and \"incorrect detection\".", "subpage_snippet": "", "source": "www.superannotate.com", "link": "https://www.superannotate.com/blog/intersection-over-union-for-object-detection", "content": "Intersection over Union (IoU) is a measure that shows how well the prediction bounding box aligns with the ground truth box. It's one of the main metrics for evaluating the accuracy of object detection algorithms and helps distinguish between \"correct detection\" and \"incorrect detection\"."}
diff --git a/data/sampled_jsons/EventPS_Real-time_photometric_stereo_using_an_event_camera_CVPR_2022_Yu_year_2022.jsonl b/data/sampled_jsons/EventPS_Real-time_photometric_stereo_using_an_event_camera_CVPR_2022_Yu_year_2022.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3c677a4875fd3d525194321c1a406ede804c27ec
--- /dev/null
+++ b/data/sampled_jsons/EventPS_Real-time_photometric_stereo_using_an_event_camera_CVPR_2022_Yu_year_2022.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "PDF EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , EventPS estimates surface nor-mal only from the radiance changes, significantly enhanc-ing data efficiency.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2024/papers/Yu_EventPS_Real-Time_Photometric_Stereo_Using_an_Event_Camera_CVPR_2024_paper.pdf", "content": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , EventPS estimates surface nor-mal only from the radiance changes, significantly enhanc-ing data efficiency."}
+{"idx": 1, "title": "EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "Photometric stereo is a well-established technique to es-timate the surface normal of an object. However, the re-quirement of capturing multiple high dynamic range images under different illumination conditions limits the speed and real-time applications. This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/10655016", "content": "Photometric stereo is a well-established technique to es-timate the surface normal of an object. However, the re-quirement of capturing multiple high dynamic range images under different illumination conditions limits the speed and real-time applications. This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional ..."}
+{"idx": 2, "title": "EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera that capitalizes on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , significantly enhancing data efficiency. Photometric stereo is a well-established technique to es-timate the surface normal of an object. However, the re-quirement of ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/EventPS:-Real-Time-Photometric-Stereo-Using-an-Yu-Ren/7f72975f58ceff79a3762464ba7e5f8c29c54aaf", "content": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera that capitalizes on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , significantly enhancing data efficiency. Photometric stereo is a well-established technique to es-timate the surface normal of an object. However, the re-quirement of ..."}
+{"idx": 3, "title": "EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution dynamic range and low bandwidth characteristics of event cameras EventPS estimates surface normal only from the radiance changes significantly enhancing data efficiency.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2024/html/Yu_EventPS_Real-Time_Photometric_Stereo_Using_an_Event_Camera_CVPR_2024_paper.html", "content": "This paper introduces EventPS a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution dynamic range and low bandwidth characteristics of event cameras EventPS estimates surface normal only from the radiance changes significantly enhancing data efficiency."}
+{"idx": 4, "title": "CVPR Poster EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth characteristics of event cameras , EventPS estimates surface normal only from the radiance changes, significantly enhancing data efficiency.", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2024/poster/31806", "content": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth characteristics of event cameras , EventPS estimates surface normal only from the radiance changes, significantly enhancing data efficiency."}
+{"idx": 5, "title": "CVPR Poster PS-EIP: Robust Photometric Stereo Based on Event Interval ...", "date": "", "ddg_snippet": "Abstract: Recently, the energy-efficient photometric stereo method using an event camera ( EventPS ) has been proposed to recover surface normals from events triggered by changes in logarithmic Lambertian reflections under a moving directional light source. However, EventPS treats each event interval independently, making it sensitive to noise, shadows, and non-Lambertian reflections. This paper ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/poster/33500", "content": "Abstract: Recently, the energy-efficient photometric stereo method using an event camera ( EventPS ) has been proposed to recover surface normals from events triggered by changes in logarithmic Lambertian reflections under a moving directional light source. However, EventPS treats each event interval independently, making it sensitive to noise, shadows, and non-Lambertian reflections. This paper ..."}
+{"idx": 6, "title": "PDF EventPSR: Surface Normal and Reflectance Estimation from Photometric ...", "date": "", "ddg_snippet": "Re-searchers have explored applying event cameras in various photometric tasks [6, 19, 29, 33]. Particularly, EventPS [29] has demonstrated the superiority of event cameras for real-time surface normal estimation, showing the potential to accelerate and improve accuracy for shape and reflectance estimation tasks.", "subpage_snippet": "", "source": "assets.ctfassets.net", "link": "https://assets.ctfassets.net/yreyglvi5sud/5jrOdfPf8UFsyTqMjOdgIe/41df47af774e585c44a96a247be7ecfd/Yu_CVPR25a.pdf", "content": "Re-searchers have explored applying event cameras in various photometric tasks [6, 19, 29, 33]. Particularly, EventPS [29] has demonstrated the superiority of event cameras for real-time surface normal estimation, showing the potential to accelerate and improve accuracy for shape and reflectance estimation tasks."}
+{"idx": 7, "title": "EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth characteristics of event cameras , EventPS estimates surface normal only from the radiance changes, significantly enhancing data efficiency.", "subpage_snippet": "", "source": "www.ybh1998.space", "link": "https://www.ybh1998.space/eventps-real-time-photometric-stereo-using-an-event-camera/", "content": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth characteristics of event cameras , EventPS estimates surface normal only from the radiance changes, significantly enhancing data efficiency."}
+{"idx": 8, "title": "EventPS: Real-Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , EventPS estimates surface nor-mal only from the radiance changes, significantly enhancing data efficiency.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=YH3f8jq3lz", "content": "This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional temporal resolution, dynamic range, and low bandwidth character-istics of event cameras , EventPS estimates surface nor-mal only from the radiance changes, significantly enhancing data efficiency."}
+{"idx": 9, "title": "Track: Orals 3C Medical and Physics-based vision", "date": "", "ddg_snippet": "Photometric stereo is a well-established technique to estimate the surface normal of an object. However, the requirement of capturing multiple high dynamic range images under different illumination conditions limits the speed and real-time applications. This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2024/session/32097", "content": "Photometric stereo is a well-established technique to estimate the surface normal of an object. However, the requirement of capturing multiple high dynamic range images under different illumination conditions limits the speed and real-time applications. This paper introduces EventPS , a novel approach to real-time photometric stereo using an event camera . Capitalizing on the exceptional ..."}
diff --git a/data/sampled_jsons/FD3_Sieve_MLE_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000.jsonl b/data/sampled_jsons/FD3_Sieve_MLE_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3291217c5e8685404ada812581422b62f049bd8f
--- /dev/null
+++ b/data/sampled_jsons/FD3_Sieve_MLE_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000_0.0000.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "NIST reactor : summary activities July 1990 through June ...", "date": "", "ddg_snippet": "... Fd3 (No. 203). Our BePO-X and ZnPO-X models assumed a fully ordered 1 : distribution of Be/Zn and P over the framework tetrahedra, and strict alternation of ...", "subpage_snippet": "", "source": "www.govinfo.gov", "link": "https://www.govinfo.gov/content/pkg/GOVPUB-C13-a64a50f25606b8ae2db871a896264bdc/pdf/GOVPUB-C13-a64a50f25606b8ae2db871a896264bdc.pdf", "content": "... Fd3 (No. 203). Our BePO-X and ZnPO-X models assumed a fully ordered 1 : distribution of Be/Zn and P over the framework tetrahedra, and strict alternation of ..."}
+{"idx": 1, "title": "1310 CLEGG STREET (2) - IIS Windows Server", "date": "", "ddg_snippet": "25 Nov 2020 — ... FD3 WILLIAMS 4' 0\"L X 2' D\"W X 4 1/2\"D , FLUORESCENTTROFFER ... Nodal Displacements Node Load Case DX ft DY ft RZ deg N1 DL 0.0000 0.0000 ...", "subpage_snippet": "", "source": "petalumadocs.cityofpetaluma.net", "link": "https://petalumadocs.cityofpetaluma.net/WebLink/DocView.aspx?id=373223&dbid=0&repo=Petaluma", "content": "25 Nov 2020 — ... FD3 WILLIAMS 4' 0\"L X 2' D\"W X 4 1/2\"D , FLUORESCENTTROFFER ... Nodal Displacements Node Load Case DX ft DY ft RZ deg N1 DL 0.0000 0.0000 ..."}
+{"idx": 2, "title": "NIST reactor : summary activities July 1990 through June 1991", "date": "", "ddg_snippet": "... Fd3 (No. 203). Our BePO-X and ZnPO-X models assumed a fully ordered 1 : distribution of Be/Zn and P over the framework tetrahedra, and strict alternation of ...", "subpage_snippet": "", "source": "nvlpubs.nist.gov", "link": "https://nvlpubs.nist.gov/nistpubs/Legacy/TN/nbstechnicalnote1292.pdf", "content": "... Fd3 (No. 203). Our BePO-X and ZnPO-X models assumed a fully ordered 1 : distribution of Be/Zn and P over the framework tetrahedra, and strict alternation of ..."}
+{"idx": 3, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}
+{"idx": 4, "title": "Wheat streak mosaic virus alters the transcriptome of its vector ...", "date": "", "ddg_snippet": "by AK Gupta · Cited by 18 — Abstract. Wheat streak mosaic virus (WSMV; genus Tritimovirus; family Potyviridae) is an economically important wheat virus that is.", "subpage_snippet": "", "source": "digitalcommons.unl.edu", "link": "https://digitalcommons.unl.edu/context/plantpathpapers/article/1630/viewcontent/Gupta_JGV_2019_Wheat_streak_mosaic.pdf", "content": "by AK Gupta · Cited by 18 — Abstract. Wheat streak mosaic virus (WSMV; genus Tritimovirus; family Potyviridae) is an economically important wheat virus that is."}
+{"idx": 5, "title": "lake champlain: partnersrips and research in the new ...", "date": "", "ddg_snippet": "Lake Champlain: partnerships and research in the new millenniurnledited by Thomas 0. Manley, PaUicia L. Manley, and Timothy B. Mihuc.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/978-1-4757-4080-6.pdf", "content": "Lake Champlain: partnerships and research in the new millenniurnledited by Thomas 0. Manley, PaUicia L. Manley, and Timothy B. Mihuc."}
+{"idx": 6, "title": "EHY223 HYSYS Dynamics Introduction To ...", "date": "", "ddg_snippet": "At the end of this course you will be able to: ll Develop the skills and techniques required for creating and running dynamic simulations.", "subpage_snippet": "", "source": "www.scribd.com", "link": "https://www.scribd.com/document/607789019/EHY223-HYSYS-Dynamics-Introduction-to-Dynamic-Modeling", "content": "At the end of this course you will be able to: ll Develop the skills and techniques required for creating and running dynamic simulations."}
+{"idx": 7, "title": "Final Report Volume I of II Testing of a 4-Stroke Diesel Cycle ...", "date": "", "ddg_snippet": "Testing of a 4-stroke diesel cycle oil-fired reciprocating internal combustion engine to determine the effectiveness of an oxidation reduction catalyst system.", "subpage_snippet": "", "source": "nepis.epa.gov", "link": "https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100VGMK.TXT", "content": "Testing of a 4-stroke diesel cycle oil-fired reciprocating internal combustion engine to determine the effectiveness of an oxidation reduction catalyst system."}
+{"idx": 8, "title": "Chitosan and Chitosan Derivatives as Chelating Agents", "date": "", "ddg_snippet": "This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission and sources are indicated. A ...", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/23222791/Chitosan_and_Chitosan_Derivatives_as_Chelating_Agents", "content": "This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission and sources are indicated. A ..."}
+{"idx": 9, "title": "(PDF) Compilation of Full Papers Presented at ICMSM2013", "date": "", "ddg_snippet": "The compilation of full papers presented at the International Congress of the Malaysian Society for Microbiology (ICMSM2013) highlights the diverse research ...", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/8810012/Compilation_of_Full_Papers_Presented_at_ICMSM2013", "content": "The compilation of full papers presented at the International Congress of the Malaysian Society for Microbiology (ICMSM2013) highlights the diverse research ..."}
diff --git a/data/sampled_jsons/Figure_3_Contrastive_CRL_MCC_0.95_0.15_synthetic_ablation_real_data.jsonl b/data/sampled_jsons/Figure_3_Contrastive_CRL_MCC_0.95_0.15_synthetic_ablation_real_data.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..19938c66d8d3dd39144cc6089ec98636dc860f45
--- /dev/null
+++ b/data/sampled_jsons/Figure_3_Contrastive_CRL_MCC_0.95_0.15_synthetic_ablation_real_data.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Sanity Checking Causal Representation Learning on a Simple...", "date": "", "ddg_snippet": "Figure 4: Ground-truth graph relating the underlying causal factors, shown in bold print, to the different views employed in the Multiview CRL experiment. The views are disjoint sets of the output variables produced by the tunnel.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2502.20099", "content": "Figure 4: Ground-truth graph relating the underlying causal factors, shown in bold print, to the different views employed in the Multiview CRL experiment. The views are disjoint sets of the output variables produced by the tunnel."}
+{"idx": 1, "title": "identification of nonparametric dynamic causal structure ...", "date": "", "ddg_snippet": "by M Fu · 2025 — As shown in Figure 1, climate time-series data satisfied the. 3 -measurement model, which supports the enough information for identification of ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2501.12500", "content": "by M Fu · 2025 — As shown in Figure 1, climate time-series data satisfied the. 3 -measurement model, which supports the enough information for identification of ..."}
+{"idx": 2, "title": "Autophagy-Dependent Generation of Free Fatty Acids Is ...", "date": "", "ddg_snippet": "by T Riffelmacher · 2017 · Cited by 316 — We observed a profound ablation of Atg7 expression (>80%) by qPCR in mature neutrophils, their myeloblast precursors, and at the level of ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC5610174/", "content": "by T Riffelmacher · 2017 · Cited by 316 — We observed a profound ablation of Atg7 expression (>80%) by qPCR in mature neutrophils, their myeloblast precursors, and at the level of ..."}
+{"idx": 3, "title": "Impurities and defects in, and isotope compositions of, ...", "date": "", "ddg_snippet": "The diamonds were analyzed for their carbon isotopic compositions and nitrogen aggregation and concentration characteristics, and based on this dataset ...", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/pdf/impurities-and-defects-in-and-isotope-compositions-of-1kt50j4rru.pdf", "content": "The diamonds were analyzed for their carbon isotopic compositions and nitrogen aggregation and concentration characteristics, and based on this dataset ..."}
+{"idx": 4, "title": "Advanced Materials for Exploration Task Research Results", "date": "", "ddg_snippet": "by MB Cook · 2008 · Cited by 1 — Includes compilations of significant scientific and technical data and information deemed to be of continuing reference value. NASA's counterpart of peer-. 332 pages", "subpage_snippet": "", "source": "ntrs.nasa.gov", "link": "https://ntrs.nasa.gov/api/citations/20110008064/downloads/20110008064.pdf", "content": "by MB Cook · 2008 · Cited by 1 — Includes compilations of significant scientific and technical data and information deemed to be of continuing reference value. NASA's counterpart of peer-. 332 pages"}
+{"idx": 5, "title": "Sunday, April 27, 2025", "date": "", "ddg_snippet": "27 Apr 2025 — This study investigates the real -world rates of genetics referrals in eligible cancer patients at Singapore's largest healthcare cluster using ... 6,076 pages", "subpage_snippet": "", "source": "www.aacr.org", "link": "https://www.aacr.org/wp-content/uploads/2025/05/AACR2025_Proceedings_050725.pdf", "content": "27 Apr 2025 — This study investigates the real -world rates of genetics referrals in eligible cancer patients at Singapore's largest healthcare cluster using ... 6,076 pages"}
+{"idx": 6, "title": "ADVANCES IN MUCOSAL IMMUNOLOGY", "date": "", "ddg_snippet": "Figure 3 . Secondary IgM and IgG responses rccalled by dIfferent doses 01' SRBC 80 days after priming. PFC. 200. 150. %. 100. 50. --. 0--0 60 ...__'SO. ~j160.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/978-1-4615-1941-6.pdf", "content": "Figure 3 . Secondary IgM and IgG responses rccalled by dIfferent doses 01' SRBC 80 days after priming. PFC. 200. 150. %. 100. 50. --. 0--0 60 ...__'SO. ~j160."}
+{"idx": 7, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}
+{"idx": 8, "title": "Material", "date": "", "ddg_snippet": "The injured group showed a decrease of fibril moduli with an increase of GAG release from the cartilage ECM (see Fig . 3 ). Inversely, the ECM hydraulic ...", "subpage_snippet": "", "source": "www.biomomentum.com", "link": "https://www.biomomentum.com/material/", "content": "The injured group showed a decrease of fibril moduli with an increase of GAG release from the cartilage ECM (see Fig . 3 ). Inversely, the ECM hydraulic ..."}
+{"idx": 9, "title": "Top 10 Basic Science Abstract Presentations", "date": "", "ddg_snippet": "13 Mar 2025 — We have demonstrated that optogenetic stimulation of the urothelial cells can produce urothelial cell-mediated local bladder contractions.", "subpage_snippet": "", "source": "onlinelibrary.wiley.com", "link": "https://onlinelibrary.wiley.com/doi/full/10.1002/nau.25673", "content": "13 Mar 2025 — We have demonstrated that optogenetic stimulation of the urothelial cells can produce urothelial cell-mediated local bladder contractions."}
diff --git a/data/sampled_jsons/FlowDec_paper_Table_8_SIGMOS_values_4.5_kbps_FlowDec-75m_3.83_DAC-75.jsonl b/data/sampled_jsons/FlowDec_paper_Table_8_SIGMOS_values_4.5_kbps_FlowDec-75m_3.83_DAC-75.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..79895f3f6f1bc6f413929e21dcc78defd60e7b46
--- /dev/null
+++ b/data/sampled_jsons/FlowDec_paper_Table_8_SIGMOS_values_4.5_kbps_FlowDec-75m_3.83_DAC-75.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "FlowDec : A flow -based full-band general audio codec with high...", "date": "", "ddg_snippet": "FlowDec - 75 s: 75 Hz, single-bitrate.Figure 16: Spectrograms comparing FlowDec - 75 m against DAC - 75 as well as the initial decoder output (NDAC-75) on the three audio files where FlowDec - 75 m produces the worst fwSSNR values on the whole test set. A.9 Full objective metrics table .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.01485v1", "content": "FlowDec - 75 s: 75 Hz, single-bitrate.Figure 16: Spectrograms comparing FlowDec - 75 m against DAC - 75 as well as the initial decoder output (NDAC-75) on the three audio files where FlowDec - 75 m produces the worst fwSSNR values on the whole test set. A.9 Full objective metrics table ."}
+{"idx": 1, "title": "GitHub - hammerlab/ flowdec : TensorFlow Deconvolution for...", "date": "", "ddg_snippet": "Flowdec /TensorFlow with full GPU acceleration: ~1 second. Signal Dimensions - Flowdec can support 1, 2, or 3 dimensional images/signals.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/hammerlab/flowdec", "content": "Flowdec /TensorFlow with full GPU acceleration: ~1 second. Signal Dimensions - Flowdec can support 1, 2, or 3 dimensional images/signals."}
+{"idx": 2, "title": "FlowDec | A flow -based full-band general audio codec with high...", "date": "", "ddg_snippet": "For FlowDec , DAC, and DAC - 75 , we show both examples at a higher bitrate (7.0 - 7.75 kbps ) as well as at a lower bitrate (4.3 - 4 . 5 kbps ). You can choose between Speech, Music and Audio Files.", "subpage_snippet": "", "source": "sp-uhh.github.io", "link": "https://sp-uhh.github.io/FlowDec/", "content": "For FlowDec , DAC, and DAC - 75 , we show both examples at a higher bitrate (7.0 - 7.75 kbps ) as well as at a lower bitrate (4.3 - 4 . 5 kbps ). You can choose between Speech, Music and Audio Files."}
+{"idx": 3, "title": "Математика 4 класс Моро. Часть 1. Страница 18 Номер 2", "date": "", "ddg_snippet": "Часть 1. Страница 18 Номер 2 900 − 2 * 50 + 140; 600 + 90 : 3 − 200; 700 − 25 * 2 + 100; 120 − 75 : 3 * 4 + 65; 200 − 80 : 4 * 5 − 35; 108 − 54 : 9 * 6 + 58; 342 : 3; 564 : 2; 721 : 7.", "subpage_snippet": "", "source": "Reshalka.com", "link": "https://Reshalka.com/uchebniki/4-klass/matematika/moro/95", "content": "Часть 1. Страница 18 Номер 2 900 − 2 * 50 + 140; 600 + 90 : 3 − 200; 700 − 25 * 2 + 100; 120 − 75 : 3 * 4 + 65; 200 − 80 : 4 * 5 − 35; 108 − 54 : 9 * 6 + 58; 342 : 3; 564 : 2; 721 : 7."}
+{"idx": 4, "title": "Дмитрий Никотин – Telegram", "date": "", "ddg_snippet": "Про политику, простым языком о сложном. Контрпропаганда и интерпретация новостей.", "subpage_snippet": "", "source": "t.me", "link": "https://t.me/s/dmitrynikotin", "content": "Про политику, простым языком о сложном. Контрпропаганда и интерпретация новостей."}
+{"idx": 5, "title": "ГДЗ по математике 5 класс учебник Виленкин 2 часть – номер 6.365", "date": "", "ddg_snippet": "Найдите корень уравнения: а) 2,442 : (6, 8 - х) = 8 ,14; б) 3,6 ∙ (1,3 + у) = 7,56; в) 0,3n + 1,6n - 0,31 = 0,45; г) 8 ,3m - 4m - 1, 8 m + 4,25 = 8 . Ответ", "subpage_snippet": "", "source": "gdz-raketa.ru", "link": "https://gdz-raketa.ru/matematika/5-klass/uchebnik-vilenkin/2-chast-nomer-6-365/", "content": "Найдите корень уравнения: а) 2,442 : (6, 8 - х) = 8 ,14; б) 3,6 ∙ (1,3 + у) = 7,56; в) 0,3n + 1,6n - 0,31 = 0,45; г) 8 ,3m - 4m - 1, 8 m + 4,25 = 8 . Ответ"}
+{"idx": 6, "title": "Составь программу действий и вычисли: a) 4045 : 5 + 451 * 75 ...", "date": "", "ddg_snippet": "Выполним вычитание в скобках: 729 - 642 = 87. Выполним деление: 4045 : 5 = 809. Выполним умножение: 451 * 75 = 33825.", "subpage_snippet": "", "source": "www.euroki.org", "link": "https://www.euroki.org/koza/sostav-programmu-deystviy-i-vychisli-a------------b", "content": "Выполним вычитание в скобках: 729 - 642 = 87. Выполним деление: 4045 : 5 = 809. Выполним умножение: 451 * 75 = 33825."}
+{"idx": 7, "title": "Как решить: Ореховая смесь из фундука и миндаля имеет массу...", "date": "", "ddg_snippet": "После съедения 25% миндаля осталось: 0. 75 (692−x). Составим уравнение по условиюОстаток миндаля: 0. 75 ⋅652=489 г. Разница: 489−14=475 г. автор вопроса выбрал этот ответ лучшим.", "subpage_snippet": "", "source": "www.bolshoyvopros.ru", "link": "https://www.bolshoyvopros.ru/questions/4608232-kak-reshit-orehovaja-smes-iz-funduka-i-mindalja-imeet-massu-692-g.html", "content": "После съедения 25% миндаля осталось: 0. 75 (692−x). Составим уравнение по условиюОстаток миндаля: 0. 75 ⋅652=489 г. Разница: 489−14=475 г. автор вопроса выбрал этот ответ лучшим."}
+{"idx": 8, "title": "Выбираем смартфон до 15000 рублей | сравнение...", "date": "", "ddg_snippet": "ООО \"Алибаба.ком (ру)\" ИНН 7703380158. Сегодня мы сравним новый POCO M7 Pro 5G и защищенный бюджетник Realme C 75 . POCO мощнее и получил лучшие возможности камер, зато Realme имеет сертификат IP69 и готов выдерживать любые невзгоды.", "subpage_snippet": "", "source": "vk.com", "link": "https://vk.com/video-94973034_456239955", "content": "ООО \"Алибаба.ком (ру)\" ИНН 7703380158. Сегодня мы сравним новый POCO M7 Pro 5G и защищенный бюджетник Realme C 75 . POCO мощнее и получил лучшие возможности камер, зато Realme имеет сертификат IP69 и готов выдерживать любые невзгоды."}
+{"idx": 9, "title": "Почему так дешево? Экстремальная распаковка и обзор игрового...", "date": "", "ddg_snippet": "Экстремальная распаковка и обзор игрового монитора Thunderobot F23H 75 » на канале «Обзоры и тесты игровых устройств» в хорошем качестве и бесплатно, опубликованное 29 декабря 2023 года в 17:18, длительностью 00:12:46, на видеохостинге RUTUBE.", "subpage_snippet": "", "source": "rutube.ru", "link": "https://rutube.ru/video/8ffc944c4385ae5343cda60132a8a2e4/", "content": "Экстремальная распаковка и обзор игрового монитора Thunderobot F23H 75 » на канале «Обзоры и тесты игровых устройств» в хорошем качестве и бесплатно, опубликованное 29 декабря 2023 года в 17:18, длительностью 00:12:46, на видеохостинге RUTUBE."}
diff --git a/data/sampled_jsons/Foret_et_al._2021_Sharpness-Aware_Minimization_SAM_abstract.jsonl b/data/sampled_jsons/Foret_et_al._2021_Sharpness-Aware_Minimization_SAM_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c4569de5ece0309d1ead505b4f795f0a7a280079
--- /dev/null
+++ b/data/sampled_jsons/Foret_et_al._2021_Sharpness-Aware_Minimization_SAM_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "QUERY function - Google Docs Editors Help", "date": "", "ddg_snippet": "QUERY function Runs a Google Visualization API Query Language query across data. Sample Usage QUERY(A2:E6,\"select avg(A) pivot B\") QUERY(A2:E6,F2,FALSE) Syntax QUERY(data, query, [headers]) data - The range of cells to perform the query on. Each column of data can only hold boolean, numeric (including date/time types) or string values.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/docs/answer/3093343?hl=en", "content": "QUERY function Runs a Google Visualization API Query Language query across data. Sample Usage QUERY(A2:E6,\"select avg(A) pivot B\") QUERY(A2:E6,F2,FALSE) Syntax QUERY(data, query, [headers]) data - The range of cells to perform the query on. Each column of data can only hold boolean, numeric (including date/time types) or string values."}
+{"idx": 1, "title": "QUERY function - Google Docs Editors Help", "date": "", "ddg_snippet": "QUERY function Runs a Google Visualisation API Query Language query across data. Sample usage QUERY(A2:E6,'select avg(A) pivot B') QUERY(A2:E6,F2,FALSE) Syntax QUERY(data, query, [headers]) data – The range of cells to perform the query on. Each column of data can only hold boolean, numeric (including date/time types) or string values.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/docs/answer/3093343?hl=en-GB", "content": "QUERY function Runs a Google Visualisation API Query Language query across data. Sample usage QUERY(A2:E6,'select avg(A) pivot B') QUERY(A2:E6,F2,FALSE) Syntax QUERY(data, query, [headers]) data – The range of cells to perform the query on. Each column of data can only hold boolean, numeric (including date/time types) or string values."}
+{"idx": 2, "title": "[video] [GOOGLE SHEETS] FUNCIÓN QUERY: FUNCIONES DE AGREGACIÓN:...", "date": "", "ddg_snippet": "Ver en [GOOGLE SHEETS] FUNCIÓN QUERY: FUNCIONES DE AGREGACIÓN: SUM, AVG, COUNT, MIN y MAX 35 mil visualizaciones 6 votos a favor", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/docs/community-video/231737600/google-sheets-funciÓn-query-funciones-de-agregaciÓn-sum-avg-count-min-y-max?hl=es", "content": "Ver en [GOOGLE SHEETS] FUNCIÓN QUERY: FUNCIONES DE AGREGACIÓN: SUM, AVG, COUNT, MIN y MAX 35 mil visualizaciones 6 votos a favor"}
+{"idx": 3, "title": "Refine searches in Gmail - Computer - Gmail Help - Google Help", "date": "", "ddg_snippet": "Use a search operator On your computer, go to Gmail. At the top, click the search box. Enter a search operator. Tips: After you search, you can use the results to set up a filter for these messages. When using numbers as part of your query, a space or a dash (-) will separate a number while a dot (.) will be a decimal. For example, 01.2047-100 is considered 2 numbers: 01.2047 and 100.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/mail/answer/7190?hl=en&co=GENIE.Platform=Desktop", "content": "Use a search operator On your computer, go to Gmail. At the top, click the search box. Enter a search operator. Tips: After you search, you can use the results to set up a filter for these messages. When using numbers as part of your query, a space or a dash (-) will separate a number while a dot (.) will be a decimal. For example, 01.2047-100 is considered 2 numbers: 01.2047 and 100."}
+{"idx": 4, "title": "Google payments center help", "date": "", "ddg_snippet": "Official Google payments center Help Center where you can find tips and tutorials on using Google payments center and other answers to frequently asked questions.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/paymentscenter/?hl=en", "content": "Official Google payments center Help Center where you can find tips and tutorials on using Google payments center and other answers to frequently asked questions."}
+{"idx": 5, "title": "QUERY - Google 文件編輯器說明", "date": "", "ddg_snippet": "QUERY(A2:E6,F2,FALSE) 語法 QUERY(資料, 查詢, [標題]) 資料 - 查詢作業針對的儲存格範圍。 Each column of data can only hold boolean, numeric (including date/time types) or string values. In case of mixed data types in a single column, the majority data type determines the data type of the column for query purposes.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/docs/answer/3093343?hl=zh-Hant", "content": "QUERY(A2:E6,F2,FALSE) 語法 QUERY(資料, 查詢, [標題]) 資料 - 查詢作業針對的儲存格範圍。 Each column of data can only hold boolean, numeric (including date/time types) or string values. In case of mixed data types in a single column, the majority data type determines the data type of the column for query purposes."}
+{"idx": 6, "title": "Set default search engine and site search shortcuts", "date": "", "ddg_snippet": "Enter the web address for the search engine's results page, and use %s where the query would go. To find and edit the web address of the results page: Copy and paste the web address of the search results page into the URL field. The address for the search results page is different from the website address.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/chrome/answer/95426?hl=en&co=GENIE.Platform=Desktop", "content": "Enter the web address for the search engine's results page, and use %s where the query would go. To find and edit the web address of the results page: Copy and paste the web address of the search results page into the URL field. The address for the search results page is different from the website address."}
+{"idx": 7, "title": "BigQuery - Google Cloud Platform Console Help", "date": "", "ddg_snippet": "Use a variety of third-party tools to access data on BigQuery, such as tools that load or visualize your data. Use datasets to organize and control access to tables, and construct jobs for BigQuery to execute (load, export, query, or copy data). Find BigQuery in the left side menu of the Google Cloud Platform Console, under Big Data.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/cloud/answer/6255052?hl=en", "content": "Use a variety of third-party tools to access data on BigQuery, such as tools that load or visualize your data. Use datasets to organize and control access to tables, and construct jobs for BigQuery to execute (load, export, query, or copy data). Find BigQuery in the left side menu of the Google Cloud Platform Console, under Big Data."}
+{"idx": 8, "title": "Url with %s in place of query - Google Chrome Community", "date": "", "ddg_snippet": "Url with %s in place of query What is google chrome's query link? I know this sounds stupid but is there a search engine called Google chrome instead of google, I told my friend about my search engine situation and he said that Google search engine is the problem and that I should get Google chrome, but I already have google chrome.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/chrome/thread/171175871/url-with-s-in-place-of-query?hl=en", "content": "Url with %s in place of query What is google chrome's query link? I know this sounds stupid but is there a search engine called Google chrome instead of google, I told my friend about my search engine situation and he said that Google search engine is the problem and that I should get Google chrome, but I already have google chrome."}
+{"idx": 9, "title": "Query on/in/about/regarding... | WordReference Forums", "date": "", "ddg_snippet": "Jan 21, 2017 · Good afternoon all, I was wondering if I could use the following prepositions or prepositional phrases with \"query\" I have a question in this matter I have a question on this matter I have a question about this matter I have a question regarding / with regard to / concerning / other similar...", "subpage_snippet": "", "source": "forum.wordreference.com", "link": "https://forum.wordreference.com/threads/query-on-in-about-regarding.3278284/", "content": "Jan 21, 2017 · Good afternoon all, I was wondering if I could use the following prepositions or prepositional phrases with \"query\" I have a question in this matter I have a question on this matter I have a question about this matter I have a question regarding / with regard to / concerning / other similar..."}
diff --git a/data/sampled_jsons/GUI-Xplore_Empowering_Generalizable_GUI_Agents_Table_4_rule-based_GPT-based_clustering_year_2024.jsonl b/data/sampled_jsons/GUI-Xplore_Empowering_Generalizable_GUI_Agents_Table_4_rule-based_GPT-based_clustering_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..052a1f34513e168f5c86886c70b9f7bb3e5651a8
--- /dev/null
+++ b/data/sampled_jsons/GUI-Xplore_Empowering_Generalizable_GUI_Agents_Table_4_rule-based_GPT-based_clustering_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Empowering Generalizable GUI Agents with One Exploration", "date": "", "ddg_snippet": "by Y Sun · 2025 · Cited by 9 — Xplore - Agent encodes GUI modeling-generated linear se- quences into a GUI Transition Graph through GUI clus- tering, where clustering accuracy is critical for ... 10 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Sun_GUI-Xplore_Empowering_Generalizable_GUI_Agents_with_One_Exploration_CVPR_2025_paper.pdf", "content": "by Y Sun · 2025 · Cited by 9 — Xplore - Agent encodes GUI modeling-generated linear se- quences into a GUI Transition Graph through GUI clus- tering, where clustering accuracy is critical for ... 10 pages"}
+{"idx": 1, "title": "Empowering Generalizable GUI Agents with One Exploration", "date": "", "ddg_snippet": "22 Mar 2025 — We introduce GUI-Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17709v1", "content": "22 Mar 2025 — We introduce GUI-Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework."}
+{"idx": 2, "title": "Graphs Meet AI Agents: Taxonomy, Progress, and Future ...", "date": "", "ddg_snippet": "22 Jun 2025 — This survey presents a first systematic review of how graphs can empower AI agents . Specifically, we explore the integration of graph techniques with core ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.18019v1", "content": "22 Jun 2025 — This survey presents a first systematic review of how graphs can empower AI agents . Specifically, we explore the integration of graph techniques with core ..."}
+{"idx": 3, "title": "GUI-Xplore: Empowering Generalizable GUI Agents with One ... GitHub - 921112343/GUI-Xplore: [CVPR 2025] GUI-Xplore ... CVPR 2025 Open Access Repository GUI-Xplore: Empowering Generalizable GUI Agents with One ... GUI-Xplore: Empowering Generalizable GUI Agents with One ... GUI-Xplore/README.md at main · 921112343/GUI-Xplore · GitHub GitHub - 921112343/ GUI-Xplore : [CVPR 2025] GUI-Xplore : Empowering GUI -Xplore: Empowering Generalizable GUI Agents with One Exploration GUI - Xplore : Empowering Generalizable GUI Agents with One Exploratio… GUI -Xplore: Empowering Generalizable GUI Agents with One Exploration GUI -Xplore: Empowering Generalizable GUI Agents with One Exploration GUI - Xplore : Empowering Generalizable GUI Agents with One Exploratio… GUI-Xplore: Empowering Generalizable GUI Agents with One ...", "date": "", "ddg_snippet": "Mar 22, 2025 · To address these challenges, we introduce GUI - Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. To address these challenges, we introduce GUI - Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. To address these challenges, we in-troduce GUI - Xplore , a dataset meticulously designed to en-hance cross-application and cross-task generalization via an exploration-and-reasoning framework. Abstract: GUI agents hold significant potential to enhance the experience and efficiency of human-device interaction. However, current methods face challenges in generalizing across applications (apps) and tasks, primarily due to two fundamental limitations in existing datasets. Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. What is GUI-Xplore dataset? GUI-Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. [2025-03-18] Initial release of GUI-Xplore dataset and benchmark. What is GUI-Xplore? To address these challenges, we introduce GUI-Xplore, a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. What are the challenges of GUI-Xplore app exploration? The App Exploration task in the GUI-Xplore dataset presents challenges due to its information density and the multi-layered nature of the downstream tasks . To address these challenges, we introduce a two-stage framework for GUI agents, as shown in Fig 3. What is GUI-Xplore agent? To fully exploit GUI-Xplore's unique features, we propose Xplore-Agent, a GUI agent framework that combines Action-aware GUI Modeling with Graph-Guided Environment Reasoning. What is GUI Xplore? GUI-Xplore integrates pre-recorded exploration videos providing contextual insights, alongside five hierarchically structured downstream tasks designed to comprehensively evaluate GUI agent capabilities. Do task discrepancies affect cross-app generalization? In the Cross-App Generalization experiment, to mitigate the impact of task discrepancies when comparing against exist-ing GUI agent methods, we constructed a Cross-App Au-tomation test set based on the Application Usage task from GUI - Xplore dataset. While GUI - Xplore represents a significant advancement in GUI agents , several limitations deserve consideration. First, the exploration phase still requires substantial computational resources and time, which could limit practical applications in resource-constrained environments.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.17709", "content": "Mar 22, 2025 · To address these challenges, we introduce GUI - Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. To address these challenges, we introduce GUI - Xplore , a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. To address these challenges, we in-troduce GUI - Xplore , a dataset meticulously designed to en-hance cross-application and cross-task generalization via an exploration-and-reasoning framework. Abstract: GUI agents hold significant potential to enhance the experience and efficiency of human-device interaction. However, current methods face challenges in generalizing across applications (apps) and tasks, primarily due to two fundamental limitations in existing datasets. Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. What is GUI-Xplore dataset? GUI-Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation. [2025-03-18] Initial release of GUI-Xplore dataset and benchmark. What is GUI-Xplore? To address these challenges, we introduce GUI-Xplore, a dataset meticulously designed to enhance cross-application and cross-task generalization via an exploration-and-reasoning framework. What are the challenges of GUI-Xplore app exploration? The App Exploration task in the GUI-Xplore dataset presents challenges due to its information density and the multi-layered nature of the downstream tasks . To address these challenges, we introduce a two-stage framework for GUI agents, as shown in Fig 3. What is GUI-Xplore agent? To fully exploit GUI-Xplore's unique features, we propose Xplore-Agent, a GUI agent framework that combines Action-aware GUI Modeling with Graph-Guided Environment Reasoning. What is GUI Xplore? GUI-Xplore integrates pre-recorded exploration videos providing contextual insights, alongside five hierarchically structured downstream tasks designed to comprehensively evaluate GUI agent capabilities. Do task discrepancies affect cross-app generalization? In the Cross-App Generalization experiment, to mitigate the impact of task discrepancies when comparing against exist-ing GUI agent methods, we constructed a Cross-App Au-tomation test set based on the Application Usage task from GUI - Xplore dataset. While GUI - Xplore represents a significant advancement in GUI agents , several limitations deserve consideration. First, the exploration phase still requires substantial computational resources and time, which could limit practical applications in resource-constrained environments."}
+{"idx": 4, "title": "GitHub - 921112343/GUI-Xplore: [CVPR 2025] GUI-Xplore ...", "date": "", "ddg_snippet": "Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/921112343/GUI-Xplore", "content": "Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation."}
+{"idx": 5, "title": "GUI-Xplore: Empowering Generalizable GUI Agents with One ...", "date": "", "ddg_snippet": "To address these challenges, we in-troduce GUI - Xplore , a dataset meticulously designed to en-hance cross-application and cross-task generalization via an exploration-and-reasoning framework.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.17709", "content": "To address these challenges, we in-troduce GUI - Xplore , a dataset meticulously designed to en-hance cross-application and cross-task generalization via an exploration-and-reasoning framework."}
+{"idx": 6, "title": "GUI-Xplore: Empowering Generalizable GUI Agents with One ...", "date": "", "ddg_snippet": "Abstract: GUI agents hold significant potential to enhance the experience and efficiency of human-device interaction. However, current methods face challenges in generalizing across applications (apps) and tasks, primarily due to two fundamental limitations in existing datasets.", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/overview/2503.17709v1", "content": "Abstract: GUI agents hold significant potential to enhance the experience and efficiency of human-device interaction. However, current methods face challenges in generalizing across applications (apps) and tasks, primarily due to two fundamental limitations in existing datasets."}
+{"idx": 7, "title": "GUI-Xplore/README.md at main · 921112343/GUI-Xplore · GitHub", "date": "", "ddg_snippet": "Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/921112343/GUI-Xplore/blob/main/README.md", "content": "Feb 27, 2025 · GUI - Xplore is a novel dataset designed to enhance cross-application and cross-task generalization for GUI agents . It provides exploration videos for diverse applications, combined with five structured downstream tasks, aiming to push the boundaries of generalizable GUI automation."}
+{"idx": 8, "title": "GUI-Xplore: Empowering Generalizable GUI Agents with One ...", "date": "", "ddg_snippet": "While GUI - Xplore represents a significant advancement in GUI agents , several limitations deserve consideration. First, the exploration phase still requires substantial computational resources and time, which could limit practical applications in resource-constrained environments.", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/gui-xplore-empowering-generalizable-gui-agents-one", "content": "While GUI - Xplore represents a significant advancement in GUI agents , several limitations deserve consideration. First, the exploration phase still requires substantial computational resources and time, which could limit practical applications in resource-constrained environments."}
+{"idx": 9, "title": "Transformation of industrial robotics with natural language ...", "date": "", "ddg_snippet": "by Z Yu · 2026 — Rule - based interactive natural language processing, combined with 3D visual perception for environmental mapping and rule-scene matching to translate human ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S073658452500167X", "content": "by Z Yu · 2026 — Rule - based interactive natural language processing, combined with 3D visual perception for environmental mapping and rule-scene matching to translate human ..."}
diff --git a/data/sampled_jsons/Grassmannian_cluster_algebra_148658_S18_characters_100000_training_examples.jsonl b/data/sampled_jsons/Grassmannian_cluster_algebra_148658_S18_characters_100000_training_examples.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4f14655a220947d3246e8556421e1237420feaed
--- /dev/null
+++ b/data/sampled_jsons/Grassmannian_cluster_algebra_148658_S18_characters_100000_training_examples.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Grassmannian - Wikipedia", "date": "", "ddg_snippet": "In mathematics, the Grassmannian (named in honour of Hermann Grassmann) is a differentiable manifold that parameterizes the set of all - dimensional linear subspaces of an -dimensional vector space over a field that has a differentiable structure.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Grassmannian", "content": "In mathematics, the Grassmannian (named in honour of Hermann Grassmann) is a differentiable manifold that parameterizes the set of all - dimensional linear subspaces of an -dimensional vector space over a field that has a differentiable structure."}
+{"idx": 1, "title": "[math/0311148] Grassmannians and Cluster Algebras - arXiv.org", "date": "", "ddg_snippet": "Nov 10, 2003 · This paper demonstrates that the homogeneous coordinate ring of the Grassmannian G(k, n) is a {\\it cluster algebra of geometric type} - as defined by S. Fomin and A. Zelevinsky. Grassmannians having {\\it finite cluster type} are classified and the associated cluster variables are studied in connection with the geometry of configurations of points in RP2.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/math/0311148", "content": "Nov 10, 2003 · This paper demonstrates that the homogeneous coordinate ring of the Grassmannian G(k, n) is a {\\it cluster algebra of geometric type} - as defined by S. Fomin and A. Zelevinsky. Grassmannians having {\\it finite cluster type} are classified and the associated cluster variables are studied in connection with the geometry of configurations of points in RP2."}
+{"idx": 2, "title": "Lecture Notes Introduction to Cluster Algebra", "date": "", "ddg_snippet": "The cluster algebra structure of the Grassmannian leads to the notion of \\Positive Grassmannian \", where each cluster gives a \\total positivity cri-terion\" on the positive variety of Gr(k; n), generalizing the results for total positivity of matrices.", "subpage_snippet": "", "source": "www.math.hkust.edu.hk", "link": "https://www.math.hkust.edu.hk/~ivanip/teaching/cluster/Lecture11-Grassmannian.pdf", "content": "The cluster algebra structure of the Grassmannian leads to the notion of \\Positive Grassmannian \", where each cluster gives a \\total positivity cri-terion\" on the positive variety of Gr(k; n), generalizing the results for total positivity of matrices."}
+{"idx": 3, "title": "Grassmannians and Cluster Structures - PMC Grassmannian categories of infinite rank - joint with Jenny ... Grassmannian - Wikipedia Clustering cluster algebras with clusters - LIMS Lecture Notes Introduction to Cluster Algebra Grassmannian - Wikipedia Lecture Notes Introduction to Cluster Algebra Grassmannian - Wikipedia Grassmannian - Wikipedia Grassmannian - Wikipedia GRASSMANNIANS AND CLUSTER ALGEBRAS - University of Bath", "date": "", "ddg_snippet": "Apr 22, 2021 · In the finite types, both the cluster algebra and the cluster category are determined. However, in general, the categories F k, n are of infinite type and so far not well understood. In these lectures, we recall the cluster algebras and cluster categories of the Grassmannian and some of the key ingredients for studying them. Grassmannian cluster algebras of Theorem (Scott 2006) nite rank Maximal sets of compatible Plucker coordinates are ( examples of) clusters. In mathematics, the Grassmannian (named in honour of Hermann Grassmann) is a differentiable manifold that parameterizes the set of all - dimensional linear subspaces of an -dimensional vector space over a field that has a differentiable structure. For example , in mathematics, cluster variables in Grassmannian clus-ter algebras C[Gr(k, n)] correspond to real prime modules of the quantum affine algebra modules in Grassmannian cluster categories [9]. Is Grassmannian a cluster algebra? In this lecture, we will look at the general Grassmannian Gr(k; n), and show that it also prosesses a cluster algebra structure of geometric type . We will follow [Scott]. Remark 10.1. Is Grassmannian an algebraic group? More generally, over a ground field , the group is an algebraic group , and this construction shows that the Grassmannian is a non-singular algebraic variety. It follows from the existence of the Plücker embedding that the Grassmannian is complete as an algebraic variety. In particular, is a parabolic subgroup of . What is a positive Grassmannian? We will follow [Scott]. Remark 10.1. The cluster algebra structure of the Grassmannian leads to the notion of \\Positive Grassmannian\", where each cluster gives a otal positivity cri-terion \" on the positive variety of Gr(k; n), generalizing the results for total positivity of matrices. What is an example of a Grassmannian? For example, the Grassmannian is the space of lines through the origin in , so it is the same as the projective space of one dimension lower than . When is a real or complex vector space, Grassmannians are compact smooth manifolds, of dimension . In general they have the structure of a nonsingular projective algebraic variety. What is integral cohomology of Grassmannians? The integral cohomology of the Grassmannians is generated, as a ring , by the Chern classes of . In particular, all of the integral cohomology is at even degree as in the case of a projective space. These generators are subject to a set of relations, which defines the ring. What is the Plücker embedding of the Grassmannian? The Plücker embedding of the Grassmannian satisfies a set of simple quadratic relations called the Plücker relations . These show that the Grassmannian embeds as a nonsingular projective algebraic subvariety of the projectivization of the th exterior power of and give another method for constructing the Grassmannian. A cluster algebra is of finite type if it possess only finitely many cluster variables. We have already seen an example , namely the homogeneous coordinate ring of the Grassmannian (2, n).", "subpage_snippet": "", "source": "www.ncbi.nlm.nih.gov", "link": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8641944/", "content": "Apr 22, 2021 · In the finite types, both the cluster algebra and the cluster category are determined. However, in general, the categories F k, n are of infinite type and so far not well understood. In these lectures, we recall the cluster algebras and cluster categories of the Grassmannian and some of the key ingredients for studying them. Grassmannian cluster algebras of Theorem (Scott 2006) nite rank Maximal sets of compatible Plucker coordinates are ( examples of) clusters. In mathematics, the Grassmannian (named in honour of Hermann Grassmann) is a differentiable manifold that parameterizes the set of all - dimensional linear subspaces of an -dimensional vector space over a field that has a differentiable structure. For example , in mathematics, cluster variables in Grassmannian clus-ter algebras C[Gr(k, n)] correspond to real prime modules of the quantum affine algebra modules in Grassmannian cluster categories [9]. Is Grassmannian a cluster algebra? In this lecture, we will look at the general Grassmannian Gr(k; n), and show that it also prosesses a cluster algebra structure of geometric type . We will follow [Scott]. Remark 10.1. Is Grassmannian an algebraic group? More generally, over a ground field , the group is an algebraic group , and this construction shows that the Grassmannian is a non-singular algebraic variety. It follows from the existence of the Plücker embedding that the Grassmannian is complete as an algebraic variety. In particular, is a parabolic subgroup of . What is a positive Grassmannian? We will follow [Scott]. Remark 10.1. The cluster algebra structure of the Grassmannian leads to the notion of \\Positive Grassmannian\", where each cluster gives a otal positivity cri-terion \" on the positive variety of Gr(k; n), generalizing the results for total positivity of matrices. What is an example of a Grassmannian? For example, the Grassmannian is the space of lines through the origin in , so it is the same as the projective space of one dimension lower than . When is a real or complex vector space, Grassmannians are compact smooth manifolds, of dimension . In general they have the structure of a nonsingular projective algebraic variety. What is integral cohomology of Grassmannians? The integral cohomology of the Grassmannians is generated, as a ring , by the Chern classes of . In particular, all of the integral cohomology is at even degree as in the case of a projective space. These generators are subject to a set of relations, which defines the ring. What is the Plücker embedding of the Grassmannian? The Plücker embedding of the Grassmannian satisfies a set of simple quadratic relations called the Plücker relations . These show that the Grassmannian embeds as a nonsingular projective algebraic subvariety of the projectivization of the th exterior power of and give another method for constructing the Grassmannian. A cluster algebra is of finite type if it possess only finitely many cluster variables. We have already seen an example , namely the homogeneous coordinate ring of the Grassmannian (2, n)."}
+{"idx": 4, "title": "Grassmannian categories of infinite rank - joint with Jenny ...", "date": "", "ddg_snippet": "Grassmannian cluster algebras of Theorem (Scott 2006) nite rank Maximal sets of compatible Plucker coordinates are ( examples of) clusters.", "subpage_snippet": "", "source": "dept.math.lsa.umich.edu", "link": "https://dept.math.lsa.umich.edu/~speyer/DimerConference2020/slides/Schroll.pdf", "content": "Grassmannian cluster algebras of Theorem (Scott 2006) nite rank Maximal sets of compatible Plucker coordinates are ( examples of) clusters."}
+{"idx": 5, "title": "Clustering cluster algebras with clusters - LIMS", "date": "", "ddg_snippet": "For example , in mathematics, cluster variables in Grassmannian clus-ter algebras C[Gr(k, n)] correspond to real prime modules of the quantum affine algebra modules in Grassmannian cluster categories [9].", "subpage_snippet": "", "source": "lims.ac.uk", "link": "https://lims.ac.uk/documents/undefined-6.pdf", "content": "For example , in mathematics, cluster variables in Grassmannian clus-ter algebras C[Gr(k, n)] correspond to real prime modules of the quantum affine algebra modules in Grassmannian cluster categories [9]."}
+{"idx": 6, "title": "GRASSMANNIANS AND CLUSTER ALGEBRAS - University of Bath", "date": "", "ddg_snippet": "A cluster algebra is of finite type if it possess only finitely many cluster variables. We have already seen an example , namely the homogeneous coordinate ring of the Grassmannian (2, n).", "subpage_snippet": "", "source": "people.bath.ac.uk", "link": "https://people.bath.ac.uk/masadk/zzz/scott06.pdf", "content": "A cluster algebra is of finite type if it possess only finitely many cluster variables. We have already seen an example , namely the homogeneous coordinate ring of the Grassmannian (2, n)."}
+{"idx": 7, "title": "[1309.7301] A categorification of Grassmannian cluster algebras", "date": "", "ddg_snippet": "View a PDF of the paper titled A categorification of Grassmannian cluster algebras , by Bernt Tore Jensen and 1 other authors.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1309.7301", "content": "View a PDF of the paper titled A categorification of Grassmannian cluster algebras , by Bernt Tore Jensen and 1 other authors."}
+{"idx": 8, "title": "A categorification of Grassmannian cluster algebras - the University...", "date": "", "ddg_snippet": "More precisely, there is a cluster character dened on the category which maps the rigid indecomposable objects to the cluster variables and the maximal rigid objects to clusters .Dive into the research topics of 'A categorification of Grassmannian cluster algebras '.", "subpage_snippet": "", "source": "researchportal.bath.ac.uk", "link": "https://researchportal.bath.ac.uk/en/publications/a-categorification-of-grassmannian-cluster-algebras", "content": "More precisely, there is a cluster character dened on the category which maps the rigid indecomposable objects to the cluster variables and the maximal rigid objects to clusters .Dive into the research topics of 'A categorification of Grassmannian cluster algebras '."}
+{"idx": 9, "title": "A categorification of Grassmannian cluster algebras | Request PDF", "date": "", "ddg_snippet": "[18,Section 5]. In fact, the Grassmannian cluster category MCM μ n R (k,n) is stably equivalent to the category SubQ k studied by Geiß et al.The Jacobian algebra of this dual graph gives a categorification of the Grassmannian G(2, 10) in the example , see [JKS16].", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/305315125_A_categorification_of_Grassmannian_cluster_algebras", "content": "[18,Section 5]. In fact, the Grassmannian cluster category MCM μ n R (k,n) is stably equivalent to the category SubQ k studied by Geiß et al.The Jacobian algebra of this dual graph gives a categorification of the Grassmannian G(2, 10) in the example , see [JKS16]."}
diff --git a/data/sampled_jsons/Gumiho_speculative_decoding_Figure_4_serial_head_depth_accepted_tokens_speedup_year_2024.jsonl b/data/sampled_jsons/Gumiho_speculative_decoding_Figure_4_serial_head_depth_accepted_tokens_speedup_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..76fbd8becdf678b64f524bea3315126d320427be
--- /dev/null
+++ b/data/sampled_jsons/Gumiho_speculative_decoding_Figure_4_serial_head_depth_accepted_tokens_speedup_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative ...", "date": "", "ddg_snippet": "Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens , where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens , enabling efficient multi- token generation ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.10135", "content": "Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens , where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens , enabling efficient multi- token generation ..."}
+{"idx": 1, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative ...", "date": "", "ddg_snippet": "Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based on the theoretical finding that initial tokens in the draft sequence have a more significant impact on the overall accepted length. Gumiho employs a hybrid head design that combines serial and parallel components.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/AMD-AGI/Gumiho", "content": "Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based on the theoretical finding that initial tokens in the draft sequence have a more significant impact on the overall accepted length. Gumiho employs a hybrid head design that combines serial and parallel components."}
+{"idx": 2, "title": "Gumiho - a amd Collection - Hugging Face", "date": "", "ddg_snippet": "Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Paper • 2503.10135 •Published Mar 13 Upvote - Share collection View history Collection guide Browse collections", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/collections/amd/gumiho-684a3b7cbbe86ab23b393e9f", "content": "Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Paper • 2503.10135 •Published Mar 13 Upvote - Share collection View history Collection guide Browse collections"}
+{"idx": 3, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in...", "date": "", "ddg_snippet": "This paper proposes Gumiho , a hybrid architecture for speculative decoding in Large Language Models. The core idea is motivated by the insight that early tokens in a draft sequence are more critical for the overall acceptance rate than later ones. Gumiho operationalizes this by using a more accurate, serial Transformer-based head for early tokens and faster parallel MLP-based heads for later ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0ObGn4e1IS", "content": "This paper proposes Gumiho , a hybrid architecture for speculative decoding in Large Language Models. The core idea is motivated by the insight that early tokens in a draft sequence are more critical for the overall acceptance rate than later ones. Gumiho operationalizes this by using a more accurate, serial Transformer-based head for early tokens and faster parallel MLP-based heads for later ..."}
+{"idx": 4, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative ...", "date": "", "ddg_snippet": "To address this, Gumiho proposes an innovative hybrid model that integrates both serial and parallel heads . 🔍 Key Features: - Early Tokens : Utilizes an advanced Transformer architecture in a ...", "subpage_snippet": "", "source": "www.linkedin.com", "link": "https://www.linkedin.com/posts/abdullah-kasri_gumiho-a-hybrid-architecture-to-prioritize-activity-7306519551876161536-HoDU", "content": "To address this, Gumiho proposes an innovative hybrid model that integrates both serial and parallel heads . 🔍 Key Features: - Early Tokens : Utilizes an advanced Transformer architecture in a ..."}
+{"idx": 5, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative ...", "date": "", "ddg_snippet": "Specifically, given the critical importance of early tokens , we employ a sophisticated Transformer architecture for the early draft heads in a serial configuration to improve accuracy.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/389821466_Gumiho_A_Hybrid_Architecture_to_Prioritize_Early_Tokens_in_Speculative_Decoding", "content": "Specifically, given the critical importance of early tokens , we employ a sophisticated Transformer architecture for the early draft heads in a serial configuration to improve accuracy."}
+{"idx": 6, "title": "Speculative Decoding via Hybrid Drafting and Rollback-Aware Branch ...", "date": "", "ddg_snippet": "Recently, speculative decoding (SD) has emerged as a promising technique to ac- celerate LLM inference by employing a small draft model to propose draft tokens in advance, and validating them in parallel with the large target model. However, the existing SD methods still remain fundamentally constrained by their serialized execution, which causes the mutual waiting bubbles between the draft ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2506.01979", "content": "Recently, speculative decoding (SD) has emerged as a promising technique to ac- celerate LLM inference by employing a small draft model to propose draft tokens in advance, and validating them in parallel with the large target model. However, the existing SD methods still remain fundamentally constrained by their serialized execution, which causes the mutual waiting bubbles between the draft ..."}
+{"idx": 7, "title": "Geralt-Targaryen/Awesome-Speculative-Decoding - GitHub", "date": "", "ddg_snippet": "\" Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding \" [2025-03] [paper] EAGLE for the first two draft tokens , Medusa for the next 5.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/Geralt-Targaryen/Awesome-Speculative-Decoding", "content": "\" Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding \" [2025-03] [paper] EAGLE for the first two draft tokens , Medusa for the next 5."}
+{"idx": 8, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative ...", "date": "", "ddg_snippet": "View recent discussion. Abstract: Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens , where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens , enabling ...", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/overview/2503.10135v2", "content": "View recent discussion. Abstract: Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens , where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens , enabling ..."}
+{"idx": 9, "title": "Accelerating LLMs with Smarter Token Prioritization", "date": "", "ddg_snippet": "Combines serial and parallel processing for optimal token generation Prioritizes early tokens that have higher likelihood of being correct Achieves up to 1.6x speedup over baseline speculative decoding methods Offers flexible deployment options with minimal overhead", "subpage_snippet": "", "source": "www.zerna.io", "link": "https://www.zerna.io/page/engineering/presentation_set/engineering-llm-research/presentation/engineering-model-optimization/slide/engineering-paper-2503_10135", "content": "Combines serial and parallel processing for optimal token generation Prioritizes early tokens that have higher likelihood of being correct Achieves up to 1.6x speedup over baseline speculative decoding methods Offers flexible deployment options with minimal overhead"}
diff --git a/data/sampled_jsons/Gumiho_speculative_decoding_Table_3_wall_time_tradeoff_openreview_year_2024.jsonl b/data/sampled_jsons/Gumiho_speculative_decoding_Table_3_wall_time_tradeoff_openreview_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7e96da3d1f6b1224c1807afcd73bc3c29d2ad0fd
--- /dev/null
+++ b/data/sampled_jsons/Gumiho_speculative_decoding_Table_3_wall_time_tradeoff_openreview_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Gumiho: A Hybrid Architecture to Prioritize ... - OpenReview", "date": "", "ddg_snippet": "May 1, 2025 · This paper proposes a new speculative decoding method called Gumiho . It combines the parallel draft head architecture and sequential draft head architecture to derive a hybrid architecture.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0ObGn4e1IS", "content": "May 1, 2025 · This paper proposes a new speculative decoding method called Gumiho . It combines the parallel draft head architecture and sequential draft head architecture to derive a hybrid architecture."}
+{"idx": 1, "title": "[2503.10135] Gumiho: A Hybrid Architecture to Prioritize ... GitHub - AMD-AGI/Gumiho: Official Implementation of \"Gumiho ... (PDF) Gumiho: A Hybrid Architecture to Prioritize Early ... Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ... Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ... Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ...", "date": "", "ddg_snippet": "Mar 13, 2025 · Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens, where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens, enabling efficient multi-token generation ... Jul 11, 2025 · This project implements Gumiho , a novel hybrid architecture designed to accelerate the auto-regressive token generation process of Large Language Models (LLMs) using speculative decoding . Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based ... Mar 13, 2025 · PDF | Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). In the following, we define yt+1 = ML(yt) ≜ ML(yt, kv<t) and ignore kv<t for simplicity. Speculative Decoding Auto-regressive text generation in LLMs is time -consuming. Speculative decoding addresses this limitation by employing a smaller, faster draft model MS to generate candidate tokens ahead of time . Abstract Gumiho , a hybrid model combining serial and parallel heads with varying architectures, improves speculative decoding for large language models by prioritizing accuracy for early tokens. Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Jinze Li · Yixing Xu · Haiduo Huang · Xuanwu Yin · Dong Li · Edith Ngai · Emad Barsoum", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.10135", "content": "Mar 13, 2025 · Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multiple heads to predict a sequence of future tokens, where each head handles a token in the sequence. The target LLM verifies the predicted sequence and accepts aligned tokens, enabling efficient multi-token generation ... Jul 11, 2025 · This project implements Gumiho , a novel hybrid architecture designed to accelerate the auto-regressive token generation process of Large Language Models (LLMs) using speculative decoding . Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based ... Mar 13, 2025 · PDF | Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). In the following, we define yt+1 = ML(yt) ≜ ML(yt, kv<t) and ignore kv<t for simplicity. Speculative Decoding Auto-regressive text generation in LLMs is time -consuming. Speculative decoding addresses this limitation by employing a smaller, faster draft model MS to generate candidate tokens ahead of time . Abstract Gumiho , a hybrid model combining serial and parallel heads with varying architectures, improves speculative decoding for large language models by prioritizing accuracy for early tokens. Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Jinze Li · Yixing Xu · Haiduo Huang · Xuanwu Yin · Dong Li · Edith Ngai · Emad Barsoum"}
+{"idx": 2, "title": "GitHub - AMD-AGI/Gumiho: Official Implementation of \"Gumiho ...", "date": "", "ddg_snippet": "Jul 11, 2025 · This project implements Gumiho , a novel hybrid architecture designed to accelerate the auto-regressive token generation process of Large Language Models (LLMs) using speculative decoding . Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/AMD-AGI/Gumiho", "content": "Jul 11, 2025 · This project implements Gumiho , a novel hybrid architecture designed to accelerate the auto-regressive token generation process of Large Language Models (LLMs) using speculative decoding . Unlike existing methods that treat all tokens within a generated sequence as equally important, Gumiho is based ..."}
+{"idx": 3, "title": "(PDF) Gumiho: A Hybrid Architecture to Prioritize Early ...", "date": "", "ddg_snippet": "Mar 13, 2025 · PDF | Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM).", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/389821466_Gumiho_A_Hybrid_Architecture_to_Prioritize_Early_Tokens_in_Speculative_Decoding", "content": "Mar 13, 2025 · PDF | Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM)."}
+{"idx": 4, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ...", "date": "", "ddg_snippet": "In the following, we define yt+1 = ML(yt) ≜ ML(yt, kv<t) and ignore kv<t for simplicity. Speculative Decoding Auto-regressive text generation in LLMs is time -consuming. Speculative decoding addresses this limitation by employing a smaller, faster draft model MS to generate candidate tokens ahead of time .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=0ObGn4e1IS", "content": "In the following, we define yt+1 = ML(yt) ≜ ML(yt, kv<t) and ignore kv<t for simplicity. Speculative Decoding Auto-regressive text generation in LLMs is time -consuming. Speculative decoding addresses this limitation by employing a smaller, faster draft model MS to generate candidate tokens ahead of time ."}
+{"idx": 5, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ...", "date": "", "ddg_snippet": "Abstract Gumiho , a hybrid model combining serial and parallel heads with varying architectures, improves speculative decoding for large language models by prioritizing accuracy for early tokens.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/2503.10135", "content": "Abstract Gumiho , a hybrid model combining serial and parallel heads with varying architectures, improves speculative decoding for large language models by prioritizing accuracy for early tokens."}
+{"idx": 6, "title": "Gumiho: A Hybrid Architecture to Prioritize Early Tokens in ...", "date": "", "ddg_snippet": "Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Jinze Li · Yixing Xu · Haiduo Huang · Xuanwu Yin · Dong Li · Edith Ngai · Emad Barsoum", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46686", "content": "Gumiho : A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding Jinze Li · Yixing Xu · Haiduo Huang · Xuanwu Yin · Dong Li · Edith Ngai · Emad Barsoum"}
+{"idx": 7, "title": "Policy-driven collaborative decoding for latency tradeoff | OpenReview", "date": "", "ddg_snippet": "Speculative decoding has shown promising improvements in decode times .In this paper, we propose to extend speculative decoding by learning a dynamic decoding policy that considers the trade - off between smaller and more complex models for generation and verification.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=_G6L10mWQkP", "content": "Speculative decoding has shown promising improvements in decode times .In this paper, we propose to extend speculative decoding by learning a dynamic decoding policy that considers the trade - off between smaller and more complex models for generation and verification."}
+{"idx": 8, "title": "Reducing AI Inference Latency With Speculative Decoding", "date": "", "ddg_snippet": "Explore how speculative decoding techniques, including EAGLE- 3 , reduce latency and enhance efficiency in AI inference, optimizing large language model performance on NVIDIA GPUs.", "subpage_snippet": "", "source": "digiitallife.com", "link": "https://digiitallife.com/reducing-ai-inference-latency-with-speculative-decoding/", "content": "Explore how speculative decoding techniques, including EAGLE- 3 , reduce latency and enhance efficiency in AI inference, optimizing large language model performance on NVIDIA GPUs."}
+{"idx": 9, "title": "timewall.io", "date": "", "ddg_snippet": "لموقع time wall .", "subpage_snippet": "", "source": "timewall.io", "link": "https://timewall.io/", "content": "لموقع time wall ."}
diff --git a/data/sampled_jsons/HDT_history-driven_target_empirical_measure_LRU_cache_implementation_memory_reduction.jsonl b/data/sampled_jsons/HDT_history-driven_target_empirical_measure_LRU_cache_implementation_memory_reduction.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7c6fb7f70beb41da33b8ef08b94ec6b8e13d1691
--- /dev/null
+++ b/data/sampled_jsons/HDT_history-driven_target_empirical_measure_LRU_cache_implementation_memory_reduction.jsonl
@@ -0,0 +1,8 @@
+{"idx": 0, "title": "Beyond Self-Repellent Kernels: History - Driven Target Towards...", "date": "", "ddg_snippet": "Scalable Implementation for Large Graphs: We implement an LRU ( Least Recently Used ) cache scheme to manage. 𝐱{\\mathbf{x}}bold_x. with limited memory .In this paper, we propose a history - driven target ( HDT ) framework for MCMC sampling on general graphs.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.18300v3", "content": "Scalable Implementation for Large Graphs: We implement an LRU ( Least Recently Used ) cache scheme to manage. 𝐱{\\mathbf{x}}bold_x. with limited memory .In this paper, we propose a history - driven target ( HDT ) framework for MCMC sampling on general graphs."}
+{"idx": 1, "title": "Beyond Self-Repellent Kernels: History-Driven Target Towards Efficient ...", "date": "", "ddg_snippet": "To handle memory limitations, we introduce a Least Recently Used ( LRU ) cache scheme, enabling partial tracking of the empirical measure without loss in sampling efficiency.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.18300v1", "content": "To handle memory limitations, we introduce a Least Recently Used ( LRU ) cache scheme, enabling partial tracking of the empirical measure without loss in sampling efficiency."}
+{"idx": 2, "title": "Beyond Self-Repellent Kernels: History-Driven Target ... - 道客巴巴", "date": "", "ddg_snippet": "To overcome theselimitations, instead of direct modif ication of tran-sition kernel, HDT introduces a history-dependenttarget distribution π [x] to replace the original tar-get µ in any graph sampler, where x representsthe empirical measure of past visits.", "subpage_snippet": "", "source": "www.doc88.com", "link": "https://www.doc88.com/p-28437834991562.html", "content": "To overcome theselimitations, instead of direct modif ication of tran-sition kernel, HDT introduces a history-dependenttarget distribution π [x] to replace the original tar-get µ in any graph sampler, where x representsthe empirical measure of past visits."}
+{"idx": 3, "title": "論文の概要: Beyond Self-Repellent Kernels: History-Driven Target Towards ...", "date": "", "ddg_snippet": "To overcome these limitations, instead of direct modification of transition kernel, HDT introduces a history-dependent target distribution $\\boldsymbol {\\pi} [\\mathbf {x}]$ to replace the original target $\\boldsymbol {\\mu}$ in any graph sampler, where $\\mathbf {x}$ represents the empirical measure of past visits.", "subpage_snippet": "", "source": "fugumt.com", "link": "https://fugumt.com/fugumt/paper_check/2505.18300v1", "content": "To overcome these limitations, instead of direct modification of transition kernel, HDT introduces a history-dependent target distribution $\\boldsymbol {\\pi} [\\mathbf {x}]$ to replace the original target $\\boldsymbol {\\mu}$ in any graph sampler, where $\\mathbf {x}$ represents the empirical measure of past visits."}
+{"idx": 4, "title": "Beyond Self-Repellent Kernels: History-Driven Target ...", "date": "", "ddg_snippet": "by J Hu — We propose a * history-driven target ( HDT )* framework in Markov Chain ... empirical measure of past visits. This design preserves ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0yzOEMbShU", "content": "by J Hu — We propose a * history-driven target ( HDT )* framework in Markov Chain ... empirical measure of past visits. This design preserves ..."}
+{"idx": 5, "title": "History-Driven Target Towards Efficient Nonlinear MCMC ...", "date": "", "ddg_snippet": "This findings open a possibility that we may only requires an approximated empirical measure to implement HDT -MCMC while remaining fast convergence.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46659", "content": "This findings open a possibility that we may only requires an approximated empirical measure to implement HDT -MCMC while remaining fast convergence."}
+{"idx": 6, "title": "Track: Poster Session 5 East", "date": "", "ddg_snippet": "17 Jul 2025 — We propose a * history-driven target ( HDT )* framework in Markov Chain ... empirical measure of past visits. This design preserves ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/session/50267", "content": "17 Jul 2025 — We propose a * history-driven target ( HDT )* framework in Markov Chain ... empirical measure of past visits. This design preserves ..."}
+{"idx": 7, "title": "Jie Hu", "date": "", "ddg_snippet": "Abstract:We propose a history-driven target ( HDT ) framework in Markov Chain ... empirical measure of past visits. This design preserves lightweight ...", "subpage_snippet": "", "source": "www.catalyzex.com", "link": "https://www.catalyzex.com/author/Jie+Hu", "content": "Abstract:We propose a history-driven target ( HDT ) framework in Markov Chain ... empirical measure of past visits. This design preserves lightweight ..."}
diff --git a/data/sampled_jsons/HTML_is_Better_Than_Plain_Text_for_Modeling_Retrieved_Knowledge_in_RAG_Systems_Table_4_Prune-Gen_LLM_year_2024.jsonl b/data/sampled_jsons/HTML_is_Better_Than_Plain_Text_for_Modeling_Retrieved_Knowledge_in_RAG_Systems_Table_4_Prune-Gen_LLM_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7e048c0d16facda51f64a897f2a5a4d788dd9efc
--- /dev/null
+++ b/data/sampled_jsons/HTML_is_Better_Than_Plain_Text_for_Modeling_Retrieved_Knowledge_in_RAG_Systems_Table_4_Prune-Gen_LLM_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved ...", "date": "", "ddg_snippet": "To alleviate this problem, we propose HtmlRAG, which uses HTML instead of plain text as the format of retrieved knowledge in RAG . We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new challenges.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2411.02959", "content": "To alleviate this problem, we propose HtmlRAG, which uses HTML instead of plain text as the format of retrieved knowledge in RAG . We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new challenges."}
+{"idx": 1, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieval Results ...", "date": "", "ddg_snippet": "We propose HtmlRAG, which uses HTML instead of plain text as the format of external knowledge in RAG systems . To tackle the long context brought by HTML , we propose Lossless HTML Cleaning and Two-Step Block-Tree-Based HTML Pruning.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/plageon/HtmlRAG", "content": "We propose HtmlRAG, which uses HTML instead of plain text as the format of external knowledge in RAG systems . To tackle the long context brought by HTML , we propose Lossless HTML Cleaning and Two-Step Block-Tree-Based HTML Pruning."}
+{"idx": 2, "title": "Understanding HtmlRAG: HTML is Better Than Plain Text for Modeling ...", "date": "", "ddg_snippet": "HtmlRAG introduces an innovative method for enhancing Retrieval-Augmented Generation ( RAG ) systems by using HTML over plain text for integrating external knowledge into large language models (LLMs). This approach preserves essential data structure, improving LLM performance and accuracy.", "subpage_snippet": "", "source": "techchilli.com", "link": "https://techchilli.com/artificial-intelligence/understanding-htmlrag-html-is-better-than-plain-text-for-modeling-retrieved-knowledge-in-rag-systems/", "content": "HtmlRAG introduces an innovative method for enhancing Retrieval-Augmented Generation ( RAG ) systems by using HTML over plain text for integrating external knowledge into large language models (LLMs). This approach preserves essential data structure, improving LLM performance and accuracy."}
+{"idx": 3, "title": "Implementing HtmlRAG: Enhancing Retrieval-Augmented ... - Dev Genius", "date": "", "ddg_snippet": "Retrieval-Augmented Generation ( RAG ) systems have gained significant traction in enhancing the knowledge capabilities of Large Language Models (LLMs) and mitigating issues like hallucinations. Traditional RAG systems typically use plain text extracted from HTML documents as the format for retrieved knowledge . However, this approach often leads to the loss of structural and semantic information ...", "subpage_snippet": "", "source": "blog.devgenius.io", "link": "https://blog.devgenius.io/implementing-htmlrag-enhancing-retrieval-augmented-generation-with-html-knowledge-91cdd6278e23", "content": "Retrieval-Augmented Generation ( RAG ) systems have gained significant traction in enhancing the knowledge capabilities of Large Language Models (LLMs) and mitigating issues like hallucinations. Traditional RAG systems typically use plain text extracted from HTML documents as the format for retrieved knowledge . However, this approach often leads to the loss of structural and semantic information ..."}
+{"idx": 4, "title": "Table 4 from HtmlRAG: HTML is Better Than Plain Text for Modeling ...", "date": "", "ddg_snippet": "Table 4 : Analysis of inference cost on ELI5 dataset We compare the chunking-based refiner using BGE (BGE), the two HTML pruning steps basing on the text embedding (PruneEmbed) and the generative model ( Prune -Gen) in HtmlRAG, and LLM chatting (LLM Chat) by model parameters, storage, average input tokens, and average output tokens.", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/HtmlRAG:-HTML-is-Better-Than-Plain-Text-for-in-RAG-Tan-Dou/7cfd2426ca908c8c5a81bd7c7ca01f914a972de4/figure/6", "content": "Table 4 : Analysis of inference cost on ELI5 dataset We compare the chunking-based refiner using BGE (BGE), the two HTML pruning steps basing on the text embedding (PruneEmbed) and the generative model ( Prune -Gen) in HtmlRAG, and LLM chatting (LLM Chat) by model parameters, storage, average input tokens, and average output tokens."}
+{"idx": 5, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved ...", "date": "", "ddg_snippet": "We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new challenges. HTML contains additional content such as tags, JavaScript, and CSS specifications, which bring extra input tokens and noise to the RAG system .", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/abs/10.1145/3696410.3714546", "content": "We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new challenges. HTML contains additional content such as tags, JavaScript, and CSS specifications, which bring extra input tokens and noise to the RAG system ."}
+{"idx": 6, "title": "HTML > Plain Text for RAG - by Pascal Biese - LLM Watch", "date": "", "ddg_snippet": "HtmlRAG addresses this issue by using HTML instead of plain text as the format for retrieved knowledge in RAG systems . The authors believe that HTML is better suited for modeling knowledge in external documents, and most LLMs have robust capabilities to understand HTML .", "subpage_snippet": "", "source": "www.llmwatch.com", "link": "https://www.llmwatch.com/p/html-plain-text-for-rag", "content": "HtmlRAG addresses this issue by using HTML instead of plain text as the format for retrieved knowledge in RAG systems . The authors believe that HTML is better suited for modeling knowledge in external documents, and most LLMs have robust capabilities to understand HTML ."}
+{"idx": 7, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved ...", "date": "", "ddg_snippet": "To alleviate this problem, we propose HtmlRAG, which uses HTML instead of plain text as the format of retrieved knowledge in RAG . We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new chal-lenges.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2411.02959v1", "content": "To alleviate this problem, we propose HtmlRAG, which uses HTML instead of plain text as the format of retrieved knowledge in RAG . We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML . However, utilizing HTML presents new chal-lenges."}
+{"idx": 8, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved ...", "date": "", "ddg_snippet": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems Jiejun Tan, Zhicheng Dou, Wen Wang, Mang Wang, Weipeng Chen, Ji-Rong Wen", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=leCEFyMyxg", "content": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems Jiejun Tan, Zhicheng Dou, Wen Wang, Mang Wang, Weipeng Chen, Ji-Rong Wen"}
+{"idx": 9, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved ...", "date": "", "ddg_snippet": "HTML contains a lot of intrinsic information, so we propose taking HTML as the format of retrieved knowledge in RAG systems , and design HTML cleaning and block-tree-based HTML pruning to shorten the length and preserve information.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/2411.02959", "content": "HTML contains a lot of intrinsic information, so we propose taking HTML as the format of retrieved knowledge in RAG systems , and design HTML cleaning and block-tree-based HTML pruning to shorten the length and preserve information."}
diff --git a/data/sampled_jsons/Here_we_introduce_trial-by-trial_error_consistency,_a_quantitative_analysis_for_measuring_whether_tw.jsonl b/data/sampled_jsons/Here_we_introduce_trial-by-trial_error_consistency,_a_quantitative_analysis_for_measuring_whether_tw.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..739e2fe022ceb96923881b5c947162e7d4a5f4bc
--- /dev/null
+++ b/data/sampled_jsons/Here_we_introduce_trial-by-trial_error_consistency,_a_quantitative_analysis_for_measuring_whether_tw.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2006.16736] Beyond accuracy: quantifying trial - by - trial behaviour of...", "date": "", "ddg_snippet": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2006.16736", "content": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 1, "title": "Beyond accuracy: quantifying trial-by-trial behaviour of CNNs and ...", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper/2020/hash/9f6992966d4c363ea0162a056cb45fe5-Abstract.html", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 2, "title": "Psychophysical experiment. A fixation cross is presented for 500 ms....", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Psychophysical-experiment-A-fixation-cross-is-presented-for-500-ms-Then-an-image-is_fig2_264038033", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 3, "title": "Beyond accuracy: quantifying trial - by - trial behaviour of CNNs and...", "date": "", "ddg_snippet": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "papertalk.org", "link": "https://papertalk.org/papertalks/10006", "content": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 4, "title": "Beyond accuracy: quantifying trial - by - trial behaviour of CNNs and...", "date": "", "ddg_snippet": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/342587815_Beyond_accuracy_quantifying_trial-by-trial_behaviour_of_CNNs_and_humans_by_measuring_error_consistency", "content": "Here we introduce trial - by - trial error consistency , a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 5, "title": "Beyond accuracy | Proceedings of the 34th International Conference on ...", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3495724.3496889", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 6, "title": "arXiv:2006.16736v3 [cs.CV] 18 Dec 2020", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs. Making consistent errors on a trial-by-trial basis is a necessary condition if we want to ascertain similar processing strategies between decision makers.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2006.16736v3", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs. Making consistent errors on a trial-by-trial basis is a necessary condition if we want to ascertain similar processing strategies between decision makers."}
+{"idx": 7, "title": "New conference paper accepted at NeurIPS 2020", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "uni-tuebingen.de", "link": "https://uni-tuebingen.de/en/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/neuronale-informationsverarbeitung/news/newsfullview-archive/article/new-conference-paper-accepted-at-neurips-2020/", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 8, "title": "New arcticle updoaded to arXiv - Universität Tübingen", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "uni-tuebingen.de", "link": "https://uni-tuebingen.de/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/neuronale-informationsverarbeitung/news/newsfullview-archive/article/new-arcticle-updoaded-to-arxiv/", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
+{"idx": 9, "title": "New conference paper accepted at NeurIPS 2020 - Universität Tübingen", "date": "", "ddg_snippet": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs.", "subpage_snippet": "", "source": "uni-tuebingen.de", "link": "https://uni-tuebingen.de/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/neuronale-informationsverarbeitung/news/newsfullview-aktuell/article/new-conference-paper-accepted-at-neurips-2020/", "content": "Here we introduce trial-by-trial error consistency, a quantitative analysis for measuring whether two decision making systems systematically make errors on the same inputs."}
diff --git a/data/sampled_jsons/Hierarchical_Overlapping_Clustering_on_Graphs.jsonl b/data/sampled_jsons/Hierarchical_Overlapping_Clustering_on_Graphs.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..140a9d026c5a6cc0c0be56d70b79a046cf4f97f9
--- /dev/null
+++ b/data/sampled_jsons/Hierarchical_Overlapping_Clustering_on_Graphs.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ICML Poster Hierarchical Overlapping Clustering on Graphs : Cost...", "date": "", "ddg_snippet": "Overlap and hierarchy are two prevalent phenomena in clustering , and usually coexist in a single system.To address this issue, we initiate the study of hierarchical overlapping clustering on graphs by introducing a new cost function for it.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46447", "content": "Overlap and hierarchy are two prevalent phenomena in clustering , and usually coexist in a single system.To address this issue, we initiate the study of hierarchical overlapping clustering on graphs by introducing a new cost function for it."}
+{"idx": 1, "title": "Hierarchical Single-Linkage Clustering for Community Detection with...", "date": "", "ddg_snippet": "Keywords Graph Clustering , Community Detection, Edge Clustering , Overlapping Clustering , Hierarchical Clustering . 1 Introduction.For overlapping clusters , we use ABCD+o2 [21] graphs that have both outliers and overlapping ground truth clusters .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.02334", "content": "Keywords Graph Clustering , Community Detection, Edge Clustering , Overlapping Clustering , Hierarchical Clustering . 1 Introduction.For overlapping clusters , we use ABCD+o2 [21] graphs that have both outliers and overlapping ground truth clusters ."}
+{"idx": 2, "title": "Overlapping community detection in weighted networks via...", "date": "", "ddg_snippet": "4 Proposed algorithm for overlapping hierarchical community detection in weighted networks. Hierarchical agglomerative clustering on the graph (GHAC) detects nested communities in a network using novel dissimilarity.", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11515960/", "content": "4 Proposed algorithm for overlapping hierarchical community detection in weighted networks. Hierarchical agglomerative clustering on the graph (GHAC) detects nested communities in a network using novel dissimilarity."}
+{"idx": 3, "title": "(PDF) Overlapping community detection in weighted networks via...", "date": "", "ddg_snippet": "In this work, we propose an algorithm for hierarchical overlapping community detection in weighted networks. The overlap between clusters is realized via maximal cliques that are used as base elements for hierarchical agglomerative clustering on the graph (GHAC).", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/385311779_Overlapping_community_detection_in_weighted_networks_via_hierarchical_clustering", "content": "In this work, we propose an algorithm for hierarchical overlapping community detection in weighted networks. The overlap between clusters is realized via maximal cliques that are used as base elements for hierarchical agglomerative clustering on the graph (GHAC)."}
+{"idx": 4, "title": "Overlapping Hierarchical Clustering (OHC) | SpringerLink", "date": "", "ddg_snippet": "In this paper we propose a new method that allows clusters to overlap until a strong cluster attraction is reached, based on a density criterion. The resulting hierarchical structure, called a quasi-dendrogram, is represented as a directed acyclic graph and combines the advantages of...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-030-44584-3_21", "content": "In this paper we propose a new method that allows clusters to overlap until a strong cluster attraction is reached, based on a density criterion. The resulting hierarchical structure, called a quasi-dendrogram, is represented as a directed acyclic graph and combines the advantages of..."}
+{"idx": 5, "title": "Overlapping Hierarchical Clustering (OHC)", "date": "", "ddg_snippet": "Overlapping Hierarchical Clustering (OHC). Ian Jeantet, Zoltan Miklos, David Gross-Amblard. Overlapping Hierarchical Clustering (OHC). IDA 2020, April 27 – 29 13 / 11. Cluster growth. How are the clusters growing? Let’s start with this (δi = 2.23)-neighbourhood graph .", "subpage_snippet": "", "source": "inria.hal.science", "link": "https://inria.hal.science/hal-02452729/file/55_Overlapping_Hierarchical_Clustering_IDA_2020.pdf", "content": "Overlapping Hierarchical Clustering (OHC). Ian Jeantet, Zoltan Miklos, David Gross-Amblard. Overlapping Hierarchical Clustering (OHC). IDA 2020, April 27 – 29 13 / 11. Cluster growth. How are the clusters growing? Let’s start with this (δi = 2.23)-neighbourhood graph ."}
+{"idx": 6, "title": "Finding communities by clustering a graph into overlapping subgraphs", "date": "", "ddg_snippet": "In this paper we present OSLOM (Order Statistics Local Optimization Method), the first method capable to detect clusters in networks accounting for edge directions, edge weights, overlapping communities, hierarchies and community dynamics.", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/54240030/Finding_communities_by_clustering_a_graph_into_overlapping_subgraphs", "content": "In this paper we present OSLOM (Order Statistics Local Optimization Method), the first method capable to detect clusters in networks accounting for edge directions, edge weights, overlapping communities, hierarchies and community dynamics."}
+{"idx": 7, "title": "Unsupervised Clustering : A Guide | Built In", "date": "", "ddg_snippet": "A hierarchical clustering algorithm is based on the union between the two nearest clusters . The beginning condition is realized by setting every data point as a cluster . After a few iterations it reaches the final clusters wanted.", "subpage_snippet": "", "source": "builtin.com", "link": "https://builtin.com/articles/unsupervised-clustering", "content": "A hierarchical clustering algorithm is based on the union between the two nearest clusters . The beginning condition is realized by setting every data point as a cluster . After a few iterations it reaches the final clusters wanted."}
+{"idx": 8, "title": "Overview of overlapping partitional clustering", "date": "", "ddg_snippet": "Basically, the existing overlapping clustering methods are extensions from usual clustering models such as hierarchical , generative, graph -based or partitional mod-els. Graph -based data clustering with overlaps . Discrete Optimization, 8(1):2–17.", "subpage_snippet": "", "source": "www.larodec.com", "link": "https://www.larodec.com/download/file/fid/432", "content": "Basically, the existing overlapping clustering methods are extensions from usual clustering models such as hierarchical , generative, graph -based or partitional mod-els. Graph -based data clustering with overlaps . Discrete Optimization, 8(1):2–17."}
+{"idx": 9, "title": "An Edge-Based Approach to Partitioning and Overlapping Graph ...", "date": "", "ddg_snippet": "Graph clustering is an NP-class problem [9] that can be classified into three distinct categories, partitioning, overlapping , and hierarchical clustering [2]. Section 3 elaborates on this further.", "subpage_snippet": "", "source": "www.mdpi.com", "link": "https://www.mdpi.com/2076-3417/14/1/380", "content": "Graph clustering is an NP-class problem [9] that can be classified into three distinct categories, partitioning, overlapping , and hierarchical clustering [2]. Section 3 elaborates on this further."}
diff --git a/data/sampled_jsons/HtmlRAG_LLM_Chat.jsonl b/data/sampled_jsons/HtmlRAG_LLM_Chat.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ec99ede4c5d2692a61ebc8dcda6d28e59a532dc6
--- /dev/null
+++ b/data/sampled_jsons/HtmlRAG_LLM_Chat.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling ...", "date": "", "ddg_snippet": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2411.02959v1", "content": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML."}
+{"idx": 1, "title": "Table 4 from HtmlRAG: HTML is Better Than Plain Text for ...", "date": "", "ddg_snippet": "HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG, and designs a two-step block-tree-based pruning method that prunes useless HTML blocks and keeps only the relevant part of the HTML.", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/HtmlRAG:-HTML-is-Better-Than-Plain-Text-for-in-RAG-Tan-Dou/7cfd2426ca908c8c5a81bd7c7ca01f914a972de4/figure/6", "content": "HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG, and designs a two-step block-tree-based pruning method that prunes useless HTML blocks and keeps only the relevant part of the HTML."}
+{"idx": 2, "title": "HtmlRAG: HTML is Better Than Plain Text for Modeling ...", "date": "", "ddg_snippet": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2411.02959v1", "content": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML."}
+{"idx": 3, "title": "Htmlrag: HTML Is Better Than Plain Text For Modeling ...", "date": "", "ddg_snippet": "To address this issue, we propose HTML cleaning, ( RAG ) utilizes a retrieval system to fetch external knowledge and. compression, and pruning strategies, to shorten the HTML while augment the LLM . It has proved effective in mitigating hallucina- minimizing the loss of information. Specifically, we design a two- tions of LLMs [41, 76].", "subpage_snippet": "", "source": "www.scribd.com", "link": "https://www.scribd.com/document/788595083/2411-02959v1", "content": "To address this issue, we propose HTML cleaning, ( RAG ) utilizes a retrieval system to fetch external knowledge and. compression, and pruning strategies, to shorten the HTML while augment the LLM . It has proved effective in mitigating hallucina- minimizing the loss of information. Specifically, we design a two- tions of LLMs [41, 76]."}
+{"idx": 4, "title": "[2411.02959] HtmlRAG: HTML is Better Than Plain Text for ...", "date": "", "ddg_snippet": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML.", "subpage_snippet": "", "source": "ar5iv.labs.arxiv.org", "link": "https://ar5iv.labs.arxiv.org/html/2411.02959", "content": "To alleviate this problem, we propose HtmlRAG , which uses HTML instead of plain text as the format of retrieved knowledge in RAG. We believe HTML is better than plain text in modeling knowledge in external documents, and most LLMs possess robust capacities to understand HTML."}
+{"idx": 5, "title": "LLM Research Papers: The 2024 List", "date": "", "ddg_snippet": "In the meantime, I want to share my running bookmark list of many fascinating (mostly LLM -related) papers I stumbled upon in 2024.", "subpage_snippet": "", "source": "magazine.sebastianraschka.com", "link": "https://magazine.sebastianraschka.com/p/llm-research-papers-the-2024-list", "content": "In the meantime, I want to share my running bookmark list of many fascinating (mostly LLM -related) papers I stumbled upon in 2024."}
+{"idx": 6, "title": "LLM Research Papers: The 2024 List", "date": "", "ddg_snippet": "I still plan to do so, but in the meantime, I want to share my running bookmark list of many fascinating (mostly LLM - related) papers I stumbled upon ...", "subpage_snippet": "", "source": "sebastianraschka.com", "link": "https://sebastianraschka.com/blog/2024/llm-research-papers-the-2024-list.html", "content": "I still plan to do so, but in the meantime, I want to share my running bookmark list of many fascinating (mostly LLM - related) papers I stumbled upon ..."}
+{"idx": 7, "title": "HtmlRAG論文読んでみた - Zenn", "date": "", "ddg_snippet": "Nov 24, 2024 · その中でも H tml RA G と呼ばれる手法の提案がおもしろそうだったので読んでみました。 そこで備忘録がてら記事にして投稿してみようと思った次第です。", "subpage_snippet": "", "source": "zenn.dev", "link": "https://zenn.dev/argus1/articles/06319726c51874", "content": "Nov 24, 2024 · その中でも H tml RA G と呼ばれる手法の提案がおもしろそうだったので読んでみました。 そこで備忘録がてら記事にして投稿してみようと思った次第です。"}
+{"idx": 8, "title": "‘ML dataset’ directory · Gwern.net", "date": "", "ddg_snippet": "... Memorization in Multiple-Choice LLM ... HtmlRAG : HTML Is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems ”, Tan et al 2024", "subpage_snippet": "", "source": "gwern.net", "link": "https://gwern.net/doc/ai/dataset/index", "content": "... Memorization in Multiple-Choice LLM ... HtmlRAG : HTML Is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems ”, Tan et al 2024"}
+{"idx": 9, "title": "‘retrieval AI’ directory · Gwern.net", "date": "", "ddg_snippet": "How ChatGPT Remembers You: A Deep ... HtmlRAG : HTML Is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems ”, Tan et al 2024", "subpage_snippet": "", "source": "gwern.net", "link": "https://gwern.net/doc/ai/nn/retrieval/index", "content": "How ChatGPT Remembers You: A Deep ... HtmlRAG : HTML Is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems ”, Tan et al 2024"}
diff --git a/data/sampled_jsons/ICML_2025_Normalizing_Flows_papers.jsonl b/data/sampled_jsons/ICML_2025_Normalizing_Flows_papers.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..660d85aea10328e170a28c54a06e988574e28006
--- /dev/null
+++ b/data/sampled_jsons/ICML_2025_Normalizing_Flows_papers.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ICML 2025 Normalizing Flows are Capable Generative Models Oral", "date": "", "ddg_snippet": "Oral Normalizing Flows are Capable Generative Models Shuangfei Zhai · Ruixiang Zhang · Preetum Nakkiran · David Berthelot · Jiatao Gu · Huangjie Zheng · Tianrong Chen · Miguel Angel Bautista Martin · Navdeep Jaitly · Joshua M Susskind West Ballroom A [ Abstract ] [ Visit Oral 6B Deep Learning Architectures ] Thu 17 Jul 4 p.m. — 4:15 ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/oral/47266", "content": "Oral Normalizing Flows are Capable Generative Models Shuangfei Zhai · Ruixiang Zhang · Preetum Nakkiran · David Berthelot · Jiatao Gu · Huangjie Zheng · Tianrong Chen · Miguel Angel Bautista Martin · Navdeep Jaitly · Joshua M Susskind West Ballroom A [ Abstract ] [ Visit Oral 6B Deep Learning Architectures ] Thu 17 Jul 4 p.m. — 4:15 ..."}
+{"idx": 1, "title": "[2412.06329] Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "View a PDF of the paper titled Normalizing Flows are Capable Generative Models, by Shuangfei Zhai and 9 other authors", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2412.06329", "content": "View a PDF of the paper titled Normalizing Flows are Capable Generative Models, by Shuangfei Zhai and 9 other authors"}
+{"idx": 2, "title": "Normalizing Flows are Capable Generative Models - Apple Machine ...", "date": "", "ddg_snippet": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years. In this work, we demonstrate that NFs are more powerful than previously believed. We present TarFlow: a simple and scalable architecture that enables highly ...", "subpage_snippet": "", "source": "machinelearning.apple.com", "link": "https://machinelearning.apple.com/research/normalizing-flows", "content": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years. In this work, we demonstrate that NFs are more powerful than previously believed. We present TarFlow: a simple and scalable architecture that enables highly ..."}
+{"idx": 3, "title": "ICML 2025 Accepted Paper List - Paper Copilot", "date": "", "ddg_snippet": "How to use the paper list below:- Overview: This table presents papers from the ICML conference, year 2025 .- Filtering: By default, the table loads the first 100 records. You can use the filter box under each column header to search within these loaded entries.", "subpage_snippet": "", "source": "papercopilot.com", "link": "https://papercopilot.com/paper-list/icml-paper-list/icml-2025-paper-list/", "content": "How to use the paper list below:- Overview: This table presents papers from the ICML conference, year 2025 .- Filtering: By default, the table loads the first 100 records. You can use the filter box under each column header to search within these loaded entries."}
+{"idx": 4, "title": "ICML.2025 | Cool Papers - Immersive Paper Discovery", "date": "", "ddg_snippet": "The list of accepted papers for ICML.2025 , including titles, authors, and abstracts, with support for paper interpretation based on Kimi AI.", "subpage_snippet": "", "source": "papers.cool", "link": "https://papers.cool/venue/ICML.2025", "content": "The list of accepted papers for ICML.2025 , including titles, authors, and abstracts, with support for paper interpretation based on Kimi AI."}
+{"idx": 5, "title": "ICML 2025 Papers", "date": "", "ddg_snippet": "Enable Javascript in your browser to see the papers page.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/papers.html?search=agent", "content": "Enable Javascript in your browser to see the papers page."}
+{"idx": 6, "title": "[2505.23527] Normalizing Flows are Capable Models for RL", "date": "", "ddg_snippet": "View a PDF of the paper titled Normalizing Flows are Capable Models for RL, by Raj Ghugare and Benjamin Eysenbach", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.23527", "content": "View a PDF of the paper titled Normalizing Flows are Capable Models for RL, by Raj Ghugare and Benjamin Eysenbach"}
+{"idx": 7, "title": "GitHub - serendipityerr/Diffusion-Papers-ICML-2025", "date": "", "ddg_snippet": "INRFlow: Flow Matching for INRs in Ambient Space. [pdf] Ensemble Distribution Distillation via Flow Matching. [pdf] Flexible Tails for Normalizing Flows . [pdf] Wasserstein Flow Matching: Generative Modeling Over Families of Distributions. [pdf] SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations. [pdf]", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/serendipityerr/Diffusion-Papers-ICML-2025", "content": "INRFlow: Flow Matching for INRs in Ambient Space. [pdf] Ensemble Distribution Distillation via Flow Matching. [pdf] Flexible Tails for Normalizing Flows . [pdf] Wasserstein Flow Matching: Generative Modeling Over Families of Distributions. [pdf] SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations. [pdf]"}
+{"idx": 8, "title": "Stanford AI Lab Papers and Talks at ICML 2025 | SAIL Blog", "date": "", "ddg_snippet": "The Forty-Second International Conference on Machine Learning ( ICML ) 2025 is being hosted in Vancouver from July 13 to July 19. We're excited to share all the work from SAIL that's being presented, and you'll find links to papers , videos and blogs below.", "subpage_snippet": "", "source": "ai.stanford.edu", "link": "https://ai.stanford.edu/blog/icml-2025/", "content": "The Forty-Second International Conference on Machine Learning ( ICML ) 2025 is being hosted in Vancouver from July 13 to July 19. We're excited to share all the work from SAIL that's being presented, and you'll find links to papers , videos and blogs below."}
+{"idx": 9, "title": "Normalizing Flows are Capable Generative Models - OpenReview", "date": "", "ddg_snippet": "Published: 01 May 2025 , Last Modified: 23 Jul 2025 ICML 2025 oral Everyone Revisions BibTeX CC BY 4.0 TL;DR: We show that normalizing flows can work great as a generative modeling principle, and propose a simple architecture and set of techniques to achieve it.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=2uheUFcFsM", "content": "Published: 01 May 2025 , Last Modified: 23 Jul 2025 ICML 2025 oral Everyone Revisions BibTeX CC BY 4.0 TL;DR: We show that normalizing flows can work great as a generative modeling principle, and propose a simple architecture and set of techniques to achieve it."}
diff --git a/data/sampled_jsons/Intervention_and_Conditioning_in_Causal_Bayesian_Networks_year_None.jsonl b/data/sampled_jsons/Intervention_and_Conditioning_in_Causal_Bayesian_Networks_year_None.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d88394a3e870cfda0a1927ebb08aa9b1b467cac7
--- /dev/null
+++ b/data/sampled_jsons/Intervention_and_Conditioning_in_Causal_Bayesian_Networks_year_None.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Watch Intervention Full Episodes, Video & More | A&E", "date": "", "ddg_snippet": "The Emmy award-winning series “Intervention” profiles people whose uncontrollable addiction to drugs, alcohol or compulsive behavior has brought them to the brink of destruction and has …", "subpage_snippet": "", "source": "www.aetv.com", "link": "https://www.aetv.com/shows/intervention", "content": "The Emmy award-winning series “Intervention” profiles people whose uncontrollable addiction to drugs, alcohol or compulsive behavior has brought them to the brink of destruction and has …"}
+{"idx": 1, "title": "INTERVENTION | English meaning - Cambridge Dictionary", "date": "", "ddg_snippet": "INTERVENTION definition: 1. intentional action to change a situation, with the aim of improving it or preventing it from…. Learn more.", "subpage_snippet": "", "source": "dictionary.cambridge.org", "link": "https://dictionary.cambridge.org/dictionary/english/intervention", "content": "INTERVENTION definition: 1. intentional action to change a situation, with the aim of improving it or preventing it from…. Learn more."}
+{"idx": 2, "title": "INTERVENTION Definition & Meaning - Merriam-Webster", "date": "", "ddg_snippet": "The meaning of INTERVENTION is the act or an instance of intervening. How to use intervention in a sentence.", "subpage_snippet": "", "source": "www.merriam-webster.com", "link": "https://www.merriam-webster.com/dictionary/intervention", "content": "The meaning of INTERVENTION is the act or an instance of intervening. How to use intervention in a sentence."}
+{"idx": 3, "title": "INTERVENTION Definition & Meaning | Dictionary.com", "date": "", "ddg_snippet": "Intervention definition: the act or fact of coming or occurring between two people, things, or times.. See examples of INTERVENTION used in a sentence.", "subpage_snippet": "", "source": "www.dictionary.com", "link": "https://www.dictionary.com/browse/intervention", "content": "Intervention definition: the act or fact of coming or occurring between two people, things, or times.. See examples of INTERVENTION used in a sentence."}
+{"idx": 4, "title": "Intervention - Definition, Meaning & Synonyms | Vocabulary.com", "date": "", "ddg_snippet": "An intervention is the act of inserting one thing between others, like a person trying to help. You could be the subject of a school intervention if your teachers call your parents about the bad …", "subpage_snippet": "", "source": "www.vocabulary.com", "link": "https://www.vocabulary.com/dictionary/intervention", "content": "An intervention is the act of inserting one thing between others, like a person trying to help. You could be the subject of a school intervention if your teachers call your parents about the bad …"}
+{"idx": 5, "title": "INTERVENTION definition and meaning | Collins English Dictionary", "date": "", "ddg_snippet": "An intervention is an occasion when friends or family tell someone who has a serious problem that they need to get help. Finally, his wife staged an intervention and he agreed to go into rehab. …", "subpage_snippet": "", "source": "www.collinsdictionary.com", "link": "https://www.collinsdictionary.com/dictionary/english/intervention", "content": "An intervention is an occasion when friends or family tell someone who has a serious problem that they need to get help. Finally, his wife staged an intervention and he agreed to go into rehab. …"}
+{"idx": 6, "title": "Intervention - definition of intervention by The Free Dictionary", "date": "", "ddg_snippet": "The act or process of intervening: a nation's military interventions in neighboring countries; a politician opposed to government intervention in the market economy.", "subpage_snippet": "", "source": "www.thefreedictionary.com", "link": "https://www.thefreedictionary.com/intervention", "content": "The act or process of intervening: a nation's military interventions in neighboring countries; a politician opposed to government intervention in the market economy."}
+{"idx": 7, "title": "intervention , n. meanings, etymology and more | Oxford English …", "date": "", "ddg_snippet": "intervention, n. meanings, etymology, pronunciation and more in the Oxford English Dictionary", "subpage_snippet": "", "source": "www.oed.com", "link": "https://www.oed.com/dictionary/intervention_n", "content": "intervention, n. meanings, etymology, pronunciation and more in the Oxford English Dictionary"}
+{"idx": 8, "title": "Intervention : Meaning, Definition, and Practical Applications", "date": "", "ddg_snippet": "Nov 3, 2024 · Meaning of the Word Intervention refers to the action or process of intervening in a situation, often to change or influence the outcome. It involves stepping into a situation with …", "subpage_snippet": "", "source": "meaningdictionary.com", "link": "https://meaningdictionary.com/intervention-meaning-definition-and-practical-applications/", "content": "Nov 3, 2024 · Meaning of the Word Intervention refers to the action or process of intervening in a situation, often to change or influence the outcome. It involves stepping into a situation with …"}
+{"idx": 9, "title": "Early Intervention | Learn the Signs. Act Early. | CDC", "date": "", "ddg_snippet": "Early Intervention... Is the term used to describe services and supports available to babies and young children with developmental delays and disabilities and their families. May include …", "subpage_snippet": "", "source": "www.cdc.gov", "link": "https://www.cdc.gov/act-early/early-intervention/index.html", "content": "Early Intervention... Is the term used to describe services and supports available to babies and young children with developmental delays and disabilities and their families. May include …"}
diff --git a/data/sampled_jsons/LAuReL_arxiv.org_LAUREL-LR_equation_implementation_formula.jsonl b/data/sampled_jsons/LAuReL_arxiv.org_LAUREL-LR_equation_implementation_formula.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f343eccf680d864c504e21597a224418db33d5a1
--- /dev/null
+++ b/data/sampled_jsons/LAuReL_arxiv.org_LAUREL-LR_equation_implementation_formula.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "arXiv . org e-Print archive", "date": "", "ddg_snippet": "Help Improve arXiv . arXiv is working with academic researchers to investigate ways to improve the service for our users. This requires a rich dataset to better inform the research. Users who wish to contribute to the future of arXiv may opt in to all...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/", "content": "Help Improve arXiv . arXiv is working with academic researchers to investigate ways to improve the service for our users. This requires a rich dataset to better inform the research. Users who wish to contribute to the future of arXiv may opt in to all..."}
+{"idx": 1, "title": "Markup for Rich Contents (The Org Manual)", "date": "", "ddg_snippet": "Or you may need to insert a formula or image in your writing. Org offers syntax for all of this and more. Used in conjunction with the export framework (see Exporting), you can author beautiful documents in Org —like the fine manual you are currently reading.", "subpage_snippet": "", "source": "orgmode.org", "link": "https://orgmode.org/manual/Markup-for-Rich-Contents.html", "content": "Or you may need to insert a formula or image in your writing. Org offers syntax for all of this and more. Used in conjunction with the export framework (see Exporting), you can author beautiful documents in Org —like the fine manual you are currently reading."}
+{"idx": 2, "title": "Winlator, Новый эмулятор для запуска Windows (ПК) игр для Android", "date": "", "ddg_snippet": "DXVK_ENABLE_D3D12=0: Disables the translation of DirectX 12 calls. Setting it to 0 can be useful for compatibility if a game has issues with DXVK's D3D12 implementation , forcing it to use D3D11 instead.", "subpage_snippet": "", "source": "4pda.to", "link": "https://4pda.to/forum/index.php?showtopic=1073499&st=27980", "content": "DXVK_ENABLE_D3D12=0: Disables the translation of DirectX 12 calls. Setting it to 0 can be useful for compatibility if a game has issues with DXVK's D3D12 implementation , forcing it to use D3D11 instead."}
+{"idx": 3, "title": "Aman's AI Journal • Papers List", "date": "", "ddg_snippet": "chrF Score Formula LAUREL -RW+ LR +PA: 75.25% (+2.40% params), outperforming naive scaling with fewer parameters. 1B LLM Pretraining (LLM-1)", "subpage_snippet": "", "source": "aman.ai", "link": "https://aman.ai/papers/", "content": "chrF Score Formula LAUREL -RW+ LR +PA: 75.25% (+2.40% params), outperforming naive scaling with fewer parameters. 1B LLM Pretraining (LLM-1)"}
+{"idx": 4, "title": "Анимированные курсоры для Windows (200+)", "date": "", "ddg_snippet": "Наборы с преобладанием анимированных (подвижных) указателей Windows в формате ani. Для тех, кто любит подолгу залипать на курсор.", "subpage_snippet": "", "source": "VSThemes.org", "link": "https://VSThemes.org/cursors/animated/", "content": "Наборы с преобладанием анимированных (подвижных) указателей Windows в формате ani. Для тех, кто любит подолгу залипать на курсор."}
+{"idx": 5, "title": "Ulysses | The Poetry Foundation", "date": "", "ddg_snippet": "We are not now that strength which in old daysOne equal temper of heroic hearts", "subpage_snippet": "", "source": "www.poetryfoundation.org", "link": "https://www.poetryfoundation.org/poems/45392/ulysses", "content": "We are not now that strength which in old daysOne equal temper of heroic hearts"}
+{"idx": 6, "title": "SimPlaza – MSFS 2020 Aircraft & More", "date": "", "ddg_snippet": "SimPlaza Recommended, iniBuilds, MSFS 2020 Aircraft, MSFS 2020 Airliner. PMDG – Boeing 777–200 LR v2.0.128.", "subpage_snippet": "", "source": "simplaza.org", "link": "https://simplaza.org/", "content": "SimPlaza Recommended, iniBuilds, MSFS 2020 Aircraft, MSFS 2020 Airliner. PMDG – Boeing 777–200 LR v2.0.128."}
+{"idx": 7, "title": "Angular Material UI Component Library", "date": "", "ddg_snippet": "UI component infrastructure and Material Design components for mobile and desktop Angular web applications.", "subpage_snippet": "", "source": "material.angular.dev", "link": "https://material.angular.dev/components/datepicker/overview", "content": "UI component infrastructure and Material Design components for mobile and desktop Angular web applications."}
+{"idx": 8, "title": "Ancient Grome - TV Tropes", "date": "", "ddg_snippet": "A sister trope to Mayincatec, Spexico, Far East and Scotireland, a tendency for writers to overlap the Greek and Roman civilizations and confuse aspects of the two Classical civilizations, e.g. Roman numerals in an otherwise Greek setting, Greek …", "subpage_snippet": "", "source": "tvtropes.org", "link": "https://tvtropes.org/pmwiki/pmwiki.php/Main/AncientGrome", "content": "A sister trope to Mayincatec, Spexico, Far East and Scotireland, a tendency for writers to overlap the Greek and Roman civilizations and confuse aspects of the two Classical civilizations, e.g. Roman numerals in an otherwise Greek setting, Greek …"}
+{"idx": 9, "title": "On the Theoretical Limitations of Embedding-Based Retrieval | alphaXiv", "date": "", "ddg_snippet": "View 1 comments: 严格正交 (Orthogonal)：在 n 维欧几里得空间 R^n 中，一个向量组如果两两严格正交（夹角精确为90°，或点积为0），那么这个向量组的大小（基数）最多为 n。这构成了该空间的一组正交基。近似正交 (Almost/Nearly Orthogonal)：如果我们...", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/pdf/2508.21038", "content": "View 1 comments: 严格正交 (Orthogonal)：在 n 维欧几里得空间 R^n 中，一个向量组如果两两严格正交（夹角精确为90°，或点积为0），那么这个向量组的大小（基数）最多为 n。这构成了该空间的一组正交基。近似正交 (Almost/Nearly Orthogonal)：如果我们..."}
diff --git a/data/sampled_jsons/LUNO_Linearization_Neural_Operators_Magnani_2022_Related_Work_section_FNO.jsonl b/data/sampled_jsons/LUNO_Linearization_Neural_Operators_Magnani_2022_Related_Work_section_FNO.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b121cbaf47f0570dbf90862fb13b39cf237cd77c
--- /dev/null
+++ b/data/sampled_jsons/LUNO_Linearization_Neural_Operators_Magnani_2022_Related_Work_section_FNO.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Linearization Turns Neural Operators into Function-Valued Gaussian ...", "date": "", "ddg_snippet": "More closely related to our work , Magnani et al. ( 2022 ) developed uncer-tainty estimates for graph neural operators using Laplace approximation, though their approach does not extend to FNOs , nor does it consider function space formulations.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=4Z04wVQ9FY", "content": "More closely related to our work , Magnani et al. ( 2022 ) developed uncer-tainty estimates for graph neural operators using Laplace approximation, though their approach does not extend to FNOs , nor does it consider function space formulations."}
+{"idx": 1, "title": "Linearization Turns Neural Operators into Function-Valued Gaussian ...", "date": "", "ddg_snippet": "Spotlight Poster Linearization Turns Neural Operators into Function-Valued Gaussian Processes Emilia Magnani · Marvin Pförtner · Tobias Weber · Philipp Hennig East Exhibition Hall A-B #E-1207", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46474", "content": "Spotlight Poster Linearization Turns Neural Operators into Function-Valued Gaussian Processes Emilia Magnani · Marvin Pförtner · Tobias Weber · Philipp Hennig East Exhibition Hall A-B #E-1207"}
+{"idx": 2, "title": "Linearization Turns Neural Operators into Function-Valued Gaussian ...", "date": "", "ddg_snippet": "We introduce LUNO , a novel framework for approximate Bayesian uncertainty quantification in trained neural operators . Our approach leverages model linearization to push (Gaussian) weight-space uncertainty forward to the neural operator's predictions.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2406.05072", "content": "We introduce LUNO , a novel framework for approximate Bayesian uncertainty quantification in trained neural operators . Our approach leverages model linearization to push (Gaussian) weight-space uncertainty forward to the neural operator's predictions."}
+{"idx": 3, "title": "LUNO: Linearized Predictive Uncertainty in Neural Operators", "date": "", "ddg_snippet": "luno - Linearized Uncertainty for Neural Operators This repository contains the main algorithm of the paper \" Linearization Turns Neural Operators into Function-Valued Gaussian Processes\" by Magnani et al. (2025).", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/MethodsOfMachineLearning/luno", "content": "luno - Linearized Uncertainty for Neural Operators This repository contains the main algorithm of the paper \" Linearization Turns Neural Operators into Function-Valued Gaussian Processes\" by Magnani et al. (2025)."}
+{"idx": 4, "title": "(PDF) Linearization Turns Neural Operators into Function-Valued ...", "date": "", "ddg_snippet": "Neural operators are deep neural networks designed to learn nontrivial solution operators of such differential equations from data.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/381294298_Linearization_Turns_Neural_Operators_into_Function-Valued_Gaussian_Processes", "content": "Neural operators are deep neural networks designed to learn nontrivial solution operators of such differential equations from data."}
+{"idx": 5, "title": "Linearization Turns Neural Operators into Function-Valued Gaussian ...", "date": "", "ddg_snippet": "In Section 4, we discuss prior work on operator learning and related uncertainty quantifica-tion. Finally, we demonstrate in Section 5 the effectiveness of LUNO in common PDE learning settings.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2406.05072v2", "content": "In Section 4, we discuss prior work on operator learning and related uncertainty quantifica-tion. Finally, we demonstrate in Section 5 the effectiveness of LUNO in common PDE learning settings."}
+{"idx": 6, "title": "PDF Abstract Linearization Turns Neural Operators into ... - ResearchGate", "date": "", "ddg_snippet": "Abstract Modeling dynamical systems, e.g. in climate and engineering sciences, often ne-cessitates solving partial differential equations. Neural operators are deep neural networks designed to ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/381294298_Linearization_Turns_Neural_Operators_into_Function-Valued_Gaussian_Processes/fulltext/6666938685a4ee7261b375f0/Linearization-Turns-Neural-Operators-into-Function-Valued-Gaussian-Processes.pdf", "content": "Abstract Modeling dynamical systems, e.g. in climate and engineering sciences, often ne-cessitates solving partial differential equations. Neural operators are deep neural networks designed to ..."}
+{"idx": 7, "title": "Experiments for the paper \"Linearization Turns Neural Operators into ...", "date": "", "ddg_snippet": "Experiments for the paper ` Linearization Turns Neural Operators Into Function-Valued Gaussian Processes` - 2bys/ luno -experiments", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/2bys/luno-experiments", "content": "Experiments for the paper ` Linearization Turns Neural Operators Into Function-Valued Gaussian Processes` - 2bys/ luno -experiments"}
+{"idx": 8, "title": "A comprehensive and fair comparison of two neural operators (with ...", "date": "", "ddg_snippet": "Neural operators can learn nonlinear mappings between function spaces and offer a new simulation paradigm for real-time prediction of complex dynamics for realistic diverse applications as well as for system identification in science and engineering.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0045782522001207", "content": "Neural operators can learn nonlinear mappings between function spaces and offer a new simulation paradigm for real-time prediction of complex dynamics for realistic diverse applications as well as for system identification in science and engineering."}
+{"idx": 9, "title": "deeponet-fno/README.md at main · lu-group/deeponet-fno", "date": "", "ddg_snippet": "A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data - lu-group/deeponet- fno", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/lu-group/deeponet-fno/blob/main/README.md?plain=1", "content": "A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data - lu-group/deeponet- fno"}
diff --git a/data/sampled_jsons/Langevin_unlearning_A_new_perspective_of_noisy_gradient_descent_for_machine_unlearning_Chien_et_al.__year_2024.jsonl b/data/sampled_jsons/Langevin_unlearning_A_new_perspective_of_noisy_gradient_descent_for_machine_unlearning_Chien_et_al.__year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d38aeb1752663aa15427ac3a079016c630b6ae5c
--- /dev/null
+++ b/data/sampled_jsons/Langevin_unlearning_A_new_perspective_of_noisy_gradient_descent_for_machine_unlearning_Chien_et_al.__year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Langevin Unlearning: A New Perspective of Noisy Gradient ... Langevin unlearning | Proceedings of the 38th International ... Graph-COM/Langevin_unlearning - GitHub dblp: Langevin Unlearning: A New Perspective of Noisy ... LANGEVIN UNLEARNING - OpenReview Certified Machine Unlearning via Noisy Stochastic Gradient ... The Challenge of Machine Unlearning in Data Privacy", "date": "", "ddg_snippet": "Jan 18, 2024 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. Jun 5, 2025 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . By Eli Chien , Haoyu Wang, Ziang Chen and Pan Li. Please also check our NeurIPS 2024 paper. Certified Machine Unlearning via Noisy Stochastic Gradient Descent. The code is runnable under the following enveironment: Feb 13, 2025 · dblp: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . For some months now, the dblp team has been receiving an exceptionally high number of support and error correction requests from the community. We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate un- learning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. 4 days ago · Title: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning Abstract : Machine unlearning has raised significant interest with the adoption of laws ensuring the ``right to be forgotten''.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2401.10371", "content": "Jan 18, 2024 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. Jun 5, 2025 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . By Eli Chien , Haoyu Wang, Ziang Chen and Pan Li. Please also check our NeurIPS 2024 paper. Certified Machine Unlearning via Noisy Stochastic Gradient Descent. The code is runnable under the following enveironment: Feb 13, 2025 · dblp: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . For some months now, the dblp team has been receiving an exceptionally high number of support and error correction requests from the community. We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate un- learning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits. 4 days ago · Title: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning Abstract : Machine unlearning has raised significant interest with the adoption of laws ensuring the ``right to be forgotten''."}
+{"idx": 1, "title": "Langevin unlearning | Proceedings of the 38th International ...", "date": "", "ddg_snippet": "Jun 5, 2025 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3737916.3740446", "content": "Jun 5, 2025 · We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits."}
+{"idx": 2, "title": "Graph-COM/Langevin_unlearning - GitHub", "date": "", "ddg_snippet": "Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . By Eli Chien , Haoyu Wang, Ziang Chen and Pan Li. Please also check our NeurIPS 2024 paper. Certified Machine Unlearning via Noisy Stochastic Gradient Descent. The code is runnable under the following enveironment:", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/Graph-COM/Langevin_unlearning", "content": "Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . By Eli Chien , Haoyu Wang, Ziang Chen and Pan Li. Please also check our NeurIPS 2024 paper. Certified Machine Unlearning via Noisy Stochastic Gradient Descent. The code is runnable under the following enveironment:"}
+{"idx": 3, "title": "dblp: Langevin Unlearning: A New Perspective of Noisy ...", "date": "", "ddg_snippet": "Feb 13, 2025 · dblp: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . For some months now, the dblp team has been receiving an exceptionally high number of support and error correction requests from the community.", "subpage_snippet": "", "source": "dblp.org", "link": "https://dblp.org/rec/conf/nips/ChienWCL24a", "content": "Feb 13, 2025 · dblp: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning . For some months now, the dblp team has been receiving an exceptionally high number of support and error correction requests from the community."}
+{"idx": 4, "title": "LANGEVIN UNLEARNING - OpenReview", "date": "", "ddg_snippet": "We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate un- learning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=VRrdtaaYiy", "content": "We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate un- learning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits."}
+{"idx": 5, "title": "Certified Machine Unlearning via Noisy Stochastic Gradient ...", "date": "", "ddg_snippet": "We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits.", "subpage_snippet": "", "source": "bohrium.dp.tech", "link": "https://bohrium.dp.tech/paper/arxiv/2403.17105", "content": "We propose Langevin unlearning , an unlearning framework based on noisy gradient descent with privacy guarantees for approximate unlearning problems. Langevin unlearning unifies the DP learning process and the privacy-certified unlearning process with many algorithmic benefits."}
+{"idx": 6, "title": "The Challenge of Machine Unlearning in Data Privacy", "date": "", "ddg_snippet": "4 days ago · Title: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning Abstract : Machine unlearning has raised significant interest with the adoption of laws ensuring the ``right to be forgotten''.", "subpage_snippet": "", "source": "scisimple.com", "link": "https://scisimple.com/en/articles/2025-09-16-the-challenge-of-machine-unlearning-in-data-privacy--ak2m4j1", "content": "4 days ago · Title: Langevin Unlearning : A New Perspective of Noisy Gradient Descent for Machine Unlearning Abstract : Machine unlearning has raised significant interest with the adoption of laws ensuring the ``right to be forgotten''."}
+{"idx": 7, "title": "Certified Unlearning for Neural Networks", "date": "", "ddg_snippet": "Recent works aim to achieve certified unlearning for non-convex tasks (Golatkar et al ., 2020 ; Chourasia & Shah, 2023 ; Chien et al ., 2024 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.06985v2", "content": "Recent works aim to achieve certified unlearning for non-convex tasks (Golatkar et al ., 2020 ; Chourasia & Shah, 2023 ; Chien et al ., 2024 ..."}
+{"idx": 8, "title": "Towards Reliable Forgetting: A Survey on Machine Unlearning", "date": "", "ddg_snippet": "... for privacy protection, security, and legal compliance ( e .g., GDPR), machine unlearning has emerged as a critical technique for ensuring the ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.15115v1", "content": "... for privacy protection, security, and legal compliance ( e .g., GDPR), machine unlearning has emerged as a critical technique for ensuring the ..."}
+{"idx": 9, "title": "Distributional Unlearning: Forgetting Distributions, Not Just", "date": "", "ddg_snippet": "While these lines of work tackle important aspects of unlearning or robustness, they do not provide a unified, distribution‑level forgetting ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.15112v1", "content": "While these lines of work tackle important aspects of unlearning or robustness, they do not provide a unified, distribution‑level forgetting ..."}
diff --git a/data/sampled_jsons/Learning_stochastic_dynamics_snapshots_regularized_unbalanced_optimal_transport_Theorem_4.1_Fisher_i.jsonl b/data/sampled_jsons/Learning_stochastic_dynamics_snapshots_regularized_unbalanced_optimal_transport_Theorem_4.1_Fisher_i.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9622ee1dd0a54b32582a10ad896d3bac850f8d09
--- /dev/null
+++ b/data/sampled_jsons/Learning_stochastic_dynamics_snapshots_regularized_unbalanced_optimal_transport_Theorem_4.1_Fisher_i.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Learning stochastic dynamics from snapshots through regularized ...", "date": "", "ddg_snippet": "regularized unbalanced optimal transport ; Schrödinger bridge; trajectory inference; deep learning .In this section, we first introduce the RUOT problem, then derive the Fisher regularization form of RUOT, and discuss its connections with existing literature.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.00844v1", "content": "regularized unbalanced optimal transport ; Schrödinger bridge; trajectory inference; deep learning .In this section, we first introduce the RUOT problem, then derive the Fisher regularization form of RUOT, and discuss its connections with existing literature."}
+{"idx": 1, "title": "(PDF) Learning Stochastic Dynamics from Snapshots through...", "date": "", "ddg_snippet": "regularized unbalanced optimal transport (RUOT) and continuous unbalanced stochas-. tic dynamics from time-series snapshot data. By leveraging Fisher regularization , our. method transforms a stochastic differential equation (SDE) problem into an ordinary.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/384563451_Learning_Stochastic_Dynamics_from_Snapshots_through_Regularized_Unbalanced_Optimal_Transport", "content": "regularized unbalanced optimal transport (RUOT) and continuous unbalanced stochas-. tic dynamics from time-series snapshot data. By leveraging Fisher regularization , our. method transforms a stochastic differential equation (SDE) problem into an ordinary."}
+{"idx": 2, "title": "Learning stochastic dynamics from snapshots through regularized ...", "date": "", "ddg_snippet": "Reconstructing dynamics using samples from sparsely time-resolved snapshots is an important problem in both natural sciences and machine learning . Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT)...", "subpage_snippet": "", "source": "synthical.com", "link": "https://synthical.com/article/Learning-Stochastic-Dynamics-from-Snapshots-through-Regularized-Unbalanced-Optimal-Transport-39225f33-522f-44ce-b836-70679229a4ab", "content": "Reconstructing dynamics using samples from sparsely time-resolved snapshots is an important problem in both natural sciences and machine learning . Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT)..."}
+{"idx": 3, "title": "Learning stochastic dynamics from snapshots through regularized ...", "date": "", "ddg_snippet": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=gQlxd3Mtru", "content": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots ."}
+{"idx": 4, "title": "Learning Stochastic Dynamics from Snapshots through...", "date": "", "ddg_snippet": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots .", "subpage_snippet": "", "source": "papers.cool", "link": "https://papers.cool/arxiv/2410.00844", "content": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots ."}
+{"idx": 5, "title": "Learning Stochastic Dynamics from Snapshots through...", "date": "", "ddg_snippet": "This paper presents a method for learning stochastic dynamics from snapshot data using regularized unbalanced optimal transport .", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/learning-stochastic-dynamics-from-snapshots-through-regularized", "content": "This paper presents a method for learning stochastic dynamics from snapshot data using regularized unbalanced optimal transport ."}
+{"idx": 6, "title": "GitHub - zhenyiizhang/DeepRUOT: Learning stochastic dynamics ...", "date": "", "ddg_snippet": "Learning stochastic dynamics from snapshots through regularized unbalanced optimal transport . The Thirteenth International Conference on Learning Representations.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/zhenyiizhang/DeepRUOT", "content": "Learning stochastic dynamics from snapshots through regularized unbalanced optimal transport . The Thirteenth International Conference on Learning Representations."}
+{"idx": 7, "title": "ICLR 2025 Learning stochastic dynamics from snapshots through...", "date": "", "ddg_snippet": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots .", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2025/oral/31800", "content": "Here, we introduce a new deep learning approach for solving regularized unbalanced optimal transport (RUOT) and inferring continuous unbalanced stochastic dynamics from observed snapshots ."}
+{"idx": 8, "title": "On the Relation Between Optimal Transport and Schrödinger... | CoLab", "date": "", "ddg_snippet": "We take a new look at the relation between the optimal transport problem and the Schrödinger bridge problem from a stochastic control perspective.", "subpage_snippet": "", "source": "colab.ws", "link": "https://colab.ws/articles/10.1007/s10957-015-0803-z", "content": "We take a new look at the relation between the optimal transport problem and the Schrödinger bridge problem from a stochastic control perspective."}
+{"idx": 9, "title": "Manifold Interpolating Optimal - Transport Flows... | Connected Papers", "date": "", "ddg_snippet": "TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics .Action Matching: Learning Stochastic Dynamics from Samples.", "subpage_snippet": "", "source": "www.connectedpapers.com", "link": "https://www.connectedpapers.com/main/b18d4ee8e80c5669e046ad76ff15da3b2f6d835e/Manifold-Interpolating-Optimal+Transport-Flows-for-Trajectory-Inference/graph", "content": "TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics .Action Matching: Learning Stochastic Dynamics from Samples."}
diff --git a/data/sampled_jsons/Likelihood_Based_Approach_to_Distribution_Regression_Using_Conditional_Deep_Generative_Models_PDF_ar.jsonl b/data/sampled_jsons/Likelihood_Based_Approach_to_Distribution_Regression_Using_Conditional_Deep_Generative_Models_PDF_ar.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..70a6e3542880c216a396fb3b20a187099ca0d3b5
--- /dev/null
+++ b/data/sampled_jsons/Likelihood_Based_Approach_to_Distribution_Regression_Using_Conditional_Deep_Generative_Models_PDF_ar.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Graphical model - Wikipedia", "date": "", "ddg_snippet": "Generally, probabilistic graphical models use a graph- based representation as the foundation for encoding a distribution over a multi-dimensional ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Graphical_model", "content": "Generally, probabilistic graphical models use a graph- based representation as the foundation for encoding a distribution over a multi-dimensional ..."}
+{"idx": 1, "title": "Generative model - Wikipedia", "date": "", "ddg_snippet": "... generative model \" is also used to describe models that generate instances of output variables in a way that has no clear relationship to probability ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Generative_model", "content": "... generative model \" is also used to describe models that generate instances of output variables in a way that has no clear relationship to probability ..."}
+{"idx": 2, "title": "Flow-based generative model - Wikipedia", "date": "", "ddg_snippet": "... based generative model is a generative model used in machine learning that explicitly models a probability distribution by leveraging normalizing flow ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Flow-based_generative_model", "content": "... based generative model is a generative model used in machine learning that explicitly models a probability distribution by leveraging normalizing flow ..."}
+{"idx": 3, "title": "Generative adversarial network - Wikipedia", "date": "", "ddg_snippet": "The generator's task is to approach G ref {\\displaystyle \\mu _{G}\\approx \\mu _{\\text{ref}}} , that is, to match its own output distribution as ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Generative_adversarial_network", "content": "The generator's task is to approach G ref {\\displaystyle \\mu _{G}\\approx \\mu _{\\text{ref}}} , that is, to match its own output distribution as ..."}
+{"idx": 4, "title": "A Likelihood Based Approach to Distribution Regression Using ...", "date": "", "ddg_snippet": "Oct 2, 2024 · In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.02025", "content": "Oct 2, 2024 · In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ..."}
+{"idx": 5, "title": "A Likelihood Based Approach to Distribution Regression Using ...", "date": "", "ddg_snippet": "Conclusion This paper presents an innovative likelihood - based approach for distribution regression using conditional deep generative models . The key contributions are the principled formulation, efficient inference, and demonstrated improvements over existing methods in terms of predictive performance and uncertainty quantification.", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/likelihood-based-approach-to-distribution-regression-using", "content": "Conclusion This paper presents an innovative likelihood - based approach for distribution regression using conditional deep generative models . The key contributions are the principled formulation, efficient inference, and demonstrated improvements over existing methods in terms of predictive performance and uncertainty quantification."}
+{"idx": 6, "title": "A LIKELIHOOD BASED APPROACH TO DISTRIBUTION REGRESSION USING ...", "date": "", "ddg_snippet": "In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the re-sponse variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=V6hhhXoTSq", "content": "In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the re-sponse variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ..."}
+{"idx": 7, "title": "ICML Poster A Likelihood Based Approach to Distribution ...", "date": "", "ddg_snippet": "In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46645", "content": "In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating these ..."}
+{"idx": 8, "title": "A Likelihood Approach to Nonparametric Estimation of a ... Conditional Diffusion Models are Minimax-Optimal and Manifold ... A Likelihood Based Approach to Distribution Regression Using ...", "date": "", "ddg_snippet": "We investigate statistical properties of a likelihood approach to nonparametric estimation of a singular distribution using deep generative models . More speci cally, a deep generative model is used to model high-dimensional data that are assumed to concentrate around some low-dimensional structure. Estimating the distribution supported on this low-... See full list on jmlr.org is the Hellinger convergence rate of the sieve MLE of p , and See full list on jmlr.org decreases as increases because becomes smoother while See full list on jmlr.org Our main theoretical results are given in this section. We rst present assumptions on the data-generating distribution P . Then, we derive the convergence rate of a sieve MLE for p with respect to the Hellinger distance in the deep generative model. We next obtain the convergence rate of the corresponding sieve MLE of Q under the Wasserstein distan... See full list on jmlr.org n where C = C(q; d; t; ; K; D; max; ; ). As one can see, the dimension d in the convergence rate of Corollary 4 is replaced by the intrinsic dimension t . If t is much smaller than d, the improvement from the structural assumption would be signi cant. See full list on jmlr.org In this section, we empirically demonstrate that the data perturbation method proposed in Section 3.4 plays an important role to improve the performance of a sieve MLE of deep generative models . In addition, we illustrate that deep generative models can detect low-dimensional structures well. Numerical studies are carried out by analyzing various s... See full list on jmlr.org Assume that the generator f = f is parametrized by . With a slight abuse of notation, let ; = pf ; , that is, Z ; (x) = See full list on jmlr.org The model is trained after perturbing the training data by an arti cial noise e For each data set, we consider various values of . e See full list on jmlr.org Abstract:In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating ... More specifically, we study the large-sample properties of a likelihood - based approach for estimating these models . Our results lead to the convergence rate of a sieve maximum likelihood estimator (MLE) for estimating the conditional distribution (and its devolved counterpart) of the response given predictors in the Hellinger (Wasserstein) metric.", "subpage_snippet": "", "source": "jmlr.org", "link": "https://jmlr.org/papers/volume24/21-1099/21-1099.pdf", "content": "We investigate statistical properties of a likelihood approach to nonparametric estimation of a singular distribution using deep generative models . More speci cally, a deep generative model is used to model high-dimensional data that are assumed to concentrate around some low-dimensional structure. Estimating the distribution supported on this low-... See full list on jmlr.org is the Hellinger convergence rate of the sieve MLE of p , and See full list on jmlr.org decreases as increases because becomes smoother while See full list on jmlr.org Our main theoretical results are given in this section. We rst present assumptions on the data-generating distribution P . Then, we derive the convergence rate of a sieve MLE for p with respect to the Hellinger distance in the deep generative model. We next obtain the convergence rate of the corresponding sieve MLE of Q under the Wasserstein distan... See full list on jmlr.org n where C = C(q; d; t; ; K; D; max; ; ). As one can see, the dimension d in the convergence rate of Corollary 4 is replaced by the intrinsic dimension t . If t is much smaller than d, the improvement from the structural assumption would be signi cant. See full list on jmlr.org In this section, we empirically demonstrate that the data perturbation method proposed in Section 3.4 plays an important role to improve the performance of a sieve MLE of deep generative models . In addition, we illustrate that deep generative models can detect low-dimensional structures well. Numerical studies are carried out by analyzing various s... See full list on jmlr.org Assume that the generator f = f is parametrized by . With a slight abuse of notation, let ; = pf ; , that is, Z ; (x) = See full list on jmlr.org The model is trained after perturbing the training data by an arti cial noise e For each data set, we consider various values of . e See full list on jmlr.org Abstract:In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating ... More specifically, we study the large-sample properties of a likelihood - based approach for estimating these models . Our results lead to the convergence rate of a sieve maximum likelihood estimator (MLE) for estimating the conditional distribution (and its devolved counterpart) of the response given predictors in the Hellinger (Wasserstein) metric."}
+{"idx": 9, "title": "Conditional Diffusion Models are Minimax-Optimal and Manifold ...", "date": "", "ddg_snippet": "Abstract:In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating ...", "subpage_snippet": "", "source": "bohrium.dp.tech", "link": "https://bohrium.dp.tech/paper/arxiv/2409.20124", "content": "Abstract:In this work, we explore the theoretical properties of conditional deep generative models under the statistical framework of distribution regression where the response variable lies in a high-dimensional ambient space but concentrates around a potentially lower-dimensional manifold. More specifically, we study the large-sample properties of a likelihood - based approach for estimating ..."}
diff --git "a/data/sampled_jsons/Linear_convergence_Sinkhorn's_algorithm_generalized_static_Schr\303\266dinger_bridge_Theorem_2.7_tameness_c.jsonl" "b/data/sampled_jsons/Linear_convergence_Sinkhorn's_algorithm_generalized_static_Schr\303\266dinger_bridge_Theorem_2.7_tameness_c.jsonl"
new file mode 100644
index 0000000000000000000000000000000000000000..78a5881cf7be47769a2af84566f9a8f72b04cf96
--- /dev/null
+++ "b/data/sampled_jsons/Linear_convergence_Sinkhorn's_algorithm_generalized_static_Schr\303\266dinger_bridge_Theorem_2.7_tameness_c.jsonl"
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Linear convergence of Sinkhorn's algorithm for generalized ...", "date": "", "ddg_snippet": "1 May 2025 — This paper studies the generalized static Schrödinger bridge (SSB) problem, which extends the classical SSB by replacing the entropy divergence ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0hrkN07DuO&noteId=O4YQkrxppf", "content": "1 May 2025 — This paper studies the generalized static Schrödinger bridge (SSB) problem, which extends the classical SSB by replacing the entropy divergence ..."}
+{"idx": 1, "title": "Linear convergence of Sinkhorn's algorithm for generalized ...", "date": "", "ddg_snippet": "16 Jul 2025 — We establish Kantorovich duality and linear convergence of Sinkhorn's algorithm for the generalized SSB problem under mild conditions . Our ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46671", "content": "16 Jul 2025 — We establish Kantorovich duality and linear convergence of Sinkhorn's algorithm for the generalized SSB problem under mild conditions . Our ..."}
+{"idx": 2, "title": "Linear Convergence of Sinkhorn's Algorithm for Generalized ...", "date": "", "ddg_snippet": "Linear Convergence of Sinkhorn's Algorithm for. Generalized Static Schrödinger Bridge ... 21 ... condition for tameness for similar but more general cases.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/a644c39247d5073d97fae0bbd2c5b921800516ea.pdf", "content": "Linear Convergence of Sinkhorn's Algorithm for. Generalized Static Schrödinger Bridge ... 21 ... condition for tameness for similar but more general cases."}
+{"idx": 3, "title": "large random matrices with given margins", "date": "", "ddg_snippet": "ABSTRACT. We study large random matrices with i.i.d. entries conditioned to have prescribed row and column sums (margins), a problem connected to relative ...", "subpage_snippet": "", "source": "hanbaeklyu.com", "link": "https://hanbaeklyu.com/wp-content/uploads/2025/08/ct_limit.pdf", "content": "ABSTRACT. We study large random matrices with i.i.d. entries conditioned to have prescribed row and column sums (margins), a problem connected to relative ..."}
+{"idx": 4, "title": "Optimization Methods for Large-Scale Machine Learning", "date": "", "ddg_snippet": "This paper provides a review and commentary on the past, present, and future of numerical optimization algorithms in the context of machine learning ...", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/10.1137/16M1080173", "content": "This paper provides a review and commentary on the past, present, and future of numerical optimization algorithms in the context of machine learning ..."}
+{"idx": 5, "title": "Review of latest Score Based Generative Modeling papers.", "date": "", "ddg_snippet": "On the Contraction Coefficient of the Schrödinger Bridge for Stochastic Linear Systems. September 12, 2023 Alexis M. H. Teter, Yongxin Chen, Abhishek Halder.", "subpage_snippet": "", "source": "scorebasedgenerativemodeling.github.io", "link": "https://scorebasedgenerativemodeling.github.io/", "content": "On the Contraction Coefficient of the Schrödinger Bridge for Stochastic Linear Systems. September 12, 2023 Alexis M. H. Teter, Yongxin Chen, Abhishek Halder."}
+{"idx": 6, "title": "Mathematics Oct 2023", "date": "", "ddg_snippet": "Title: On Linear Convergence of PI Consensus Algorithm under the Restricted Secant Inequality. Kushal Chakrabarti, Mayank Baranwal. Comments: Accepted for ...", "subpage_snippet": "", "source": "web3.arxiv.org", "link": "https://web3.arxiv.org/list/math/2023-10?skip=25&show=2000", "content": "Title: On Linear Convergence of PI Consensus Algorithm under the Restricted Secant Inequality. Kushal Chakrabarti, Mayank Baranwal. Comments: Accepted for ..."}
+{"idx": 7, "title": "Optimal Transport and Deep Learning - Kilian Fatras", "date": "", "ddg_snippet": "25 Nov 2021 — Ce manuscrit contient tous les travaux de ma thèse qui ont été réalisés de Novembre 2018 à Février 2021.", "subpage_snippet": "", "source": "kilianfatras.github.io", "link": "https://kilianfatras.github.io/pdf/thesis_kf.pdf", "content": "25 Nov 2021 — Ce manuscrit contient tous les travaux de ma thèse qui ont été réalisés de Novembre 2018 à Février 2021."}
+{"idx": 8, "title": "Mathematics Sep 2023", "date": "", "ddg_snippet": "Comments: In this revised version, all typographical errors have been meticulously corrected, enhancing clarity and precision.", "subpage_snippet": "", "source": "arxiv.org", "link": "http://arxiv.org/list/math/2023-09?skip=1050&show=2000", "content": "Comments: In this revised version, all typographical errors have been meticulously corrected, enhancing clarity and precision."}
+{"idx": 9, "title": "Track: Poster Session 2", "date": "", "ddg_snippet": "24 Apr 2025 — This paper addresses these challenges by introducing a novel data augmentation strategy that separately augments observed action sequences and next actions.", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2025/session/31972", "content": "24 Apr 2025 — This paper addresses these challenges by introducing a novel data augmentation strategy that separately augments observed action sequences and next actions."}
diff --git a/data/sampled_jsons/Longpre_2021_EMNLP_entity-based_knowledge_conflicts_NQ-Swap_abstract.jsonl b/data/sampled_jsons/Longpre_2021_EMNLP_entity-based_knowledge_conflicts_NQ-Swap_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..677ebd26a3fb715b2f32ae75f8c57170cfabedb2
--- /dev/null
+++ b/data/sampled_jsons/Longpre_2021_EMNLP_entity-based_knowledge_conflicts_NQ-Swap_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Entity-Based Knowledge Conflicts in Question Answering", "date": "", "ddg_snippet": "by S Longpre · 2021 · Cited by 290 — Entity - based Substitutions Key to our evalua- tion framework is substituting entity names with other plausible entity names. Entity based swap -.", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2021.emnlp-main.565.pdf", "content": "by S Longpre · 2021 · Cited by 290 — Entity - based Substitutions Key to our evalua- tion framework is substituting entity names with other plausible entity names. Entity based swap -."}
+{"idx": 1, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "6 Jun 2025 — (2021) Longpre, S., Perisetla, K., Chen, A., Ramesh, N., DuBois, C., and Singh, S. Entity-based knowledge conflicts in question answering .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.10996v2", "content": "6 Jun 2025 — (2021) Longpre, S., Perisetla, K., Chen, A., Ramesh, N., DuBois, C., and Singh, S. Entity-based knowledge conflicts in question answering ."}
+{"idx": 2, "title": "Interweaving Conflicting Knowledge and Reasoning Skills ...", "date": "", "ddg_snippet": "For example, NQ -Swaq requires LLMs to extract conflicting entities from the provided document without solely referring to their parametric knowledge directly ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2404.03577v1", "content": "For example, NQ -Swaq requires LLMs to extract conflicting entities from the provided document without solely referring to their parametric knowledge directly ..."}
+{"idx": 3, "title": "ADACAD: Adaptively Decoding to Balance Conflicts ...", "date": "", "ddg_snippet": "by H Wang · 2025 · Cited by 16 — NQ-SWAP (Longpre et al., 2021 ) introduces synthetic conflicts by swapping entities in the context to challenge the model's ability to manage ...", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.naacl-long.581.pdf", "content": "by H Wang · 2025 · Cited by 16 — NQ-SWAP (Longpre et al., 2021 ) introduces synthetic conflicts by swapping entities in the context to challenge the model's ability to manage ..."}
+{"idx": 4, "title": "Exploiting Contextual Knowledge in LLMs through V- ...", "date": "", "ddg_snippet": "Nikhil Ramesh, Chris DuBois, and Sameer Singh. 718. 2021 . Entity - based knowledge conflicts in question ... NQ - Swap NQ - Swap is based on the NQ dataset,. 1072.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=B3i76OuS4d", "content": "Nikhil Ramesh, Chris DuBois, and Sameer Singh. 718. 2021 . Entity - based knowledge conflicts in question ... NQ - Swap NQ - Swap is based on the NQ dataset,. 1072."}
+{"idx": 5, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "Abstract . Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46677", "content": "Abstract . Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge ..."}
+{"idx": 6, "title": "Paper - EMNLP 2023", "date": "", "ddg_snippet": "6 Dec 2023 — I am happy to welcome you to EMNLP -2023 in Singapore! Like EMNLP - 2021 , EMNLP -2022, and other. ACL-related meetings, we decided to host ...", "subpage_snippet": "", "source": "2023.emnlp.org", "link": "https://2023.emnlp.org/downloads/EMNLP-2023-Handbook-Nov-30.pdf", "content": "6 Dec 2023 — I am happy to welcome you to EMNLP -2023 in Singapore! Like EMNLP - 2021 , EMNLP -2022, and other. ACL-related meetings, we decided to host ..."}
+{"idx": 7, "title": "Analysing the Residual Stream of Language Models Under ...", "date": "", "ddg_snippet": "We present the knowledge conflict probing results on Macnoise, NQSwap , ConflictQA using Llama2-. 7B in Figure 4, Figure 5 and Figure 6. The results match the ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/33fc952b58cab08008bc207991f5af7d0d3123a8.pdf", "content": "We present the knowledge conflict probing results on Macnoise, NQSwap , ConflictQA using Llama2-. 7B in Figure 4, Figure 5 and Figure 6. The results match the ..."}
+{"idx": 8, "title": "Fewer Truncations Improve Language Modeling", "date": "", "ddg_snippet": "by H Ding · 2024 · Cited by 33 — Specifically, we evaluate with 5-shot on NQ - Swap ( Longpre et al., 2021 ) ... Entity - based knowledge conflicts in question answering. In Proceedings of ...", "subpage_snippet": "", "source": "assets.amazon.science", "link": "https://assets.amazon.science/be/1e/215dde844d49a13136e70d683d8b/fewer-truncations-improve-language-modeling.pdf", "content": "by H Ding · 2024 · Cited by 33 — Specifically, we evaluate with 5-shot on NQ - Swap ( Longpre et al., 2021 ) ... Entity - based knowledge conflicts in question answering. In Proceedings of ..."}
+{"idx": 9, "title": "Factuality and Large Language Models", "date": "", "ddg_snippet": "by A Chen · 2023 · Cited by 5 — The text in this chapter is based on the publications: • Entity - Based Knowledge Conflicts in Question Answering [97] ( EMNLP 2021 ). The results presented in ...", "subpage_snippet": "", "source": "escholarship.org", "link": "https://escholarship.org/content/qt7542q1d9/qt7542q1d9_noSplash_f1a1efa5375a76a01f6dde7fe8322d5b.pdf", "content": "by A Chen · 2023 · Cited by 5 — The text in this chapter is based on the publications: • Entity - Based Knowledge Conflicts in Question Answering [97] ( EMNLP 2021 ). The results presented in ..."}
diff --git a/data/sampled_jsons/MSVD_frames_sampled_Video-ColBERT_sitearxiv.org_OR_sitethecvf.com.jsonl b/data/sampled_jsons/MSVD_frames_sampled_Video-ColBERT_sitearxiv.org_OR_sitethecvf.com.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..50747575827aff5ec4e31a51a63c4a41050c2420
--- /dev/null
+++ b/data/sampled_jsons/MSVD_frames_sampled_Video-ColBERT_sitearxiv.org_OR_sitethecvf.com.jsonl
@@ -0,0 +1,2 @@
+{"idx": 0, "title": "Contextualized Late Interaction for Text-to-Video Retrieval", "date": "", "ddg_snippet": "24 Mar 2025 — ... MSVD and VATEX, while using 64 frames for DiDeMo and ActivityNet. ... Number of Sampled Frames . Report issue for preceding element. # of Frames ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.19009v1", "content": "24 Mar 2025 — ... MSVD and VATEX, while using 64 frames for DiDeMo and ActivityNet. ... Number of Sampled Frames . Report issue for preceding element. # of Frames ..."}
+{"idx": 1, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}
diff --git a/data/sampled_jsons/Machine_Learning_meets_Algebraic_Combinatorics_Schubert_polynomials_spurious_correlation_length.jsonl b/data/sampled_jsons/Machine_Learning_meets_Algebraic_Combinatorics_Schubert_polynomials_spurious_correlation_length.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d3879e2e59f22f88b586a82fa4c8e8714e82c1a1
--- /dev/null
+++ b/data/sampled_jsons/Machine_Learning_meets_Algebraic_Combinatorics_Schubert_polynomials_spurious_correlation_length.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Machine - Wikipedia", "date": "", "ddg_snippet": "A machine is a thermodynamic system that uses power to apply forces and control movement to perform an action. The term is commonly applied to artificial devices, such as those employing engines or motors, but also to natural biological macromolecules, such as molecular machines .", "subpage_snippet": "", "source": "en.m.wikipedia.org", "link": "https://en.m.wikipedia.org/wiki/Machine", "content": "A machine is a thermodynamic system that uses power to apply forces and control movement to perform an action. The term is commonly applied to artificial devices, such as those employing engines or motors, but also to natural biological macromolecules, such as molecular machines ."}
+{"idx": 1, "title": "MACHINE Definition & Meaning | Dictionary .com", "date": "", "ddg_snippet": "Machine definition: an apparatus consisting of interrelated parts with separate functions, used in the performance of some kind of work.. See examples of MACHINE used in a sentence.", "subpage_snippet": "", "source": "www.dictionary.com", "link": "https://www.dictionary.com/browse/machine", "content": "Machine definition: an apparatus consisting of interrelated parts with separate functions, used in the performance of some kind of work.. See examples of MACHINE used in a sentence."}
+{"idx": 2, "title": "MACHINE Definition & Meaning - Merriam-Webster", "date": "", "ddg_snippet": "The meaning of MACHINE is a mechanically, electrically, or electronically operated device for performing a task. How to use machine in a sentence.", "subpage_snippet": "", "source": "www.merriam-webster.com", "link": "https://www.merriam-webster.com/dictionary/machine", "content": "The meaning of MACHINE is a mechanically, electrically, or electronically operated device for performing a task. How to use machine in a sentence."}
+{"idx": 3, "title": "Machine | Definition, Mechanisms & Efficiency | Britannica", "date": "", "ddg_snippet": "machine , device, having a unique purpose, that augments or replaces human or animal effort for the accomplishment of physical tasks.", "subpage_snippet": "", "source": "www.britannica.com", "link": "https://www.britannica.com/technology/machine", "content": "machine , device, having a unique purpose, that augments or replaces human or animal effort for the accomplishment of physical tasks."}
+{"idx": 4, "title": "MACHINE | English meaning - Cambridge Dictionary", "date": "", "ddg_snippet": "MACHINE definition: 1. a piece of equipment with several moving parts that uses power to do a particular type of work…. Learn more.", "subpage_snippet": "", "source": "dictionary.cambridge.org", "link": "https://dictionary.cambridge.org/dictionary/english/machine", "content": "MACHINE definition: 1. a piece of equipment with several moving parts that uses power to do a particular type of work…. Learn more."}
+{"idx": 5, "title": "Machine - definition of machine by The Free Dictionary", "date": "", "ddg_snippet": "Of, relating to, or felt to resemble a machine : machine repairs; machine politics.", "subpage_snippet": "", "source": "www.thefreedictionary.com", "link": "https://www.thefreedictionary.com/machine", "content": "Of, relating to, or felt to resemble a machine : machine repairs; machine politics."}
+{"idx": 6, "title": "What Is A Machine ? Its Types and How it Works - Mech Lesson", "date": "", "ddg_snippet": "A machine is a mechanical device that uses power to apply force and control motion to perform work efficiently . Machines range from simple tools like pulleys and levers to complex systems like engines, robots, and manufacturing equipment.", "subpage_snippet": "", "source": "mechlesson.com", "link": "https://mechlesson.com/machine/", "content": "A machine is a mechanical device that uses power to apply force and control motion to perform work efficiently . Machines range from simple tools like pulleys and levers to complex systems like engines, robots, and manufacturing equipment."}
+{"idx": 7, "title": "What is a Machine ? - Computer Hope", "date": "", "ddg_snippet": "Jun 1, 2025 · A machine is a device with several parts that work together to perform a task. A machine can augment or replace the efforts required by an animal or human to make it easier to complete a task. For example, a car is a complex machine that transports humans and other things.", "subpage_snippet": "", "source": "www.computerhope.com", "link": "https://www.computerhope.com/jargon/m/machine.htm", "content": "Jun 1, 2025 · A machine is a device with several parts that work together to perform a task. A machine can augment or replace the efforts required by an animal or human to make it easier to complete a task. For example, a car is a complex machine that transports humans and other things."}
+{"idx": 8, "title": "Machine - New World Encyclopedia", "date": "", "ddg_snippet": "Modern power tools, automated machine tools, and human-operated power machinery are tools that are also machines . Machines used to transform heat or other energy into mechanical energy are known as engines.", "subpage_snippet": "", "source": "www.newworldencyclopedia.org", "link": "https://www.newworldencyclopedia.org/entry/Machine", "content": "Modern power tools, automated machine tools, and human-operated power machinery are tools that are also machines . Machines used to transform heat or other energy into mechanical energy are known as engines."}
+{"idx": 9, "title": "What is Machine ? – An Essential Guide", "date": "", "ddg_snippet": "The machine is a device that has a unique purpose, that augments or replaces human or animal effort for the accomplishment of physical tasks.", "subpage_snippet": "", "source": "www.theengineeringchoice.com", "link": "https://www.theengineeringchoice.com/what-is-machine/", "content": "The machine is a device that has a unique purpose, that augments or replaces human or animal effort for the accomplishment of physical tasks."}
diff --git a/data/sampled_jsons/MultiPDENet_MaNN_Block_OpenReview.jsonl b/data/sampled_jsons/MultiPDENet_MaNN_Block_OpenReview.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2ae39f0d5bf0d9eb1a191a09a81207a8e990da2e
--- /dev/null
+++ b/data/sampled_jsons/MultiPDENet_MaNN_Block_OpenReview.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "PDE-constrained Learning with Multi-time-stepping for ... - OpenReview", "date": "", "ddg_snippet": "This paper introduces MultiPDENet , a neural network architecture designed to accelerate fluid dynamic simulations by combining classical numerical methods, such as a multi-scale time-stepping scheme inspired by Runge-Kutta methods, finite-difference derivatives, and a Fourier Neural Operator for learned corrections. This approach embeds physical constraints from partial differential equations ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=stcN89QGfL", "content": "This paper introduces MultiPDENet , a neural network architecture designed to accelerate fluid dynamic simulations by combining classical numerical methods, such as a multi-scale time-stepping scheme inspired by Runge-Kutta methods, finite-difference derivatives, and a Fourier Neural Operator for learned corrections. This approach embeds physical constraints from partial differential equations ..."}
+{"idx": 1, "title": "MultiPDENet: PDE-embedded Learning with Multi-time-stepping for ...", "date": "", "ddg_snippet": "MultiPDENet achieves the state-of-the-art performance compared with other neural baseline models, also with clear speedup compared to classical numerical methods. Schematic of MultiPDENet for learning turbulent flows. (a), Model architecture. (b), Physics Block . (c), Learnable PDE block .", "subpage_snippet": "", "source": "chatpaper.com", "link": "https://chatpaper.com/chatpaper/paper/102640?from=search", "content": "MultiPDENet achieves the state-of-the-art performance compared with other neural baseline models, also with clear speedup compared to classical numerical methods. Schematic of MultiPDENet for learning turbulent flows. (a), Model architecture. (b), Physics Block . (c), Learnable PDE block ."}
+{"idx": 2, "title": "MultiPDENet: PDE-embedded Learning with Multi-time-stepping for ...", "date": "", "ddg_snippet": "A PDE-embedded network with multiscale time stepping ( MultiPDENet ), which fuses the scheme of numerical methods and machine learning, for accelerated simulation of flows and achieves the state-of-the-art performance compared with other neural baseline models, also with clear speedup compared to classical numerical methods. Solving partial differential equations (PDEs) by numerical methods meet ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/MultiPDENet:-PDE-embedded-Learning-with-for-Flow-Wang-Mi/6aee4adf8e7489f251995859a5f0432a2c60bb82", "content": "A PDE-embedded network with multiscale time stepping ( MultiPDENet ), which fuses the scheme of numerical methods and machine learning, for accelerated simulation of flows and achieves the state-of-the-art performance compared with other neural baseline models, also with clear speedup compared to classical numerical methods. Solving partial differential equations (PDEs) by numerical methods meet ..."}
+{"idx": 3, "title": "MultiPDENet: PDE-embedded Learning with Multi-time-stepping for ...", "date": "", "ddg_snippet": "MultiPDENet : PDE-embedded Learning with Multi-time-stepping for Accelerated Flow Simulation: Paper and Code. Solving partial differential equations (PDEs) by numerical methods meet computational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability ...", "subpage_snippet": "", "source": "www.catalyzex.com", "link": "https://www.catalyzex.com/paper/multipdenet-pde-embedded-learning-with-multi", "content": "MultiPDENet : PDE-embedded Learning with Multi-time-stepping for Accelerated Flow Simulation: Paper and Code. Solving partial differential equations (PDEs) by numerical methods meet computational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability ..."}
+{"idx": 4, "title": "[2501.15987] MultiPDENet: PDE-embedded Learning with Multi-time ...", "date": "", "ddg_snippet": "Solving partial differential equations (PDEs) by numerical methods meet computational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability, interpretability, and data dependency, as well as suffer in long-term prediction. To this end, we propose a PDE-embedded ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2501.15987", "content": "Solving partial differential equations (PDEs) by numerical methods meet computational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability, interpretability, and data dependency, as well as suffer in long-term prediction. To this end, we propose a PDE-embedded ..."}
+{"idx": 5, "title": "PDE-EMBEDDED LEARNING WITH MULTI TIME STEPPING FOR ... - OpenReview", "date": "", "ddg_snippet": "ABSTRACT Solving partial differential equations (PDEs) by numerical methods meet compu- tational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability, interpretability, and data dependency, as well as suf- fer in long-term prediction.To this end, we propose a ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=stcN89QGfL", "content": "ABSTRACT Solving partial differential equations (PDEs) by numerical methods meet compu- tational cost challenge for getting the accurate solution since fine grids and small time steps are required. Machine learning can accelerate this process, but struggle with weak generalizability, interpretability, and data dependency, as well as suf- fer in long-term prediction.To this end, we propose a ..."}
+{"idx": 6, "title": "Venues | OpenReview", "date": "", "ddg_snippet": "Promoting openness in scientific communication and the peer-review process", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/", "content": "Promoting openness in scientific communication and the peer-review process"}
+{"idx": 7, "title": "GitHub - ZichaoLong/PDE-Net: PDE-Net: Learning PDEs from Data", "date": "", "ddg_snippet": "PDE-Net: Learning PDEs from Data. Contribute to ZichaoLong/PDE-Net development by creating an account on GitHub.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/ZichaoLong/PDE-Net", "content": "PDE-Net: Learning PDEs from Data. Contribute to ZichaoLong/PDE-Net development by creating an account on GitHub."}
+{"idx": 8, "title": "Synergistic learning with multi-task DeepONet for efficient PDE problem ...", "date": "", "ddg_snippet": "Multi-task learning (MTL) is an inductive transfer mechanism designed to leverage useful information from multiple tasks to improve generalization per…", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0893608024010426", "content": "Multi-task learning (MTL) is an inductive transfer mechanism designed to leverage useful information from multiple tasks to improve generalization per…"}
+{"idx": 9, "title": "Target-specific de novo design of drug candidate molecules ... - Nature", "date": "", "ddg_snippet": "Inhibiting AKT1 kinase can have potentially positive uses against many types of cancer. To find novel molecules targeting this protein, a graph adversarial network is trained as a generative model.", "subpage_snippet": "", "source": "www.nature.com", "link": "https://www.nature.com/articles/s42256-025-01082-y", "content": "Inhibiting AKT1 kinase can have potentially positive uses against many types of cancer. To find novel molecules targeting this protein, a graph adversarial network is trained as a generative model."}
diff --git a/data/sampled_jsons/NJU-LHRSLHRS-Bot_github.jsonl b/data/sampled_jsons/NJU-LHRSLHRS-Bot_github.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7de73aab95898b9df07b92a8a310f4be53853024
--- /dev/null
+++ b/data/sampled_jsons/NJU-LHRSLHRS-Bot_github.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "NJU-LHRS/LHRS-Bot: VGI-Enhanced multimodal large ...", "date": "", "ddg_snippet": "We are excited to introduce LHRS-Bot , a multimodal large language model (MLLM) that leverages globally available volunteer geographic information (VGI) and ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/NJU-LHRS/LHRS-Bot", "content": "We are excited to introduce LHRS-Bot , a multimodal large language model (MLLM) that leverages globally available volunteer geographic information (VGI) and ..."}
+{"idx": 1, "title": "LHRS-Bot-Nova: Improved multimodal large language ...", "date": "", "ddg_snippet": "by Z Li · 2025 · Cited by 11 — LHRS-Bot-Nova features an enhanced vision encoder and a novel bridge layer, enabling efficient visual compression and better language-vision alignment.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0924271625002230", "content": "by Z Li · 2025 · Cited by 11 — LHRS-Bot-Nova features an enhanced vision encoder and a novel bridge layer, enabling efficient visual compression and better language-vision alignment."}
+{"idx": 2, "title": "Revision History for LHRS-Bot: Empowering Remote Sensing...", "date": "", "ddg_snippet": "Comprehensive experiments demonstrate that LHRS-Bot exhibits a profound understanding of RS images and the ability to perform nuanced reasoning within the RS ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/revisions?id=wx6gDnGRPi", "content": "Comprehensive experiments demonstrate that LHRS-Bot exhibits a profound understanding of RS images and the ability to perform nuanced reasoning within the RS ..."}
+{"idx": 3, "title": "LHRS-Bot: Empowering Remote Sensing with VGI ...", "date": "", "ddg_snippet": "by D Muhtar · 2024 · Cited by 103 — We introduce LHRS-Bot , an MLLM tailored for RS image understanding through a novel multi-level vision-language alignment strategy and a curriculum learning ...", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "https://ui.adsabs.harvard.edu/abs/2024arXiv240202544M/abstract", "content": "by D Muhtar · 2024 · Cited by 103 — We introduce LHRS-Bot , an MLLM tailored for RS image understanding through a novel multi-level vision-language alignment strategy and a curriculum learning ..."}
+{"idx": 4, "title": "LHRS-Bot: Empowering Remote Sensing with VGI- ...", "date": "", "ddg_snippet": "21 Nov 2024 — We introduce LHRS-Bot , an MLLM tailored for RS image understanding through a novel multi-level vision-language alignment strategy and a curriculum learning ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-031-72904-1_26", "content": "21 Nov 2024 — We introduce LHRS-Bot , an MLLM tailored for RS image understanding through a novel multi-level vision-language alignment strategy and a curriculum learning ..."}
+{"idx": 5, "title": "LHRS-Bot: Empowering Remote Sensing with VGI ...", "date": "", "ddg_snippet": "by D Muhtar · 2024 · Cited by 103 — Comprehensive experiments demonstrate that LHRS-Bot exhibits a profound understanding of RS images and the ability to perform nuanced reasoning ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1007/978-3-031-72904-1_26", "content": "by D Muhtar · 2024 · Cited by 103 — Comprehensive experiments demonstrate that LHRS-Bot exhibits a profound understanding of RS images and the ability to perform nuanced reasoning ..."}
+{"idx": 6, "title": "LHRS-Bot-Nova: Improved Multimodal Large Language ...", "date": "", "ddg_snippet": "14 Nov 2024 — In this study, we introduce LHRS-Bot-Nova , an MLLM specialized in understanding remote sensing (RS) images, designed to expertly perform a wide ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2411.09301v1", "content": "14 Nov 2024 — In this study, we introduce LHRS-Bot-Nova , an MLLM specialized in understanding remote sensing (RS) images, designed to expertly perform a wide ..."}
+{"idx": 7, "title": "Revision History for LHRS-Bot-Nova: Improved Multimodal...", "date": "", "ddg_snippet": "LHRS-Bot-Nova features an enhanced vision encoder and a novel bridge layer , enabling efficient visual compression and better language-vision alignment. To ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/revisions?id=9vXs46X9Kr", "content": "LHRS-Bot-Nova features an enhanced vision encoder and a novel bridge layer , enabling efficient visual compression and better language-vision alignment. To ..."}
+{"idx": 8, "title": "LHRS", "date": "", "ddg_snippet": "The official implementation of paper \"Unified Self-Supervised Learning Framework for Remote Sensing Images\".", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/NJU-LHRS", "content": "The official implementation of paper \"Unified Self-Supervised Learning Framework for Remote Sensing Images\"."}
+{"idx": 9, "title": "LHRS-Bot only answer some of my data · Issue #30", "date": "", "ddg_snippet": "When I test some of my data, I come across a situation where only part of it has an answer (in the key of \" answer\"), what is going on? And how to solve it?", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/NJU-LHRS/LHRS-Bot/issues/30", "content": "When I test some of my data, I come across a situation where only part of it has an answer (in the key of \" answer\"), what is going on? And how to solve it?"}
diff --git a/data/sampled_jsons/Neural_Persistence_Dynamics_60-dimensional_vectorization.jsonl b/data/sampled_jsons/Neural_Persistence_Dynamics_60-dimensional_vectorization.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ac1d533ee7e9322bc1432a238801a64fadd0899d
--- /dev/null
+++ b/data/sampled_jsons/Neural_Persistence_Dynamics_60-dimensional_vectorization.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "A neural circuit for flexible control of persistent behavioral", "date": "", "ddg_snippet": "... we find that mutual inhibition between two antagonistic neuromodulatory systems underlies the persistence and mutual exclusivity of the neural ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/articles/62889", "content": "... we find that mutual inhibition between two antagonistic neuromodulatory systems underlies the persistence and mutual exclusivity of the neural ..."}
+{"idx": 1, "title": "US10553218B2 - Dimensionality reduction of baum-welch", "date": "", "ddg_snippet": "Those i-vectors are computed via dimensionality reduction of the first-order statistics through the total variability procedure.", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US10553218B2/en", "content": "Those i-vectors are computed via dimensionality reduction of the first-order statistics through the total variability procedure."}
+{"idx": 2, "title": "Dynamical System Parameter Path Optimization using Persistent", "date": "", "ddg_snippet": "High- dimensional parameter spaces of dynamical systems have been explored using features of numerical solutions in [ 18 ] , but this method relies on ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.00782v1", "content": "High- dimensional parameter spaces of dynamical systems have been explored using features of numerical solutions in [ 18 ] , but this method relies on ..."}
+{"idx": 3, "title": "Embodied sensorimotor control: computational modeling of the", "date": "", "ddg_snippet": "We then summarize evidence that neural population activity occupies low- dimensional , dynamically evolving manifolds during planning and execution of ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.14360v1", "content": "We then summarize evidence that neural population activity occupies low- dimensional , dynamically evolving manifolds during planning and execution of ..."}
+{"idx": 4, "title": "Persistent vocal learning in an aging open-ended learner", "date": "", "ddg_snippet": "We find neural evidence to support persistent vocal learning in the budgerigar, suggesting resilience to aging in the open-ended learning program of ...", "subpage_snippet": "", "source": "bmcneurosci.biomedcentral.com", "link": "https://bmcneurosci.biomedcentral.com/articles/10.1186/s12868-024-00879-8", "content": "We find neural evidence to support persistent vocal learning in the budgerigar, suggesting resilience to aging in the open-ended learning program of ..."}
+{"idx": 5, "title": "Exploring the repository of de novo-designed bifunctional", "date": "", "ddg_snippet": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10– 60 amino acids ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/articles/97330", "content": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10– 60 amino acids ..."}
+{"idx": 6, "title": "Exploring the repository of de novo designed bifunctional", "date": "", "ddg_snippet": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/reviewed-preprints/97330", "content": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ..."}
+{"idx": 7, "title": "Exploring the repository of de novo designed bifunctional", "date": "", "ddg_snippet": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/reviewed-preprints/97330v1", "content": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ..."}
+{"idx": 8, "title": "Exploring the repository of de novo designed bifunctional", "date": "", "ddg_snippet": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/reviewed-preprints/97330v2", "content": "As a naturally evolved defense line of innate immune system, antimicrobial peptides (AMPs) are cationic macromolecules (usually 10- 60 amino acids ..."}
+{"idx": 9, "title": "US5371853A - Method and system for CELP speech coding and", "date": "", "ddg_snippet": "vectors ... 239000013598 vector Substances 0.000 claims abstract description 102", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US5371853A/en", "content": "vectors ... 239000013598 vector Substances 0.000 claims abstract description 102"}
diff --git a/data/sampled_jsons/No_Free_Delivery_Service_Epistemic_limits_of_passive_data_collection_train-test_paradigm_conclusion.jsonl b/data/sampled_jsons/No_Free_Delivery_Service_Epistemic_limits_of_passive_data_collection_train-test_paradigm_conclusion.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c74bb019a04e83d585aeded943f03942eb857550
--- /dev/null
+++ b/data/sampled_jsons/No_Free_Delivery_Service_Epistemic_limits_of_passive_data_collection_train-test_paradigm_conclusion.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Undergraduate engineering students’ epistemic cognition and", "date": "", "ddg_snippet": "... of the EC and reported epistemic changes of undergraduate engineering students who served as EDP mentors in a secondary-school STEM-based community ...", "subpage_snippet": "", "source": "stemeducationjournal.springeropen.com", "link": "https://stemeducationjournal.springeropen.com/articles/10.1186/s40594-025-00564-0", "content": "... of the EC and reported epistemic changes of undergraduate engineering students who served as EDP mentors in a secondary-school STEM-based community ..."}
+{"idx": 1, "title": "1573 questions with answers in TRAINING | Science topic", "date": "", "ddg_snippet": "... epistemological and operational foundation of security technicians in South Africa amid the tectonic shifts induced by the Fourth Industrial ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/topic/Training", "content": "... epistemological and operational foundation of security technicians in South Africa amid the tectonic shifts induced by the Fourth Industrial ..."}
+{"idx": 2, "title": "Why teachers don’t know best | Ed Tech Now", "date": "", "ddg_snippet": "... through no fault of their own, teachers do not have the skill-set that is required to improve the chronic under-performance of our education service ...", "subpage_snippet": "", "source": "edtechnow.net", "link": "https://edtechnow.net/2013/08/27/blind/", "content": "... through no fault of their own, teachers do not have the skill-set that is required to improve the chronic under-performance of our education service ..."}
+{"idx": 3, "title": "Learning How to Lead Through Engagement with Enquiry", "date": "", "ddg_snippet": "... nature of EBL is a holistic experience that enables epistemic development and has all the features of a threshold concept yet it is a process and not ...", "subpage_snippet": "", "source": "abcdocz.com", "link": "https://abcdocz.com/doc/198424/learning-how-to-lead-through-engagement-with-enquiry", "content": "... nature of EBL is a holistic experience that enables epistemic development and has all the features of a threshold concept yet it is a process and not ..."}
+{"idx": 4, "title": "Creating a rich curriculum through intentional teaching | New", "date": "", "ddg_snippet": "This passive positioning of teachers in early childhood education may be detrimental to young children’s learning and wellbeing.", "subpage_snippet": "", "source": "www.nzcer.org.nz", "link": "https://www.nzcer.org.nz/nzcerpress/early-childhood-folio/articles/creating-rich-curriculum-through-intentional-teaching", "content": "This passive positioning of teachers in early childhood education may be detrimental to young children’s learning and wellbeing."}
+{"idx": 5, "title": "Education | Center for International Maritime Security | Page 2", "date": "", "ddg_snippet": "... training and education include explicit calls for changing the industrial age paradigm to a post-industrial age one, as well as considerations of the ...", "subpage_snippet": "", "source": "cimsec.org", "link": "https://cimsec.org/category/education/page/2/", "content": "... training and education include explicit calls for changing the industrial age paradigm to a post-industrial age one, as well as considerations of the ..."}
+{"idx": 6, "title": "AI as Normal Technology", "date": "", "ddg_snippet": "In the case of generative AI, even failures that seem extremely obvious in hindsight were not caught during testing .", "subpage_snippet": "", "source": "www.normaltech.ai", "link": "https://www.normaltech.ai/p/ai-as-normal-technology", "content": "In the case of generative AI, even failures that seem extremely obvious in hindsight were not caught during testing ."}
+{"idx": 7, "title": "AI as Normal Technology | Knight First Amendment Institute", "date": "", "ddg_snippet": "... of controlling superintelligent AI will, in fact, make things much worse if AI turns out to be normal technology — the downsides of which will ...", "subpage_snippet": "", "source": "knightcolumbia.org", "link": "https://knightcolumbia.org/content/ai-as-normal-technology", "content": "... of controlling superintelligent AI will, in fact, make things much worse if AI turns out to be normal technology — the downsides of which will ..."}
+{"idx": 8, "title": "August | 2024 | Her2 signal | Page 2", "date": "", "ddg_snippet": "Retrospectively examined subjective ramifications of preliminary opioid misuse were contrasted in a pilot sample of opioid misusers (nonmedical use ...", "subpage_snippet": "", "source": "her2signal.com", "link": "https://her2signal.com/index.php/2024/08/page/2/", "content": "Retrospectively examined subjective ramifications of preliminary opioid misuse were contrasted in a pilot sample of opioid misusers (nonmedical use ..."}
+{"idx": 9, "title": "tarekmo – Dr. Tarek Mostafa", "date": "", "ddg_snippet": "Unlike high-performing boys, high-performing girls may not pursue a career in science simply because they are likely to be at or near the top of the ...", "subpage_snippet": "", "source": "tarekmostafa.net", "link": "http://tarekmostafa.net/author/tarekmo", "content": "Unlike high-performing boys, high-performing girls may not pursue a career in science simply because they are likely to be at or near the top of the ..."}
diff --git a/data/sampled_jsons/OSG_algorithm_Xu_et_al._2023_abstract_fully_connected_communication_graph_year_2023.jsonl b/data/sampled_jsons/OSG_algorithm_Xu_et_al._2023_abstract_fully_connected_communication_graph_year_2023.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..35d9d3dff420aebb8d10a52edc942512f6b59edf
--- /dev/null
+++ b/data/sampled_jsons/OSG_algorithm_Xu_et_al._2023_abstract_fully_connected_communication_graph_year_2023.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2309.05305] Fully-Connected Spatial-Temporal Graph for ... Fully-Connected Spatial-Temporal Graph for Multivariate Time ... Graph-based deep learning for communication networks: A ... Fully-Connected Spatial-Temporal Graph for Multivariate Time ... NEAR-OPTIMAL ONLINE LEARNING FOR MULTI-AGENT SUBMODULAR ... [2309.05305] Fully-Connected Spatial-Temporal Graph for ... Fully-Connected Spatial-Temporal Graph for Multivariate Time - Series … Graph-based deep learning for communication networks: A survey Graph-based deep learning for communication networks: A survey Deep Hierarchical Communication Graph in Multi-Agent ... - IJCAI Graph-based deep learning for communication networks: A survey Fully - Connected Spatial-Temporal Graph for Multivariate Time-Series D… Deep Hierarchical Communication Graph in Multi-Agent ...", "date": "", "ddg_snippet": "Sep 11, 2023 · Multivariate Time-Series (MTS) data is crucial in various application fields. With its sequential and multi-source (multiple sensors) properties, MTS data inherently exhibits Spatial-Temporal (ST) dependencies, involving temporal correlations between timestamps and spatial correlations between sensors in each timestamp. To effectively leverage this information, Graph Neural Network-based ... Mar 24, 2024 · To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution. Mar 1, 2022 · Communication networks are important infrastructures in contemporary society. There are still many challenges that are not fully solved and new solutions are proposed continuously in this active research area. In recent years, to model the network topology, graph -based deep learning has achieved the state-of-the-art performance in a series of problems in communication networks. In this survey ... Sep 11, 2023 · Fully - Connected Spatial-T emporal Graph for Multivariate T ime Series Data Y ucheng Wang 1,2, Y uecong Xu 2, Jianfei Y ang 1, Min Wu 2, Xiaoli Li 1,2,3, Lihua Xie 1, Zhenghua Chen 2,3* ABSTRACT Coordinating multiple agents to collaboratively maximize submodular functions in unpredictable environments is a critical task with numerous applications in machine learning, robot planning and control. The existing approaches, such as the OSG algorithm , are often hindered by their poor approximation guarantees and the rigid requirement for a fully connected communication graph . To ... Feb 28, 2024 · To address this limitation, Graph Neural Network-based methods (GNNs) have emerged as popular solutions to exploit ST dependencies within MTS data (Wang et al . 2023b; Jia et al . 2020). Figure 1: ST graphs are constructed from MTS data, creating separate graphs for each timestamp, to capture ST dependencies. What is fully connected spatial-temporal graph neural network (FC-stgnn)? To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution. What is a simple graph in communication networks? Graphs in communication networks From the graph theory, a simple graph is defined as G = (V, E) , where V is the set of nodes and E is the set of edges between nodes. In communication networks, the edges can be either directed or undirected, depending on the specific problems. Are GNNS suitable for problems in communication networks? Fig. 1. The organization of this survey. GNNs are suitable for problems in communication networks because of their strong learning ability to capture the spatial information hidden in the network topology and their generalization ability to be used in unseen topologies when the networks are dynamic. Does the hierarchical communication graph improve learning speed and coordination performance? In addition, DHCG-Q significantly outperforms all baselines, indicating that intention sharing through message-dependent directed acyclic graphs can improve learning speed and coordination performance in more complex tasks. In this section, we conduct ablation studies to demonstrate the superiority of the hierarchical communication graph. Is interference graph convolutional neural network robust to imperfect channel state information? For solving the similar problem in an unsupervised approach, Interference Graph Convolutional Neural Network (IGCNet) is proposed and validated in , which is robust to imperfect Channel State Information (CSI). Can attention graph convolutional network Fuse multi-sensor signals? Hierarchical attention graph convolutional network to fuse multi-sensor signals for remaining useful life prediction. Reliability Engineering & System Safety, 215: 107878. Li et al . (2021b) Li, T.; Zhao, Z.; Sun, C.; Yan, R.; and Chen, X. 2021b. Multireceptive Field Graph Convolutional Networks for Machine Fault Diagnosis. Coordination graph [Guestrin et al ., 2002] is a graph -based value factorization method where the local observations and the actions are shared through the edge be-tween connected agents.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2309.05305", "content": "Sep 11, 2023 · Multivariate Time-Series (MTS) data is crucial in various application fields. With its sequential and multi-source (multiple sensors) properties, MTS data inherently exhibits Spatial-Temporal (ST) dependencies, involving temporal correlations between timestamps and spatial correlations between sensors in each timestamp. To effectively leverage this information, Graph Neural Network-based ... Mar 24, 2024 · To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution. Mar 1, 2022 · Communication networks are important infrastructures in contemporary society. There are still many challenges that are not fully solved and new solutions are proposed continuously in this active research area. In recent years, to model the network topology, graph -based deep learning has achieved the state-of-the-art performance in a series of problems in communication networks. In this survey ... Sep 11, 2023 · Fully - Connected Spatial-T emporal Graph for Multivariate T ime Series Data Y ucheng Wang 1,2, Y uecong Xu 2, Jianfei Y ang 1, Min Wu 2, Xiaoli Li 1,2,3, Lihua Xie 1, Zhenghua Chen 2,3* ABSTRACT Coordinating multiple agents to collaboratively maximize submodular functions in unpredictable environments is a critical task with numerous applications in machine learning, robot planning and control. The existing approaches, such as the OSG algorithm , are often hindered by their poor approximation guarantees and the rigid requirement for a fully connected communication graph . To ... Feb 28, 2024 · To address this limitation, Graph Neural Network-based methods (GNNs) have emerged as popular solutions to exploit ST dependencies within MTS data (Wang et al . 2023b; Jia et al . 2020). Figure 1: ST graphs are constructed from MTS data, creating separate graphs for each timestamp, to capture ST dependencies. What is fully connected spatial-temporal graph neural network (FC-stgnn)? To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution. What is a simple graph in communication networks? Graphs in communication networks From the graph theory, a simple graph is defined as G = (V, E) , where V is the set of nodes and E is the set of edges between nodes. In communication networks, the edges can be either directed or undirected, depending on the specific problems. Are GNNS suitable for problems in communication networks? Fig. 1. The organization of this survey. GNNs are suitable for problems in communication networks because of their strong learning ability to capture the spatial information hidden in the network topology and their generalization ability to be used in unseen topologies when the networks are dynamic. Does the hierarchical communication graph improve learning speed and coordination performance? In addition, DHCG-Q significantly outperforms all baselines, indicating that intention sharing through message-dependent directed acyclic graphs can improve learning speed and coordination performance in more complex tasks. In this section, we conduct ablation studies to demonstrate the superiority of the hierarchical communication graph. Is interference graph convolutional neural network robust to imperfect channel state information? For solving the similar problem in an unsupervised approach, Interference Graph Convolutional Neural Network (IGCNet) is proposed and validated in , which is robust to imperfect Channel State Information (CSI). Can attention graph convolutional network Fuse multi-sensor signals? Hierarchical attention graph convolutional network to fuse multi-sensor signals for remaining useful life prediction. Reliability Engineering & System Safety, 215: 107878. Li et al . (2021b) Li, T.; Zhao, Z.; Sun, C.; Yan, R.; and Chen, X. 2021b. Multireceptive Field Graph Convolutional Networks for Machine Fault Diagnosis. Coordination graph [Guestrin et al ., 2002] is a graph -based value factorization method where the local observations and the actions are shared through the edge be-tween connected agents."}
+{"idx": 1, "title": "Fully-Connected Spatial-Temporal Graph for Multivariate Time ...", "date": "", "ddg_snippet": "Mar 24, 2024 · To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution.", "subpage_snippet": "", "source": "ojs.aaai.org", "link": "https://ojs.aaai.org/index.php/AAAI/article/view/29500", "content": "Mar 24, 2024 · To address this limitation, we propose a novel method called Fully - Connected Spatial-Temporal Graph Neural Network (FC-STGNN), including two key components namely FC graph construction and FC graph convolution."}
+{"idx": 2, "title": "Graph-based deep learning for communication networks: A ...", "date": "", "ddg_snippet": "Mar 1, 2022 · Communication networks are important infrastructures in contemporary society. There are still many challenges that are not fully solved and new solutions are proposed continuously in this active research area. In recent years, to model the network topology, graph -based deep learning has achieved the state-of-the-art performance in a series of problems in communication networks. In this survey ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0140366421004874", "content": "Mar 1, 2022 · Communication networks are important infrastructures in contemporary society. There are still many challenges that are not fully solved and new solutions are proposed continuously in this active research area. In recent years, to model the network topology, graph -based deep learning has achieved the state-of-the-art performance in a series of problems in communication networks. In this survey ..."}
+{"idx": 3, "title": "Fully-Connected Spatial-Temporal Graph for Multivariate Time ...", "date": "", "ddg_snippet": "Sep 11, 2023 · Fully - Connected Spatial-T emporal Graph for Multivariate T ime Series Data Y ucheng Wang 1,2, Y uecong Xu 2, Jianfei Y ang 1, Min Wu 2, Xiaoli Li 1,2,3, Lihua Xie 1, Zhenghua Chen 2,3*", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/373838165_Fully-Connected_Spatial-Temporal_Graph_for_Multivariate_Time_Series_Data", "content": "Sep 11, 2023 · Fully - Connected Spatial-T emporal Graph for Multivariate T ime Series Data Y ucheng Wang 1,2, Y uecong Xu 2, Jianfei Y ang 1, Min Wu 2, Xiaoli Li 1,2,3, Lihua Xie 1, Zhenghua Chen 2,3*"}
+{"idx": 4, "title": "[2309.05305] Fully-Connected Spatial-Temporal Graph for ...", "date": "", "ddg_snippet": "Feb 28, 2024 · To address this limitation, Graph Neural Network-based methods (GNNs) have emerged as popular solutions to exploit ST dependencies within MTS data (Wang et al . 2023b; Jia et al . 2020). Figure 1: ST graphs are constructed from MTS data, creating separate graphs for each timestamp, to capture ST dependencies.", "subpage_snippet": "", "source": "ar5iv.labs.arxiv.org", "link": "https://ar5iv.labs.arxiv.org/html/2309.05305", "content": "Feb 28, 2024 · To address this limitation, Graph Neural Network-based methods (GNNs) have emerged as popular solutions to exploit ST dependencies within MTS data (Wang et al . 2023b; Jia et al . 2020). Figure 1: ST graphs are constructed from MTS data, creating separate graphs for each timestamp, to capture ST dependencies."}
+{"idx": 5, "title": "Deep Hierarchical Communication Graph in Multi-Agent ...", "date": "", "ddg_snippet": "Coordination graph [Guestrin et al ., 2002] is a graph -based value factorization method where the local observations and the actions are shared through the edge be-tween connected agents.", "subpage_snippet": "", "source": "www.ijcai.org", "link": "https://www.ijcai.org/proceedings/2023/0024.pdf", "content": "Coordination graph [Guestrin et al ., 2002] is a graph -based value factorization method where the local observations and the actions are shared through the edge be-tween connected agents."}
+{"idx": 6, "title": "tandfonline.com/doi/ full /10.1080/2331186X. 2023 .2290342", "date": "", "ddg_snippet": "Full article: Critical thinking in the AI era: An exploration of EFL students'...", "subpage_snippet": "", "source": "www.tandfonline.com", "link": "https://www.tandfonline.com/doi/full/10.1080/2331186X.2023.2290342", "content": "Full article: Critical thinking in the AI era: An exploration of EFL students'..."}
+{"idx": 7, "title": "NEAR-OPTIMAL ONLINE LEARNING FOR MULTI-AGENT SUBMODULAR ...", "date": "", "ddg_snippet": "ABSTRACT Coordinating multiple agents to collaboratively maximize submodular functions in unpredictable environments is a critical task with numerous applications in machine learning, robot planning and control. The existing approaches, such as the OSG algorithm , are often hindered by their poor approximation guarantees and the rigid requirement for a fully connected communication graph . To ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2502.05028", "content": "ABSTRACT Coordinating multiple agents to collaboratively maximize submodular functions in unpredictable environments is a critical task with numerous applications in machine learning, robot planning and control. The existing approaches, such as the OSG algorithm , are often hindered by their poor approximation guarantees and the rigid requirement for a fully connected communication graph . To ..."}
+{"idx": 8, "title": "N ear -o ptimal", "date": "", "ddg_snippet": "rigid requirement for a fully connected communication graph .the communication graph . Given that the majority of applications occur in time-varying environments, Xu et al . ( 2023 ) proposed the online sequence greedy( OSG ) algorithm for online MA-SM problem", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=i8dYPGdB1C", "content": "rigid requirement for a fully connected communication graph .the communication graph . Given that the majority of applications occur in time-varying environments, Xu et al . ( 2023 ) proposed the online sequence greedy( OSG ) algorithm for online MA-SM problem"}
+{"idx": 9, "title": "Figure 1: Left: Multi-target tracking with 4 mobile sensors over a...", "date": "", "ddg_snippet": "( 2023 ) have recently proposed an online sequential greedy ( OSG ) algorithm , building upon the foundations of the classical greedy method ( Fisher et al ., 1978).", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Left-Multi-target-tracking-with-4-mobile-sensors-over-a-complete-directed-acyclic_fig1_388847678", "content": "( 2023 ) have recently proposed an online sequential greedy ( OSG ) algorithm , building upon the foundations of the classical greedy method ( Fisher et al ., 1978)."}
diff --git a/data/sampled_jsons/Olsson_2022_In-context_Learning_and_Induction_Heads_attention_mechanism_transformers_year_2022.jsonl b/data/sampled_jsons/Olsson_2022_In-context_Learning_and_Induction_Heads_attention_mechanism_transformers_year_2022.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5ebf293ef99251fe09a0537d8de243b4cea489ed
--- /dev/null
+++ b/data/sampled_jsons/Olsson_2022_In-context_Learning_and_Induction_Heads_attention_mechanism_transformers_year_2022.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2209.11895] In-context Learning and Induction Heads", "date": "", "ddg_snippet": "by C Olsson · 2022 · Cited by 614 — We present six complementary lines of evidence, arguing that induction heads may be the mechanistic source of general in-context learning in transformer models ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2209.11895", "content": "by C Olsson · 2022 · Cited by 614 — We present six complementary lines of evidence, arguing that induction heads may be the mechanistic source of general in-context learning in transformer models ..."}
+{"idx": 1, "title": "In-context Learning and Induction Heads", "date": "", "ddg_snippet": "8 Mar 2022 — The induction head, a circuit whose function is to look back over the sequence for previous instances of the current token.", "subpage_snippet": "", "source": "transformer-circuits.pub", "link": "https://transformer-circuits.pub/2022/in-context-learning-and-induction-heads/index.html", "content": "8 Mar 2022 — The induction head, a circuit whose function is to look back over the sequence for previous instances of the current token."}
+{"idx": 2, "title": "In-context Learning and Induction Heads", "date": "", "ddg_snippet": "This work discovers a new type of induction head: concept-level induction heads , which copy entire lexical units instead of individual tokens, and finds ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/In-context-Learning-and-Induction-Heads-Olsson-Elhage/c90a99eeb57019732a6cc996bb9eaf13faedf00f", "content": "This work discovers a new type of induction head: concept-level induction heads , which copy entire lexical units instead of individual tokens, and finds ..."}
+{"idx": 3, "title": "Paper page - In-context Learning and Induction Heads", "date": "", "ddg_snippet": "24 Sept 2022 — Induction heads in large transformer models may be the mechanism enabling in-context learning, with evidence from both small attention-only ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/2209.11895", "content": "24 Sept 2022 — Induction heads in large transformer models may be the mechanism enabling in-context learning, with evidence from both small attention-only ..."}
+{"idx": 4, "title": "Talk on In-Context Learning and Induction Heads", "date": "", "ddg_snippet": "The paper discusses induction heads , which are a proposed emergent structure in transformers which implements a particular algorithm.", "subpage_snippet": "", "source": "rohanhitchcock.com", "link": "https://rohanhitchcock.com/notes/2023-4-6-slt-seminar-induction-heads.html", "content": "The paper discusses induction heads , which are a proposed emergent structure in transformers which implements a particular algorithm."}
+{"idx": 5, "title": "In-context Learning and Induction Heads (2022)", "date": "", "ddg_snippet": "TL;DR: In this paper , the authors investigate the hypothesis that transformer -based in-context learners implement standard learning algorithms implicitly, by ...", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/papers/in-context-learning-and-induction-heads-u6v55ro2", "content": "TL;DR: In this paper , the authors investigate the hypothesis that transformer -based in-context learners implement standard learning algorithms implicitly, by ..."}
+{"idx": 6, "title": "In-context Learning and Induction Heads (2209.11895v1)", "date": "", "ddg_snippet": "This paper shows how induction heads drive in-context learning in transformers , enhancing model interpretability and AI safety through empirical evidence.", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/articles/2209.11895", "content": "This paper shows how induction heads drive in-context learning in transformers , enhancing model interpretability and AI safety through empirical evidence."}
+{"idx": 7, "title": "Which Attention Heads Matter for In-Context Learning?", "date": "", "ddg_snippet": "19 Feb 2025 — To better understand which of the two distinct mechanisms drives ICL, we study and compare induction heads and FV heads in 12 language models.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.14010v1", "content": "19 Feb 2025 — To better understand which of the two distinct mechanisms drives ICL, we study and compare induction heads and FV heads in 12 language models."}
+{"idx": 8, "title": "Unveiling Induction Heads: Provable Training Dynamics ...", "date": "", "ddg_snippet": "5 Nov 2024 — This paper theoretically explores how simplified transformers perform in-context learning (ICL) on n-gram Markov chain data. The authors analyze ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=4fN2REs0Ma&referrer=[the+profile+of+Siyu+Chen](/profile?id=~Siyu_Chen2)", "content": "5 Nov 2024 — This paper theoretically explores how simplified transformers perform in-context learning (ICL) on n-gram Markov chain data. The authors analyze ..."}
+{"idx": 9, "title": "Induction Heads as an Essential Mechanism for Pattern ...", "date": "", "ddg_snippet": "by J Crosbie · 2025 · Cited by 23 — Building on this foundation, Olsson et al. ( 2022 ) hypothesised that induction heads are capable of abstract pattern matching and conducted a ... 63 pages", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.findings-naacl.283.pdf", "content": "by J Crosbie · 2025 · Cited by 23 — Building on this foundation, Olsson et al. ( 2022 ) hypothesised that induction heads are capable of abstract pattern matching and conducted a ... 63 pages"}
diff --git a/data/sampled_jsons/On_the_convergence_of_the_coordinate_descent_method_for_convex_differentiable_minimization_Luo_Tseng.jsonl b/data/sampled_jsons/On_the_convergence_of_the_coordinate_descent_method_for_convex_differentiable_minimization_Luo_Tseng.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..009f3494e8479c06315318ebae73fe3053a81057
--- /dev/null
+++ b/data/sampled_jsons/On_the_convergence_of_the_coordinate_descent_method_for_convex_differentiable_minimization_Luo_Tseng.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Coordinate descent - Wikipedia", "date": "", "ddg_snippet": "Coordinate descent is an optimization algorithm that successively minimizes along coordinate directions to find the minimum of a function. At each iteration, the algorithm determines a coordinate or coordinate block via a coordinate selection rule...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Coordinate_descent", "content": "Coordinate descent is an optimization algorithm that successively minimizes along coordinate directions to find the minimum of a function. At each iteration, the algorithm determines a coordinate or coordinate block via a coordinate selection rule..."}
+{"idx": 1, "title": "On the convergence of the coordinate descent method for convex ...", "date": "", "ddg_snippet": "The coordinate descent method enjoys a long history in convex differentiable minimization . Luo , Z. Q., andTseng, P., On the Linear Convergence of Descent Methods for Convex Essentially Smooth Minimization, SIAM Journal on Control and Optimization (to appear).", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/BF00939948", "content": "The coordinate descent method enjoys a long history in convex differentiable minimization . Luo , Z. Q., andTseng, P., On the Linear Convergence of Descent Methods for Convex Essentially Smooth Minimization, SIAM Journal on Control and Optimization (to appear)."}
+{"idx": 2, "title": "On the convergence of the coordinate descent method for convex ...", "date": "", "ddg_snippet": "Author(s). Luo , Zhi-Quan.; Tseng , Paul.; Massachusetts Institute of Technology. Laboratory for Information and Decision Systems.", "subpage_snippet": "", "source": "dspace.mit.edu", "link": "https://dspace.mit.edu/handle/1721.1/3164", "content": "Author(s). Luo , Zhi-Quan.; Tseng , Paul.; Massachusetts Institute of Technology. Laboratory for Information and Decision Systems."}
+{"idx": 3, "title": "On the convergence of the coordinate descent method for convex ...", "date": "", "ddg_snippet": "ZLZ. Q. Luo . On the solution of large, structured linear complementarity problems: The block partitioned case.", "subpage_snippet": "", "source": "www.scilit.com", "link": "https://www.scilit.com/publications/1d395fdcea75f76dd04e45688a3ce4a8", "content": "ZLZ. Q. Luo . On the solution of large, structured linear complementarity problems: The block partitioned case."}
+{"idx": 4, "title": "Zhi-Quan Luo - Google Akademik", "date": "", "ddg_snippet": "On the convergence of the coordinate descent method for convex differentiable minimization . ZQ Luo , P Tseng . Journal of Optimization Theory and Applications 72 (1), 7-35, 1992 .", "subpage_snippet": "", "source": "scholar.google.com.eg", "link": "https://scholar.google.com.eg/citations?user=dW3gcXoAAAAJ&hl=tr", "content": "On the convergence of the coordinate descent method for convex differentiable minimization . ZQ Luo , P Tseng . Journal of Optimization Theory and Applications 72 (1), 7-35, 1992 ."}
+{"idx": 5, "title": "Paul Tseng | University of Washington | 123 Publications | 2964 Citations", "date": "", "ddg_snippet": "On the convergence of the coordinate descent method for convex differentiable minimization . Zhi-Quan Luo , Paul Tseng +1 moreMcMaster University, Massachusetts Institute of Technology. - 03 Jan 1992 .", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/authors/paul-tseng-3mwix2ae6s", "content": "On the convergence of the coordinate descent method for convex differentiable minimization . Zhi-Quan Luo , Paul Tseng +1 moreMcMaster University, Massachusetts Institute of Technology. - 03 Jan 1992 ."}
+{"idx": 6, "title": "ds.algorithms - Theoretical study of coordinate descent methods...", "date": "", "ddg_snippet": "In that sense, Luo & Tseng did prove linear convergence rates for some non-strongly convex objective functions in the paper \" On the convergence of the coordinate descent method for convex differentiable minimization \".", "subpage_snippet": "", "source": "cstheory.stackexchange.com", "link": "https://cstheory.stackexchange.com/questions/7715/theoretical-study-of-coordinate-descent-methods", "content": "In that sense, Luo & Tseng did prove linear convergence rates for some non-strongly convex objective functions in the paper \" On the convergence of the coordinate descent method for convex differentiable minimization \"."}
+{"idx": 7, "title": "GitHub - Evolutionary-Intelligence/CoordinateDescent: A (actively...)", "date": "", "ddg_snippet": "Luo , Z.Q. and Tseng , P., 1992 . On the convergence of the coordinate descent method for convex differentiable minimization .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/Evolutionary-Intelligence/CoordinateDescent", "content": "Luo , Z.Q. and Tseng , P., 1992 . On the convergence of the coordinate descent method for convex differentiable minimization ."}
+{"idx": 8, "title": "On the Linear Convergence of Descent Methods for Convex ...", "date": "", "ddg_snippet": "Z.-Q. Luo , P. Tseng , On the convergence of the coordinate descent method for convex differentiable minimization , J. Optim.", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/10.1137/0330025", "content": "Z.-Q. Luo , P. Tseng , On the convergence of the coordinate descent method for convex differentiable minimization , J. Optim."}
+{"idx": 9, "title": "[PDF] On the linear convergence of descent methods for convex ...", "date": "", "ddg_snippet": "The coordinate descent method enjoys a long history in convex differentiable minimization . Surprisingly, very little is known about the convergence of the iterates generated by this method.…", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/On-the-linear-convergence-of-descent-methods-for-Luo-Tseng/096893f5c8235c377d1d4e6d6ad643a9bb271a15", "content": "The coordinate descent method enjoys a long history in convex differentiable minimization . Surprisingly, very little is known about the convergence of the iterates generated by this method.…"}
diff --git a/data/sampled_jsons/Overcoming_catastrophic_forgetting_in_neural_networks_arxiv_abstract_year_2017.jsonl b/data/sampled_jsons/Overcoming_catastrophic_forgetting_in_neural_networks_arxiv_abstract_year_2017.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..38d39a7d5b71c253b0082cb5178ed83f4be563f3
--- /dev/null
+++ b/data/sampled_jsons/Overcoming_catastrophic_forgetting_in_neural_networks_arxiv_abstract_year_2017.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "by J Kirkpatrick · 2016 · Cited by 10388 — We show that it is possible to overcome this limitation and train networks that can maintain expertise on tasks which they have not experienced for a long time.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1612.00796", "content": "by J Kirkpatrick · 2016 · Cited by 10388 — We show that it is possible to overcome this limitation and train networks that can maintain expertise on tasks which they have not experienced for a long time."}
+{"idx": 1, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "by BSY Loke · 2025 · Cited by 1 — Abstract : Catastrophic forgetting is the primary challenge that hinders continual learning, which refers to a neural network ability to ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2507.10485", "content": "by BSY Loke · 2025 · Cited by 1 — Abstract : Catastrophic forgetting is the primary challenge that hinders continual learning, which refers to a neural network ability to ..."}
+{"idx": 2, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "14 Jul 2025 — This work aims to evaluate EWC's effectiveness in mitigating catastrophic forgetting and its potential as a general solution for sequential task learning in ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.10485v1", "content": "14 Jul 2025 — This work aims to evaluate EWC's effectiveness in mitigating catastrophic forgetting and its potential as a general solution for sequential task learning in ..."}
+{"idx": 3, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "by BSY Loke · 2025 · Cited by 1 — Abstract — Catastrophic forgetting is the primary challenge that hinders continual learning, which refers to a neural net- work ability to ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2507.10485", "content": "by BSY Loke · 2025 · Cited by 1 — Abstract — Catastrophic forgetting is the primary challenge that hinders continual learning, which refers to a neural net- work ability to ..."}
+{"idx": 4, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "2 Dec 2016 — A method to prevent catastrophic forgetting in neural networks by selectively slowing down learning on important weights is demonstrated ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/1612.00796", "content": "2 Dec 2016 — A method to prevent catastrophic forgetting in neural networks by selectively slowing down learning on important weights is demonstrated ..."}
+{"idx": 5, "title": "Balanced Training for Neural Machine Translation", "date": "", "ddg_snippet": "by C Shao · 2022 · Cited by 49 — Neural networks tend to gradually forget the previously learned knowledge when learning multiple tasks sequentially from dynamic data.", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2022.acl-long.143.pdf", "content": "by C Shao · 2022 · Cited by 49 — Neural networks tend to gradually forget the previously learned knowledge when learning multiple tasks sequentially from dynamic data."}
+{"idx": 6, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "Abstract . Abstract . The ability to learn tasks in a sequential fashion is crucial to the development of artificial intelligence. Until now neural networks ...", "subpage_snippet": "", "source": "www.jstor.org", "link": "https://www.jstor.org/stable/26480357", "content": "Abstract . Abstract . The ability to learn tasks in a sequential fashion is crucial to the development of artificial intelligence. Until now neural networks ..."}
+{"idx": 7, "title": "Overcoming catastrophic forgetting in neural networks - PMC", "date": "", "ddg_snippet": "by J Kirkpatrick · 2017 · Cited by 10329 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks.", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC5380101/", "content": "by J Kirkpatrick · 2017 · Cited by 10329 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks."}
+{"idx": 8, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "by J Kirkpatrick · 2017 · Cited by 10357 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks.", "subpage_snippet": "", "source": "www.pnas.org", "link": "https://www.pnas.org/doi/abs/10.1073/pnas.1611835114", "content": "by J Kirkpatrick · 2017 · Cited by 10357 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks."}
+{"idx": 9, "title": "Overcoming catastrophic forgetting in neural networks", "date": "", "ddg_snippet": "by J Kirkpatrick · 2017 · Cited by 10362 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks.", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "http://ui.adsabs.harvard.edu/abs/2017PNAS..114.3521K/abstract", "content": "by J Kirkpatrick · 2017 · Cited by 10362 — In this work we propose a practical solution to train such models sequentially by protecting the weights important for previous tasks."}
diff --git a/data/sampled_jsons/Parallel_Simulation_for_Log-concave_Sampling_and_Score-based_Diffusion_Models_deadlock.jsonl b/data/sampled_jsons/Parallel_Simulation_for_Log-concave_Sampling_and_Score-based_Diffusion_Models_deadlock.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a959dc8dfcfe6f588310325c574458c76e6ce5e4
--- /dev/null
+++ b/data/sampled_jsons/Parallel_Simulation_for_Log-concave_Sampling_and_Score-based_Diffusion_Models_deadlock.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Blog – Open3D", "date": "", "ddg_snippet": "We would like to thank all of our community contributors for their true labor of love for this release! Also thanks to the many others who helped the ...", "subpage_snippet": "", "source": "www.open3d.org", "link": "https://www.open3d.org/blog/", "content": "We would like to thank all of our community contributors for their true labor of love for this release! Also thanks to the many others who helped the ..."}
+{"idx": 1, "title": "Parallel Simulation for Log-concave Sampling and Score-based ...", "date": "", "ddg_snippet": "We first present an improved result for parallel sampling from a strongly log - concave and log-smooth distribution. Specifically, we improve the upper bound from eO log2 d ε2 (Anari et al., ε2 .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/attachment?id=qtuxDy2qEB&name=pdf", "content": "We first present an improved result for parallel sampling from a strongly log - concave and log-smooth distribution. Specifically, we improve the upper bound from eO log2 d ε2 (Anari et al., ε2 ."}
+{"idx": 2, "title": "Parallel simulation for sampling under isoperimetry and score ...", "date": "", "ddg_snippet": "Dec 10, 2024 · Inspired by the great success of the parallel simulation of the initial value problem in scientific computation, we propose parallel Picard methods for sampling tasks.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2412.07435", "content": "Dec 10, 2024 · Inspired by the great success of the parallel simulation of the initial value problem in scientific computation, we propose parallel Picard methods for sampling tasks."}
+{"idx": 3, "title": "Parallel Sampling of Diffusion Models - NeurIPS", "date": "", "ddg_snippet": "In spite of the sequential nature of the denoising steps, we show that surprisingly it is possible to parallelize sampling via Picard iterations, by guessing the solution of future denoising steps and iteratively refining until convergence.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/0d1986a61e30e5fa408c81216a616e20-Paper-Conference.pdf", "content": "In spite of the sequential nature of the denoising steps, we show that surprisingly it is possible to parallelize sampling via Picard iterations, by guessing the solution of future denoising steps and iteratively refining until convergence."}
+{"idx": 4, "title": "An Optimization Perspective on Log-concave Sampling - YouTube GitHub Pages - Sinho Chewi's Website Parallel Simulation for Log-concave Sampling and Score-based ... Parallel Sampling of Diffusion Models - NeurIPS Parallel Sampling of Diffusion Models - NeurIPS Parallel Sampling of Diffusion Models - NeurIPS Parallel Sampling of Diffusion Models - NeurIPS Parallel Sampling of Diffusion Models - NeurIPS Parallel Sampling of Diffusion Models - NeurIPS Parallel Simulation for Log-concave Sampling and Score-based...", "date": "", "ddg_snippet": "Computer Science/Discrete Mathematics Seminar II Topic: An Optimization Perspective on Log - concave Sampling ...more. My work focuses on applications of optimal transport to computational problems arising in these fields, such as log - concave sampling (see my book draft above) and variational inference (Lambert et al. (2023); Diao et al. (2023); Jiang, Chewi, and Pooladian (2024)). Spotlight Poster Parallel Simulation for Log - concave Sampling and Score - based Diffusion Models Huanjian Zhou · Masashi Sugiyama East Exhibition Hall A-B #E-1103 [ Abstract ] [ Lay Summary ] [ OpenReview] Thu 17 Jul 4:30 p.m. PDT — 7 p.m. PDT Can parallel sampling be used in a diffusion model? Parallelization techniques similar to Picard iteration have been explored in theoretical works for sampling from log-concave and determinantal distributions . Our work is the first application of parallel sampling on diffusion models , enabling a new axis of trading compute for speed. What is paradigms (parallel diffusion generative model sampling)? 37th Conference on Neural Information Processing Systems (NeurIPS 2023). Our method ParaDiGMS ( Parallel Diffusion Generative Model Sampling ) is the first general method that allows for the tradeoff between compute and sampling speed of pretrained diffusion models . What is the current landscape of sampling from pretrained diffusion models? As a summary, the current landscape of sampling from pretrained diffusion models is comprised of full-step DDPM or accelerated sampling techniques such as DDIM and DPMSolver that trade quality for speed by reducing the number of denoising steps. Can denoising diffusion probabilistic models improve sampling speed? For example, the widely-used Denoising Diffusion Probabilistic Models (DDPMs) can take 1000 denoising steps to generate one sample. In light of this, many works like DDIM and DPMSolver have proposed to improve sampling speed by reducing the number of denoising steps . What are the limitations of diffusion probabilistic models? The main limitation of diffusion models, however, is that sampling can be inconveniently slow . For example, the widely-used Denoising Diffusion Probabilistic Models (DDPMs) can take 1000 denoising steps to generate one sample. Can nave parallelization improve the sample latency of diffusion models? Indeed, naïve parallelization can let us generate multiple samples at once (improve throughput), but generating a single sample with faster wall-clock time (improving latency) appears much more dificult. We show that, surprisingly, it is possible to improve the sample latency of diffusion models by computing denoising steps in parallel. May 1, 2025 · Our research introduces a novel parallel sampling algorithm that significantly reduces the number of sequential steps—known as adaptive complexity—required to generate high-quality samples.", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=_qzJrFQ_B_Y", "content": "Computer Science/Discrete Mathematics Seminar II Topic: An Optimization Perspective on Log - concave Sampling ...more. My work focuses on applications of optimal transport to computational problems arising in these fields, such as log - concave sampling (see my book draft above) and variational inference (Lambert et al. (2023); Diao et al. (2023); Jiang, Chewi, and Pooladian (2024)). Spotlight Poster Parallel Simulation for Log - concave Sampling and Score - based Diffusion Models Huanjian Zhou · Masashi Sugiyama East Exhibition Hall A-B #E-1103 [ Abstract ] [ Lay Summary ] [ OpenReview] Thu 17 Jul 4:30 p.m. PDT — 7 p.m. PDT Can parallel sampling be used in a diffusion model? Parallelization techniques similar to Picard iteration have been explored in theoretical works for sampling from log-concave and determinantal distributions . Our work is the first application of parallel sampling on diffusion models , enabling a new axis of trading compute for speed. What is paradigms (parallel diffusion generative model sampling)? 37th Conference on Neural Information Processing Systems (NeurIPS 2023). Our method ParaDiGMS ( Parallel Diffusion Generative Model Sampling ) is the first general method that allows for the tradeoff between compute and sampling speed of pretrained diffusion models . What is the current landscape of sampling from pretrained diffusion models? As a summary, the current landscape of sampling from pretrained diffusion models is comprised of full-step DDPM or accelerated sampling techniques such as DDIM and DPMSolver that trade quality for speed by reducing the number of denoising steps. Can denoising diffusion probabilistic models improve sampling speed? For example, the widely-used Denoising Diffusion Probabilistic Models (DDPMs) can take 1000 denoising steps to generate one sample. In light of this, many works like DDIM and DPMSolver have proposed to improve sampling speed by reducing the number of denoising steps . What are the limitations of diffusion probabilistic models? The main limitation of diffusion models, however, is that sampling can be inconveniently slow . For example, the widely-used Denoising Diffusion Probabilistic Models (DDPMs) can take 1000 denoising steps to generate one sample. Can nave parallelization improve the sample latency of diffusion models? Indeed, naïve parallelization can let us generate multiple samples at once (improve throughput), but generating a single sample with faster wall-clock time (improving latency) appears much more dificult. We show that, surprisingly, it is possible to improve the sample latency of diffusion models by computing denoising steps in parallel. May 1, 2025 · Our research introduces a novel parallel sampling algorithm that significantly reduces the number of sequential steps—known as adaptive complexity—required to generate high-quality samples."}
+{"idx": 5, "title": "Parallel Simulation for Log-concave Sampling and Score-based ...", "date": "", "ddg_snippet": "Spotlight Poster Parallel Simulation for Log - concave Sampling and Score - based Diffusion Models Huanjian Zhou · Masashi Sugiyama East Exhibition Hall A-B #E-1103 [ Abstract ] [ Lay Summary ] [ OpenReview] Thu 17 Jul 4:30 p.m. PDT — 7 p.m. PDT", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/43916", "content": "Spotlight Poster Parallel Simulation for Log - concave Sampling and Score - based Diffusion Models Huanjian Zhou · Masashi Sugiyama East Exhibition Hall A-B #E-1103 [ Abstract ] [ Lay Summary ] [ OpenReview] Thu 17 Jul 4:30 p.m. PDT — 7 p.m. PDT"}
+{"idx": 6, "title": "Parallel Simulation for Log-concave Sampling and Score-based...", "date": "", "ddg_snippet": "May 1, 2025 · Our research introduces a novel parallel sampling algorithm that significantly reduces the number of sequential steps—known as adaptive complexity—required to generate high-quality samples.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=qtuxDy2qEB", "content": "May 1, 2025 · Our research introduces a novel parallel sampling algorithm that significantly reduces the number of sequential steps—known as adaptive complexity—required to generate high-quality samples."}
+{"idx": 7, "title": "Parallel Simulation for Log - concave Sampling and Score - based ...", "date": "", "ddg_snippet": "TL;DR: We propose new parallel algorithms for log - concave sampling and diffusion models , and rigorously prove that our algorithms enjoy $\\log d$ iteration complexity.Primary Area: Probabilistic Methods->Monte Carlo and Sampling Methods.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=qtuxDy2qEB&referrer=[the+profile+of+Masashi+Sugiyama](/profile?id=~Masashi_Sugiyama1)", "content": "TL;DR: We propose new parallel algorithms for log - concave sampling and diffusion models , and rigorously prove that our algorithms enjoy $\\log d$ iteration complexity.Primary Area: Probabilistic Methods->Monte Carlo and Sampling Methods."}
+{"idx": 8, "title": "Open3D – Open3D", "date": "", "ddg_snippet": "Open3D had a successful Google Season of Code 2022 with many new features added, and more in the works for the next release.", "subpage_snippet": "", "source": "www.open3d.org", "link": "https://www.open3d.org/author/administratorivcl-org/", "content": "Open3D had a successful Google Season of Code 2022 with many new features added, and more in the works for the next release."}
+{"idx": 9, "title": "GitHub Pages - Sinho Chewi's Website", "date": "", "ddg_snippet": "My work focuses on applications of optimal transport to computational problems arising in these fields, such as log - concave sampling (see my book draft above) and variational inference (Lambert et al. (2023); Diao et al. (2023); Jiang, Chewi, and Pooladian (2024)).", "subpage_snippet": "", "source": "chewisinho.github.io", "link": "https://chewisinho.github.io/", "content": "My work focuses on applications of optimal transport to computational problems arising in these fields, such as log - concave sampling (see my book draft above) and variational inference (Lambert et al. (2023); Diao et al. (2023); Jiang, Chewi, and Pooladian (2024))."}
diff --git a/data/sampled_jsons/Pathak_et_al._2022_FourCastNet_evaluation_metrics_year_2022.jsonl b/data/sampled_jsons/Pathak_et_al._2022_FourCastNet_evaluation_metrics_year_2022.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..928cdcb1862848e66be1202f05f056cc8ad5b354
--- /dev/null
+++ b/data/sampled_jsons/Pathak_et_al._2022_FourCastNet_evaluation_metrics_year_2022.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2202.11214] FourCastNet : A Global Data-driven High-resolution...", "date": "", "ddg_snippet": "Abstract: FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at $0.25^{\\circ}$ resolution. FourCastNet accurately forecasts high-resolution, fast-timescale variables...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2202.11214", "content": "Abstract: FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at $0.25^{\\circ}$ resolution. FourCastNet accurately forecasts high-resolution, fast-timescale variables..."}
+{"idx": 1, "title": "GitHub - NVlabs/ FourCastNet : Initial public release of code, data, and...", "date": "", "ddg_snippet": "FourCastNet is based on the vision transformer architecture with Adaptive Fourier Neural Operator (AFNO) attention proposed in Guibas-Mardani et al . [paper], [code]. Total Column of Water Vapor forecast using FourCastNet .Initial public release of FourCastNet Latest. Aug 6, 2022 .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/NVlabs/FourCastNet", "content": "FourCastNet is based on the vision transformer architecture with Adaptive Fourier Neural Operator (AFNO) attention proposed in Guibas-Mardani et al . [paper], [code]. Total Column of Water Vapor forecast using FourCastNet .Initial public release of FourCastNet Latest. Aug 6, 2022 ."}
+{"idx": 2, "title": "FourCastNet : A Global Data-driven High-resolution Weather Model...", "date": "", "ddg_snippet": "... Recently, the Neural Earth System Model (NESM) called FourCastNet ( Pathak et al ., 2022 b), trained on the ECMWF Reanalysis v5 (ERA5) data set, has shown impressive results on predicting various atmospheric variables globally at 0.25 degrees resolution.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/358814796_FourCastNet_A_Global_Data-driven_High-resolution_Weather_Model_using_Adaptive_Fourier_Neural_Operators", "content": "... Recently, the Neural Earth System Model (NESM) called FourCastNet ( Pathak et al ., 2022 b), trained on the ECMWF Reanalysis v5 (ERA5) data set, has shown impressive results on predicting various atmospheric variables globally at 0.25 degrees resolution."}
+{"idx": 3, "title": "FourCastNet : A Global Data-driven High-resolution... | BibSonomy", "date": "", "ddg_snippet": "pathak _ fourcastnet _ 2022 . entry type. FourCastNet . file. arXiv Fulltext PDF:/Users/pascal/Zotero/storage/8IFX2TBW/ Pathak et al .", "subpage_snippet": "", "source": "www.bibsonomy.org", "link": "https://www.bibsonomy.org/bibtex/27c725e2908978d066179e178c3bc9a61/annakrause", "content": "pathak _ fourcastnet _ 2022 . entry type. FourCastNet . file. arXiv Fulltext PDF:/Users/pascal/Zotero/storage/8IFX2TBW/ Pathak et al ."}
+{"idx": 4, "title": "FourCastNet : A Global Data-driven High-resolution Weather... | DeepAI", "date": "", "ddg_snippet": "by Jaideep Pathak , et al . FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at 0.25^∘ resolution.", "subpage_snippet": "", "source": "deepai.org", "link": "https://deepai.org/publication/fourcastnet-a-global-data-driven-high-resolution-weather-model-using-adaptive-fourier-neural-operators", "content": "by Jaideep Pathak , et al . FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at 0.25^∘ resolution."}
+{"idx": 5, "title": "arXiv Daily on Twitter: \" FourCastNet : A Global Data-driven...\"", "date": "", "ddg_snippet": "02/22/22 - FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurat... 8:08 PM · Feb 25, 2022 ·ArXiv Daily.", "subpage_snippet": "", "source": "twitter.com", "link": "https://twitter.com/arXiv_Daily/status/1497302606858711042", "content": "02/22/22 - FourCastNet , short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurat... 8:08 PM · Feb 25, 2022 ·ArXiv Daily."}
+{"idx": 6, "title": "MariNet Model in Comparision to FourCastNet and PhyDNet", "date": "", "ddg_snippet": "190 According to ( Pathak et al ., 2022 ) the use of ViT backbone is preferred due to its ability 191 to effectively model long-range dependencies.249 MariNet model shows promising results. The figures representing metrics of FourCastNet 250 model have artifacts.", "subpage_snippet": "", "source": "www.authorea.com", "link": "https://www.authorea.com/users/715258/articles/699698/master/file/data/983501_0_merged_1704924617/983501_0_merged_1704924617.pdf", "content": "190 According to ( Pathak et al ., 2022 ) the use of ViT backbone is preferred due to its ability 191 to effectively model long-range dependencies.249 MariNet model shows promising results. The figures representing metrics of FourCastNet 250 model have artifacts."}
+{"idx": 7, "title": "Using Machine Learning to Forecast the Weather... | Climate Change AI", "date": "", "ddg_snippet": "FourCastNet : A practical introduction to a state-of-the-art deep learning global weather emulator from Pathak et al ., 2022 .ClimateLearn: Machine Learning for Predicting Weather and Climate Extremes from Nguyen et al .", "subpage_snippet": "", "source": "www.climatechange.ai", "link": "https://www.climatechange.ai/blog/2024-02-07-forecast-tutorials", "content": "FourCastNet : A practical introduction to a state-of-the-art deep learning global weather emulator from Pathak et al ., 2022 .ClimateLearn: Machine Learning for Predicting Weather and Climate Extremes from Nguyen et al ."}
+{"idx": 8, "title": "Frontiers | Optimizing data-driven arctic marine forecasting...", "date": "", "ddg_snippet": "According to ( Pathak et al ., 2022 ), the FourCastNet uses such metrics as Root Mean Squared Error (RMSE), Anomaly Correlation Coefficient (ACC) at lead times of up to three days and gives results comparable to the ECMWF Integrated Forecasting System (IFS)...", "subpage_snippet": "", "source": "www.frontiersin.org", "link": "https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2024.1456480/full", "content": "According to ( Pathak et al ., 2022 ), the FourCastNet uses such metrics as Root Mean Squared Error (RMSE), Anomaly Correlation Coefficient (ACC) at lead times of up to three days and gives results comparable to the ECMWF Integrated Forecasting System (IFS)..."}
+{"idx": 9, "title": "SwinRDM: Integrate SwinRNN with Diffusion Model towards...", "date": "", "ddg_snippet": "FourCastNet ( Pathak et al . 2022 ) is the first work that directly builds the network on the 0.25 resolution data. Evaluation Metrics We follow (Rasp et al . 2020) to eval - uate forecast quality using latitude-weighted RMSE (root-mean-square error).", "subpage_snippet": "", "source": "ojs.aaai.org", "link": "https://ojs.aaai.org/index.php/AAAI/article/download/25105/24877", "content": "FourCastNet ( Pathak et al . 2022 ) is the first work that directly builds the network on the 0.25 resolution data. Evaluation Metrics We follow (Rasp et al . 2020) to eval - uate forecast quality using latitude-weighted RMSE (root-mean-square error)."}
diff --git a/data/sampled_jsons/Program_Synthesis_for_Schubert_Polynomials_n_=_4,_5_spurious_correlation_0.98.jsonl b/data/sampled_jsons/Program_Synthesis_for_Schubert_Polynomials_n_=_4,_5_spurious_correlation_0.98.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b06b85344ddb6489887b097fd4b27c68be9dfed8
--- /dev/null
+++ b/data/sampled_jsons/Program_Synthesis_for_Schubert_Polynomials_n_=_4,_5_spurious_correlation_0.98.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Tools and Algorithms for the Construction and Analysis ...", "date": "", "ddg_snippet": "The conferences cover various aspects of software systems, ranging from theo- retical computer science to foundations to programming language developments,.", "subpage_snippet": "", "source": "lumetta.web.engr.illinois.edu", "link": "https://lumetta.web.engr.illinois.edu/508-F21/project-prep/TACAS_2019_Proceedings.pdf", "content": "The conferences cover various aspects of software systems, ranging from theo- retical computer science to foundations to programming language developments,."}
+{"idx": 1, "title": "SAT-based Answer Set Programming", "date": "", "ddg_snippet": "by Y Lierler · 2010 · Cited by 14 — Chapter 4 defines traditional logic programs and introduces the notions of completion and tightness that form the foundation of this work. Chapter 5 defines.", "subpage_snippet": "", "source": "digitalcommons.unomaha.edu", "link": "https://digitalcommons.unomaha.edu/cgi/viewcontent.cgi?article=1005&context=compscifacpub", "content": "by Y Lierler · 2010 · Cited by 14 — Chapter 4 defines traditional logic programs and introduces the notions of completion and tightness that form the foundation of this work. Chapter 5 defines."}
+{"idx": 2, "title": "Unsupervised Learning Methods for Molecular Simulation Data", "date": "", "ddg_snippet": "by A Glielmo · 2021 · Cited by 397 — In this Review, we provide a comprehensive overview of the methods of unsupervised learning that have been most commonly used to investigate simulation data.", "subpage_snippet": "", "source": "pubs.acs.org", "link": "https://pubs.acs.org/doi/10.1021/acs.chemrev.0c01195", "content": "by A Glielmo · 2021 · Cited by 397 — In this Review, we provide a comprehensive overview of the methods of unsupervised learning that have been most commonly used to investigate simulation data."}
+{"idx": 3, "title": "Machine Learning meets Algebraic Combinatorics: A Suite of Datasets...", "date": "", "ddg_snippet": "Program Synthesis for Schubert Polynomials : The pre-vious example highlighted use of narrow models for con-jecture generation.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.06366", "content": "Program Synthesis for Schubert Polynomials : The pre-vious example highlighted use of narrow models for con-jecture generation."}
+{"idx": 4, "title": "Machine Learning meets Algebraic Combinatorics: A Suite of Datasets...", "date": "", "ddg_snippet": "Program Synthesis for Schubert Polynomials : The previous example highlighted use of narrow models for conjecture generation.), this process introduced a spurious correlation in the dataset222This issue has since been fixed.: structure constant.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.06366v1", "content": "Program Synthesis for Schubert Polynomials : The previous example highlighted use of narrow models for conjecture generation.), this process introduced a spurious correlation in the dataset222This issue has since been fixed.: structure constant."}
+{"idx": 5, "title": "Causal Machine Learning: A Survey and Open Problems", "date": "", "ddg_snippet": "by J Kaddour · 2022 · Cited by 260 — Causal Machine Learning (CausalML) is an umbrella term for machine learning methods that formalize the data-generation process as a ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2206.15475", "content": "by J Kaddour · 2022 · Cited by 260 — Causal Machine Learning (CausalML) is an umbrella term for machine learning methods that formalize the data-generation process as a ..."}
+{"idx": 6, "title": "polynomial regression model: Topics by ...", "date": "", "ddg_snippet": "The objective of this study was to identify the best random regression model using Legendre orthogonal polynomials to evaluate Alpine goats genetically and to ...", "subpage_snippet": "", "source": "www.science.gov", "link": "https://www.science.gov/topicpages/p/polynomial+regression+model.html", "content": "The objective of this study was to identify the best random regression model using Legendre orthogonal polynomials to evaluate Alpine goats genetically and to ..."}
+{"idx": 7, "title": "In Silico Prediction of Molecular Volumes, Heat Capacities ...", "date": "", "ddg_snippet": "Overall, for every method, the correlation between Vc and Vion is very high, with a correlation coefficient (r 2 ) > 0.98 . Hofmann's volumes, either with or ...", "subpage_snippet": "", "source": "pubs.acs.org", "link": "https://pubs.acs.org/doi/10.1021/ie801268a", "content": "Overall, for every method, the correlation between Vc and Vion is very high, with a correlation coefficient (r 2 ) > 0.98 . Hofmann's volumes, either with or ..."}
+{"idx": 8, "title": "Nonlinear dynamics of cardiovascular ageing - PMC", "date": "", "ddg_snippet": "by Y Shiogai · 2010 · Cited by 380 — The application of methods drawn from nonlinear and stochastic dynamics to the analysis of cardiovascular time series is reviewed, with particular reference ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC2853263/", "content": "by Y Shiogai · 2010 · Cited by 380 — The application of methods drawn from nonlinear and stochastic dynamics to the analysis of cardiovascular time series is reviewed, with particular reference ..."}
+{"idx": 9, "title": "EXPERIMENTAL INVESTIGATION OF A THREE- ...", "date": "", "ddg_snippet": "by BO Kolade · 2010 · Cited by 4 — High levels of the Reynolds stresses develop in the separated shear layer due to the adverse pressure gradient, as an inflection point is generated in the mean ...", "subpage_snippet": "", "source": "web.stanford.edu", "link": "https://web.stanford.edu/group/tfsa/TF_reports/TF-124_Kolade.pdf", "content": "by BO Kolade · 2010 · Cited by 4 — High levels of the Reynolds stresses develop in the separated shear layer due to the adverse pressure gradient, as an inflection point is generated in the mean ..."}
diff --git a/data/sampled_jsons/Proximal_Policy_Optimization_Schulman_2017_A2C_performance_table_results_data_Atari_games_year_2017.jsonl b/data/sampled_jsons/Proximal_Policy_Optimization_Schulman_2017_A2C_performance_table_results_data_Atari_games_year_2017.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..82150fd7d088961b19d4c97f66a44813a75e59a0
--- /dev/null
+++ b/data/sampled_jsons/Proximal_Policy_Optimization_Schulman_2017_A2C_performance_table_results_data_Atari_games_year_2017.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Proximal Policy Optimization Algorithms", "date": "", "ddg_snippet": "Proximal Policy Optimization Algorithms. John Schulman , Filip Wolski, Prafulla Dhariwal, Alec Radford, Oleg Klimov OpenAI.On Atari , it performs signicantly better (in terms of sample complexity) than A 2 C and similarly to ACER though it is much simpler.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1707.06347", "content": "Proximal Policy Optimization Algorithms. John Schulman , Filip Wolski, Prafulla Dhariwal, Alec Radford, Oleg Klimov OpenAI.On Atari , it performs signicantly better (in terms of sample complexity) than A 2 C and similarly to ACER though it is much simpler."}
+{"idx": 1, "title": "Proximal Policy Optimization Algorithms, Schulman et al, 2017 | PDF", "date": "", "ddg_snippet": "The document discusses Proximal Policy Optimization (PPO), a reinforcement learning algorithm introduced by Schulman et al. in 2017 , which aims to improve the performance and data efficiency of existing methods like TRPO while being simpler to implement.", "subpage_snippet": "", "source": "www.slideshare.net", "link": "https://www.slideshare.net/slideshow/proximal-policy-optimization-algorithms-schulman-et-al-2017/238343413", "content": "The document discusses Proximal Policy Optimization (PPO), a reinforcement learning algorithm introduced by Schulman et al. in 2017 , which aims to improve the performance and data efficiency of existing methods like TRPO while being simpler to implement."}
+{"idx": 2, "title": "Understanding Proximal Policy Optimization ( Schulman et al., 2017 )", "date": "", "ddg_snippet": "Policy Loss Function ( Schulman et al., 2017 ). The policy pi is our neural network that takes the state observation from an environment as input and suggests actions to take as an output.", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/data-science/understanding-and-implementing-proximal-policy-optimization-schulman-et-al-2017-9523078521ce", "content": "Policy Loss Function ( Schulman et al., 2017 ). The policy pi is our neural network that takes the state observation from an environment as input and suggests actions to take as an output."}
+{"idx": 3, "title": "Proximal Policy Optimization | OpenAI", "date": "", "ddg_snippet": "Proximal Policy Optimization | OpenAI. Table of contents.In tests, this algorithm has displayed the best performance on continuous control tasks and almost matches ACER’s performance on Atari , despite being far simpler to implement.", "subpage_snippet": "", "source": "openai.com", "link": "https://openai.com/index/openai-baselines-ppo/", "content": "Proximal Policy Optimization | OpenAI. Table of contents.In tests, this algorithm has displayed the best performance on continuous control tasks and almost matches ACER’s performance on Atari , despite being far simpler to implement."}
+{"idx": 4, "title": "Understanding and Implementing Proximal Policy Optimization ...", "date": "", "ddg_snippet": "Research in policy gradient methods has been prevalent in recent years, with algorithms such as TRPO, GAE, and A 2 C /A3C showing state-of-the-art performance over traditional methods such as Q-learning. One of the core algorithms in this policy gradient/actor-critic field is Proximal ...", "subpage_snippet": "", "source": "dev.to", "link": "https://dev.to/tylertaewook/understanding-and-implementing-proximal-policy-optimization-schulman-et-al-2017-20on", "content": "Research in policy gradient methods has been prevalent in recent years, with algorithms such as TRPO, GAE, and A 2 C /A3C showing state-of-the-art performance over traditional methods such as Q-learning. One of the core algorithms in this policy gradient/actor-critic field is Proximal ..."}
+{"idx": 5, "title": "Trust Region Policy Optimization (TRPO) and Proximal Policy ...", "date": "", "ddg_snippet": "Proximal Policy optimization : Clipping Objective. Form a lower bound via clipped importance ratio.Combines KFAC with A 2 C , which gives excellent results on Atari benchmark and continuous control from images. Basically, plugs in KFAC natural gradient in A 2 C .", "subpage_snippet": "", "source": "readmedium.com", "link": "https://readmedium.com/trust-region-policy-optimization-trpo-and-proximal-policy-optimization-ppo-e6e7075f39ed", "content": "Proximal Policy optimization : Clipping Objective. Form a lower bound via clipped importance ratio.Combines KFAC with A 2 C , which gives excellent results on Atari benchmark and continuous control from images. Basically, plugs in KFAC natural gradient in A 2 C ."}
+{"idx": 6, "title": "Proximal Policy Optimization Algorithms", "date": "", "ddg_snippet": "Proximal Policy Optimization Algorithms (1707.06347v2). Published 20 Jul 2017 in cs.LG.These methods, collectively termed Proximal Policy Optimization ° (PPO °), aim to balance simplicity in implementation, generality, and improved sample efficiency.", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/papers/1707.06347", "content": "Proximal Policy Optimization Algorithms (1707.06347v2). Published 20 Jul 2017 in cs.LG.These methods, collectively termed Proximal Policy Optimization ° (PPO °), aim to balance simplicity in implementation, generality, and improved sample efficiency."}
+{"idx": 7, "title": "A Proximal Policy Optimization Algorithm for", "date": "", "ddg_snippet": "One of the algorithms being researched is Proximal Policy Optimization (PPO). ICT Group chose to research PPO because PPO is proven to be useful for complex problems and deals well with large state and action sizes, which will be further dis-cussed in the literature review.", "subpage_snippet": "", "source": "arno.uvt.nl", "link": "https://arno.uvt.nl/show.cgi?fid=169254", "content": "One of the algorithms being researched is Proximal Policy Optimization (PPO). ICT Group chose to research PPO because PPO is proven to be useful for complex problems and deals well with large state and action sizes, which will be further dis-cussed in the literature review."}
+{"idx": 8, "title": "Proximal Policy Optimization Algorithms. | BibSonomy", "date": "", "ddg_snippet": "Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates.Links and resources. BibTeX key. schulman 2017 ppo.", "subpage_snippet": "", "source": "www.bibsonomy.org", "link": "https://www.bibsonomy.org/bibtex/24bbcce6aa1c42ae7f61ef8cf5475aa85/lanteunis", "content": "Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates.Links and resources. BibTeX key. schulman 2017 ppo."}
+{"idx": 9, "title": "proximal policy optimization – Water Programming: A Collaborative...", "date": "", "ddg_snippet": "Posts about proximal policy optimization written by Andrew Hamilton and Lillian Lau.PPO with a clipped surrogate objective: This method is currently the more widely used version of PPO as it has been found to perform better than the former ( Schulman et al, 2017 ; van Heeswijk, 2022).", "subpage_snippet": "", "source": "waterprogramming.wordpress.com", "link": "https://waterprogramming.wordpress.com/tag/proximal-policy-optimization/", "content": "Posts about proximal policy optimization written by Andrew Hamilton and Lillian Lau.PPO with a clipped surrogate objective: This method is currently the more widely used version of PPO as it has been found to perform better than the former ( Schulman et al, 2017 ; van Heeswijk, 2022)."}
diff --git a/data/sampled_jsons/Qwen2-VL_58.5_61.2_OR_Qwen2-VL_Avg_zh_en_XLRS-Bench.jsonl b/data/sampled_jsons/Qwen2-VL_58.5_61.2_OR_Qwen2-VL_Avg_zh_en_XLRS-Bench.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bdaa3baef26fa6f8ace31fb260becc669bfd7e8d
--- /dev/null
+++ b/data/sampled_jsons/Qwen2-VL_58.5_61.2_OR_Qwen2-VL_Avg_zh_en_XLRS-Bench.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "TemporalBench", "date": "", "ddg_snippet": "In step 2 , we leverage existing LLMs to ... Human performance is much better than GPT-4o, Qwen2 - VL -72B, LLaVA-OneVision-72B, and Gemini-1. 5 -Pro.", "subpage_snippet": "", "source": "temporalbench.github.io", "link": "https://temporalbench.github.io/", "content": "In step 2 , we leverage existing LLMs to ... Human performance is much better than GPT-4o, Qwen2 - VL -72B, LLaVA-OneVision-72B, and Gemini-1. 5 -Pro."}
+{"idx": 1, "title": "MMSearch: Benchmarking the potential of large models as", "date": "", "ddg_snippet": "... LMMs such as GPT-4o or Claude-3. 5 ... Response and middle results comparison of GPT-4o, Qwen2 - VL -7B, and LLaVA-OneVision-7B in the end-to-end task.", "subpage_snippet": "", "source": "mmsearch.github.io", "link": "https://mmsearch.github.io/", "content": "... LMMs such as GPT-4o or Claude-3. 5 ... Response and middle results comparison of GPT-4o, Qwen2 - VL -7B, and LLaVA-OneVision-7B in the end-to-end task."}
+{"idx": 2, "title": "Agent-RewardBench: Towards a Unified Benchmark for Reward", "date": "", "ddg_snippet": "... 4o follows with 61 .4%, and Claude-3. 5 ... A stronger model like GPT-4o achieves only 39. 2 % accuracy in safety, while Claude 3. 5 Sonnet scores 22.4%.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.21252v1", "content": "... 4o follows with 61 .4%, and Claude-3. 5 ... A stronger model like GPT-4o achieves only 39. 2 % accuracy in safety, while Claude 3. 5 Sonnet scores 22.4%."}
+{"idx": 3, "title": "CII-Bench", "date": "", "ddg_snippet": "As the capabilities of Multimodal Large Language Models (MLLMs) continue to improve, the need for higher-order capability evaluation of MLLMs is ...", "subpage_snippet": "", "source": "cii-bench.github.io", "link": "https://cii-bench.github.io/", "content": "As the capabilities of Multimodal Large Language Models (MLLMs) continue to improve, the need for higher-order capability evaluation of MLLMs is ..."}
+{"idx": 4, "title": "StreamBridge: Turning Your Offline Video Large Language Model", "date": "", "ddg_snippet": "... fine-tuning on Stream-IT , we successfully convert several leading offline Video-LLMs, including LLaVA-OV [ 3 ] , Oryx-1. 5 [ 1 ] , and Qwen2 - VL [ 2 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.05467v2", "content": "... fine-tuning on Stream-IT , we successfully convert several leading offline Video-LLMs, including LLaVA-OV [ 3 ] , Oryx-1. 5 [ 1 ] , and Qwen2 - VL [ 2 ..."}
+{"idx": 5, "title": "InternVL3", "date": "", "ddg_snippet": "Additionally, we compare InternVL3 with Qwen2 . 5 Chat models, whose corresponding pre-trained base models are employed as the initialization of the ...", "subpage_snippet": "", "source": "internvl.github.io", "link": "https://internvl.github.io/blog/2025-04-11-InternVL-3.0/", "content": "Additionally, we compare InternVL3 with Qwen2 . 5 Chat models, whose corresponding pre-trained base models are employed as the initialization of the ..."}
+{"idx": 6, "title": "MMG-Vid: Maximizing Marginal Gains at Segment-level and", "date": "", "ddg_snippet": "... 75% of video tokens, achieving a 3.9x speedup in the prefilling stage, all while maintaining 99. 5 % of the model’s original performance on average.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.21044v1", "content": "... 75% of video tokens, achieving a 3.9x speedup in the prefilling stage, all while maintaining 99. 5 % of the model’s original performance on average."}
+{"idx": 7, "title": "FastDriveVLA: Efficient End-to-End Driving via Plug-and-Play", "date": "", "ddg_snippet": "Current visual token pruning methods in Vision-Language Models (VLM) rely on either visual token similarity or visual-text attention, but both have ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.23318v1", "content": "Current visual token pruning methods in Vision-Language Models (VLM) rely on either visual token similarity or visual-text attention, but both have ..."}
+{"idx": 8, "title": "SemiETS: Integrating Spatial and Content Consistencies for", "date": "", "ddg_snippet": "The former is mostly the modification of SSOD frameworks to fit the irregular shape of texts [ 28 ] or improve label selection [ 52 , 37 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2504.09966v1", "content": "The former is mostly the modification of SSOD frameworks to fit the irregular shape of texts [ 28 ] or improve label selection [ 52 , 37 ..."}
+{"idx": 9, "title": "MixEval-X", "date": "", "ddg_snippet": "We identify two major issues in current evalua- tions: (1) inconsistent standards, shaped by different communities with varying protocols and ...", "subpage_snippet": "", "source": "mixeval-x.github.io", "link": "https://mixeval-x.github.io/", "content": "We identify two major issues in current evalua- tions: (1) inconsistent standards, shaped by different communities with varying protocols and ..."}
diff --git a/data/sampled_jsons/RA-PbRL_paper_challenges_risk_measures_preference-based_reinforcement_learning_quantile_estimation.jsonl b/data/sampled_jsons/RA-PbRL_paper_challenges_risk_measures_preference-based_reinforcement_learning_quantile_estimation.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6e2556bf6f791b4c1a697f15c0a25f9ac96e7812
--- /dev/null
+++ b/data/sampled_jsons/RA-PbRL_paper_challenges_risk_measures_preference-based_reinforcement_learning_quantile_estimation.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "Oct 31, 2024 · Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators. Despite ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.23569", "content": "Oct 31, 2024 · Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators. Despite ..."}
+{"idx": 1, "title": "RA-PbRL | Proceedings of the 38th International Conference on ...", "date": "", "ddg_snippet": "Jun 5, 2025 · Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators. Despite ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/abs/10.5555/3737916.3739861", "content": "Jun 5, 2025 · Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators. Despite ..."}
+{"idx": 2, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "Authors Yujie Zhao, Jose Efraim Aguilar Escamill, Weyl Lu, Huazheng Wang Abstract Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences ...", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2024/hash/7016d7b7b6e3c05b2128ac5b3aae492d-Abstract-Conference.html", "content": "Authors Yujie Zhao, Jose Efraim Aguilar Escamill, Weyl Lu, Huazheng Wang Abstract Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences ..."}
+{"idx": 3, "title": "Advances in Preference-based Reinforcement Learning: A Review", "date": "", "ddg_snippet": "Reinforcement Learning (RL) algorithms suffer from the dependency on accurately engineered reward functions to properly guide the learning agents to do the required tasks. Preference-based reinforcement learning ( PbRL ) addresses that by utilizing human preferences as feedback from the experts instead of numeric rewards. Due to its promising advantage over traditional RL, PbRL has gained more ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/9945333", "content": "Reinforcement Learning (RL) algorithms suffer from the dependency on accurately engineered reward functions to properly guide the learning agents to do the required tasks. Preference-based reinforcement learning ( PbRL ) addresses that by utilizing human preferences as feedback from the experts instead of numeric rewards. Due to its promising advantage over traditional RL, PbRL has gained more ..."}
+{"idx": 4, "title": "Risk-Aware Preference-baser Reinforcement Learning (RA-PbRL)", "date": "", "ddg_snippet": "RA-PbRL is a type of Policy-Iteration and \"Confidence Bound\" reinforcement learning algorithm designed for preference-based reinforcement learning while maximizing risk -awareness through Value-at- Risk penalties. The intuition behind the algorithm depends on the idea of confidence bounds. The ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/aguilarjose11/PbRLNeurips", "content": "RA-PbRL is a type of Policy-Iteration and \"Confidence Bound\" reinforcement learning algorithm designed for preference-based reinforcement learning while maximizing risk -awareness through Value-at- Risk penalties. The intuition behind the algorithm depends on the idea of confidence bounds. The ..."}
+{"idx": 5, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based...", "date": "", "ddg_snippet": "Sep 25, 2024 · Abstract: Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=JNDcFOczOf", "content": "Sep 25, 2024 · Abstract: Reinforcement Learning from Human Feedback (RLHF) has recently surged in popularity, particularly for aligning large language models and other AI systems with human intentions. At its core, RLHF can be viewed as a specialized instance of Preference-based Reinforcement Learning ( PbRL ), where the preferences specifically originate from human judgments rather than arbitrary evaluators ..."}
+{"idx": 6, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "Sep 26, 2024 · In preference-based reinforcement learning ( PbRL ), where rewards are implicitly defined via preferences over trajectories, quantile risk offers a powerful tool to ensure safer and more robust behavior by focusing on minimizing the risk associated with the lower quantiles of the trajectory reward distribution.", "subpage_snippet": "", "source": "deep-diver.github.io", "link": "https://deep-diver.github.io/neurips2024/posters/jndcfoczof/", "content": "Sep 26, 2024 · In preference-based reinforcement learning ( PbRL ), where rewards are implicitly defined via preferences over trajectories, quantile risk offers a powerful tool to ensure safer and more robust behavior by focusing on minimizing the risk associated with the lower quantiles of the trajectory reward distribution."}
+{"idx": 7, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based...", "date": "", "ddg_snippet": "This paper studies preference - based RL ( PbRL ) where instead of the expected return, the agent optimizes a risk measure based on preference feedback.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=JNDcFOczOf&referrer=[the+profile+of+Huazheng+Wang](/profile?id=~Huazheng_Wang1)", "content": "This paper studies preference - based RL ( PbRL ) where instead of the expected return, the agent optimizes a risk measure based on preference feedback."}
+{"idx": 8, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "To address this, we explore and prove the applicability of two risk -aware objectives to PbRL : nested and static quantile risk objectives. We also introduce Risk ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.23569v1", "content": "To address this, we explore and prove the applicability of two risk -aware objectives to PbRL : nested and static quantile risk objectives. We also introduce Risk ..."}
+{"idx": 9, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "9 Dec 2024 — We explore static and nested measures to introduce risk awareness to PbRL settings. To the best of our knowledge, our proposed RA - PbRL algorithm ...", "subpage_snippet": "", "source": "nips.cc", "link": "https://nips.cc/virtual/2024/poster/95716", "content": "9 Dec 2024 — We explore static and nested measures to introduce risk awareness to PbRL settings. To the best of our knowledge, our proposed RA - PbRL algorithm ..."}
diff --git a/data/sampled_jsons/RAGGED_framework_BM25_DPR_Contriever_retriever_paradigms.jsonl b/data/sampled_jsons/RAGGED_framework_BM25_DPR_Contriever_retriever_paradigms.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d4b9e386d4d61009dc15245c9cc0840222ff5a3a
--- /dev/null
+++ b/data/sampled_jsons/RAGGED_framework_BM25_DPR_Contriever_retriever_paradigms.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Database-Augmented Query Representation for Information", "date": "", "ddg_snippet": "Motivated by this fact, we present a novel retrieval framework called Database-Augmented Query representation (DAQu), which augments the original ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2406.16013v3", "content": "Motivated by this fact, we present a novel retrieval framework called Database-Augmented Query representation (DAQu), which augments the original ..."}
+{"idx": 1, "title": "Retrieval-Augmented Generation: Is Dense Passage Retrieval", "date": "", "ddg_snippet": "Dense passage retrieval ( DPR ) is the first step in the retrieval augmented generation (RAG) paradigm for improving the performance of large language ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2402.11035v1", "content": "Dense passage retrieval ( DPR ) is the first step in the retrieval augmented generation (RAG) paradigm for improving the performance of large language ..."}
+{"idx": 2, "title": "The Hidden Costs of Naive Retrieval: Adaptive RAG, Part 1 -", "date": "", "ddg_snippet": "... retriever can recall relevant documents using sparse vector-based methods such as TFIDF and BM25 , or dense-vector-based approaches like DPR and ...", "subpage_snippet": "", "source": "blog.reachsumit.com", "link": "https://blog.reachsumit.com/posts/2025/09/problems-with-naive-rag/", "content": "... retriever can recall relevant documents using sparse vector-based methods such as TFIDF and BM25 , or dense-vector-based approaches like DPR and ..."}
+{"idx": 3, "title": "Vinija's Notes • NLP • Retrieval Augmented Generation", "date": "", "ddg_snippet": "It leverages a retriever to find relevant contexts to condition the LLM, in this way, RAG is able to augment the knowledge-base of an LLM with ...", "subpage_snippet": "", "source": "vinija.ai", "link": "https://vinija.ai/nlp/RAG/", "content": "It leverages a retriever to find relevant contexts to condition the LLM, in this way, RAG is able to augment the knowledge-base of an LLM with ..."}
+{"idx": 4, "title": "InfoGain-RAG: Boosting Retrieval-Augmented Generation via", "date": "", "ddg_snippet": "However, current RAG frameworks often struggle with identifying whether retrieved documents meaningfully contribute to answer generation.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.12765v1", "content": "However, current RAG frameworks often struggle with identifying whether retrieved documents meaningfully contribute to answer generation."}
+{"idx": 5, "title": "Planning and Editing What You Retrieve for Enhanced Tool", "date": "", "ddg_snippet": "Experiments show that our paradigms improve the recall and NDCG of tool retrieval tasks, significantly outperforming current state-of-the-art (SOTA).", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2404.00450v2", "content": "Experiments show that our paradigms improve the recall and NDCG of tool retrieval tasks, significantly outperforming current state-of-the-art (SOTA)."}
+{"idx": 6, "title": "Retrieval-Augmented Generation (RAG): Recent Research and", "date": "", "ddg_snippet": "In today ’ s AI-driven world, Retrieval -Augmented Generation (RAG) is becoming an increasingly significant approach that combines the ...", "subpage_snippet": "", "source": "alimbekov.com", "link": "https://alimbekov.com/en/retrieval-augmented-generation-rag-recent-research-and-challenges/", "content": "In today ’ s AI-driven world, Retrieval -Augmented Generation (RAG) is becoming an increasingly significant approach that combines the ..."}
+{"idx": 7, "title": "Aman's AI Journal • NLP • Retrieval Augmented Generation", "date": "", "ddg_snippet": "Retrieval -Augmented Generation (RAG) is an advanced technique designed to enhance the output of Language Models (LMs) by incorporating external ...", "subpage_snippet": "", "source": "aman.ai", "link": "https://aman.ai/primers/ai/RAG/", "content": "Retrieval -Augmented Generation (RAG) is an advanced technique designed to enhance the output of Language Models (LMs) by incorporating external ..."}
+{"idx": 8, "title": "Aman's AI Journal • Primers • Ilya Sutskever's Top 30", "date": "", "ddg_snippet": "Dense Passage Retrieval for Open-Domain Question Answering ... Retrieval -Augmented Generation for Knowledge-Intensive NLP Tasks", "subpage_snippet": "", "source": "aman.ai", "link": "https://aman.ai/primers/ai/top-30-papers/", "content": "Dense Passage Retrieval for Open-Domain Question Answering ... Retrieval -Augmented Generation for Knowledge-Intensive NLP Tasks"}
+{"idx": 9, "title": "Jimmy Lin - ACL Anthology", "date": "", "ddg_snippet": "To this end, we propose Document Screenshot Embedding (DSE), a novel retrieval paradigm that regards document screenshots as a unified input format ...", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/people/j/jimmy-lin/", "content": "To this end, we propose Document Screenshot Embedding (DSE), a novel retrieval paradigm that regards document screenshots as a unified input format ..."}
diff --git a/data/sampled_jsons/RAGGED_framework_Section_3.1_three_retriever_paradigms.jsonl b/data/sampled_jsons/RAGGED_framework_Section_3.1_three_retriever_paradigms.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c9b0c596e1227802d2977f0cf504f0fea43e7b6d
--- /dev/null
+++ b/data/sampled_jsons/RAGGED_framework_Section_3.1_three_retriever_paradigms.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "RAGGED: Towards Informed Design of Scalable and Stable RAG ...", "date": "", "ddg_snippet": "In this work, we introduce RAGGED , a framework for system-atically evaluating RAG systems across diverse retriever -reader configurations, retrieval depths, and datasets. Our analysis reveals that reader ro-bustness to noise is the key determinant of RAG stability and scalability.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2403.09040v3", "content": "In this work, we introduce RAGGED , a framework for system-atically evaluating RAG systems across diverse retriever -reader configurations, retrieval depths, and datasets. Our analysis reveals that reader ro-bustness to noise is the key determinant of RAG stability and scalability."}
+{"idx": 1, "title": "Retrieval-Augmented Generation Architecture Framework ...", "date": "", "ddg_snippet": "Nov 29, 2024 · The RAG Architecture Framework (RAGAF) is designed to combine the strengths of retrieval-based methods and generative models, creating a robust system capable of producing accurate, contextually grounded responses. The architecture of a RAG framework consists of seven core components: Generator, Retriever , Orchestration, UI, Source, Evaluation, and Reranker (GROUSER), illustrated in Fig. 1 ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-031-77954-1_6", "content": "Nov 29, 2024 · The RAG Architecture Framework (RAGAF) is designed to combine the strengths of retrieval-based methods and generative models, creating a robust system capable of producing accurate, contextually grounded responses. The architecture of a RAG framework consists of seven core components: Generator, Retriever , Orchestration, UI, Source, Evaluation, and Reranker (GROUSER), illustrated in Fig. 1 ..."}
+{"idx": 2, "title": "[2403.09040] RAGGED: Towards Informed Design of Retrieval ...", "date": "", "ddg_snippet": "To answer this, we introduce the RAGGED framework to analyze and optimize RAG systems. On a set of representative DBQA tasks, we study two classic sparse and dense retrievers , and four top-performing LMs in encoder-decoder and decoder-only architectures. Through RAGGED , we uncover that different models suit substantially varied RAG setups.", "subpage_snippet": "", "source": "ar5iv.labs.arxiv.org", "link": "https://ar5iv.labs.arxiv.org/html/2403.09040", "content": "To answer this, we introduce the RAGGED framework to analyze and optimize RAG systems. On a set of representative DBQA tasks, we study two classic sparse and dense retrievers , and four top-performing LMs in encoder-decoder and decoder-only architectures. Through RAGGED , we uncover that different models suit substantially varied RAG setups."}
+{"idx": 3, "title": "RAGGED: Towards Informed Design of Retrieval ...", "date": "", "ddg_snippet": "14 Mar 2024 — We introduce a reusable framework that can easily be adapted to analyze new RAG components, such as retriever and reader models, as they evolve.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2403.09040v1", "content": "14 Mar 2024 — We introduce a reusable framework that can easily be adapted to analyze new RAG components, such as retriever and reader models, as they evolve."}
+{"idx": 4, "title": "RAGGED", "date": "", "ddg_snippet": "by J Hsia · Cited by 23 — RAG system components: We use three retrieval approaches: (1) BM25 (Robertson et al., 2009), a sparse retriever based on lexical matching; (2) ColBERT ( ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=KDXj60FpJr", "content": "by J Hsia · Cited by 23 — RAG system components: We use three retrieval approaches: (1) BM25 (Robertson et al., 2009), a sparse retriever based on lexical matching; (2) ColBERT ( ..."}
+{"idx": 5, "title": "CRP-RAG: A Retrieval-Augmented Generation Framework ...", "date": "", "ddg_snippet": "by K Xu · 2024 · Cited by 14 — In this section , we introduce the framework design and reasoning process of CRP- RAG ( Section 3.1 ), along with the structures and workflows of ...", "subpage_snippet": "", "source": "www.preprints.org", "link": "https://www.preprints.org/manuscript/202411.1648/v1", "content": "by K Xu · 2024 · Cited by 14 — In this section , we introduce the framework design and reasoning process of CRP- RAG ( Section 3.1 ), along with the structures and workflows of ..."}
+{"idx": 6, "title": "RAGged Edge: LLMs on a Retrieval Rollercoaster - HackerNoon", "date": "", "ddg_snippet": "Mar 25, 2025 · This paper is presented on adaptive retrieval-augmented LLM models evaluated against various baselines using comprehensive metrics and detailed implementation.", "subpage_snippet": "", "source": "hackernoon.com", "link": "https://hackernoon.com/ragged-edge-llms-on-a-retrieval-rollercoaster", "content": "Mar 25, 2025 · This paper is presented on adaptive retrieval-augmented LLM models evaluated against various baselines using comprehensive metrics and detailed implementation."}
+{"idx": 7, "title": "Next-Gen Large Language Models: The Retrieval-Augmented ...", "date": "", "ddg_snippet": "Jun 11, 2024 · 3.2 Retriever -Generator Integration Strategies Retrieval-Augmented Generation (RAG) systems rely on two key components: retrievers and generative models. Retrievers are responsible for efficiently searching and retrieving relevant information from large-scale knowledge bases. \"It involves two main phases, indexing and searching.", "subpage_snippet": "", "source": "www.freecodecamp.org", "link": "https://www.freecodecamp.org/news/retrieval-augmented-generation-rag-handbook/", "content": "Jun 11, 2024 · 3.2 Retriever -Generator Integration Strategies Retrieval-Augmented Generation (RAG) systems rely on two key components: retrievers and generative models. Retrievers are responsible for efficiently searching and retrieving relevant information from large-scale knowledge bases. \"It involves two main phases, indexing and searching."}
+{"idx": 8, "title": "RAGuard: A Novel Approach for In-Context Safe Retrieval ...", "date": "", "ddg_snippet": "2 days ago · This paper addresses this gap by presenting RAGuard, a framework that integrates safety-aware retrieval filtering with context-specific generation to enhance reliability in OSW maintenance scenarios.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-032-05073-1_13", "content": "2 days ago · This paper addresses this gap by presenting RAGuard, a framework that integrates safety-aware retrieval filtering with context-specific generation to enhance reliability in OSW maintenance scenarios."}
+{"idx": 9, "title": "arXiv:2403.09040v1 [cs.CL] 14 Mar 2024", "date": "", "ddg_snippet": "Abstract Retrieval-augmented generation (RAG) greatly benefits language models (LMs) by providing additional context for tasks such as document-based question answering (DBQA). Despite its potential, the power of RAG is highly dependent on its configuration, raising question: What is the optimal RAG config-uration? To answer this, we introduce RAGGED framework to analyze and optimize RAG ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2403.09040v1", "content": "Abstract Retrieval-augmented generation (RAG) greatly benefits language models (LMs) by providing additional context for tasks such as document-based question answering (DBQA). Despite its potential, the power of RAG is highly dependent on its configuration, raising question: What is the optimal RAG config-uration? To answer this, we introduce RAGGED framework to analyze and optimize RAG ..."}
diff --git a/data/sampled_jsons/Recursive_Allocation_Selection_RAS_algorithm_implementation_budget_workers_scores.jsonl b/data/sampled_jsons/Recursive_Allocation_Selection_RAS_algorithm_implementation_budget_workers_scores.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..84d99b5f29ff6225dd89b287a9dc6805f8a220e7
--- /dev/null
+++ b/data/sampled_jsons/Recursive_Allocation_Selection_RAS_algorithm_implementation_budget_workers_scores.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ATA: Adaptive Task Allocation for Efficient Resource Management in ...", "date": "", "ddg_snippet": "In this paper, we propose ATA(Adaptive Task Allocation ), a method that adapts to heterogeneous and random distributions of worker computation times. Through rigorous theoretical analysis, we show that ATAidentifies the optimal task allocation and performs comparably to methods with prior knowledge of computation times.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.00775v1", "content": "In this paper, we propose ATA(Adaptive Task Allocation ), a method that adapts to heterogeneous and random distributions of worker computation times. Through rigorous theoretical analysis, we show that ATAidentifies the optimal task allocation and performs comparably to methods with prior knowledge of computation times."}
+{"idx": 1, "title": "(PDF) A Critical Review of Resource Allocation ... - ResearchGate", "date": "", "ddg_snippet": "PDF | On Jul 6, 2024, Srinivasa Rao Allu and others published A Critical Review of Resource Allocation Optimization in Project Management | Find, read and cite all the research you need on ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/382049201_A_Critical_Review_of_Resource_Allocation_Optimization_in_Project_Management", "content": "PDF | On Jul 6, 2024, Srinivasa Rao Allu and others published A Critical Review of Resource Allocation Optimization in Project Management | Find, read and cite all the research you need on ..."}
+{"idx": 2, "title": "An efficient algorithm for task allocation with the budget constraint", "date": "", "ddg_snippet": "This paper studies a heterogeneous task allocation problem with the budget constraint. Existing works on task allocation mainly tackle this well-known NP-hard problem from an optimisation perspective. They have not been able to cater to the extra needs of scalability and robustness in large-scale systems.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S095741742201418X", "content": "This paper studies a heterogeneous task allocation problem with the budget constraint. Existing works on task allocation mainly tackle this well-known NP-hard problem from an optimisation perspective. They have not been able to cater to the extra needs of scalability and robustness in large-scale systems."}
+{"idx": 3, "title": "Re-balancing or disaggregating a matrix using the RAS algorithm", "date": "", "ddg_snippet": "This code is used for re-balancing or disaggregating an n by m matrix using the RAS algorithm . The application of this algorithm is to fill n unknown elements in an Input Output Table, either for projected results, or when splitting economic sectors.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/EStebb/ras", "content": "This code is used for re-balancing or disaggregating an n by m matrix using the RAS algorithm . The application of this algorithm is to fill n unknown elements in an Input Output Table, either for projected results, or when splitting economic sectors."}
+{"idx": 4, "title": "Resource Allocation System (RAS) - Citizen Network", "date": "", "ddg_snippet": "A Resource Allocation System ( RAS ) is any set of rules that allows fair allocations to be made to people who need extra support.", "subpage_snippet": "", "source": "citizen-network.org", "link": "https://citizen-network.org/library/resource-allocation-system-ras.html", "content": "A Resource Allocation System ( RAS ) is any set of rules that allows fair allocations to be made to people who need extra support."}
+{"idx": 5, "title": "MAB-Based Reinforced Worker Selection Framework for ... - IEEE Xplore", "date": "", "ddg_snippet": "Spatial crowdsensing is a special kind of crowdsourcing which allocates tasks to workers in some special places where workers can sense data for them. Due to the lack of priori information about the quality of workers and the ground truth, selecting the most suitable workers , which can guarantee the quality of the sensing tasks, remains a great challenge. In this paper, we propose a novel ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/9086123", "content": "Spatial crowdsensing is a special kind of crowdsourcing which allocates tasks to workers in some special places where workers can sense data for them. Due to the lack of priori information about the quality of workers and the ground truth, selecting the most suitable workers , which can guarantee the quality of the sensing tasks, remains a great challenge. In this paper, we propose a novel ..."}
+{"idx": 6, "title": "PDF Resource Allocation System Practice Instructions", "date": "", "ddg_snippet": "For those individuals who have received a strengths based assessment, and the conclusion of the assessment is that they have eligible, unmet needs; the worker will proceed to use the RAS (resource allocation system) to calculate an indicative personal budget . The RAS should only be used to calculate a budget when no other alternatives have been identified to meet eligible needs. This ...", "subpage_snippet": "", "source": "www.connecttosupport.org", "link": "https://www.connecttosupport.org/Resources/Councils/WestSussex/downloads/operating-model/practice-instructions/ras_practice_instructionsv1.1.pdf", "content": "For those individuals who have received a strengths based assessment, and the conclusion of the assessment is that they have eligible, unmet needs; the worker will proceed to use the RAS (resource allocation system) to calculate an indicative personal budget . The RAS should only be used to calculate a budget when no other alternatives have been identified to meet eligible needs. This ..."}
+{"idx": 7, "title": "PDF Parallel algorithm portfolios with adaptive resource allocation strategy", "date": "", "ddg_snippet": "We propose a resource allocation process based on adaptive decision-making procedures. The proposed approach is incorporated in algorithm portfolios composed of three essential types of numerical optimization algorithms , namely gradient-based, direct search, and swarm intelligence algorithms .", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/s10898-022-01162-y.pdf", "content": "We propose a resource allocation process based on adaptive decision-making procedures. The proposed approach is incorporated in algorithm portfolios composed of three essential types of numerical optimization algorithms , namely gradient-based, direct search, and swarm intelligence algorithms ."}
+{"idx": 8, "title": "Estimated budget - Resource Allocation System (RAS) principles, rules ...", "date": "", "ddg_snippet": "Assessments Estimated budget - Resource Allocation System ( RAS ) principles, rules and rates Devon's Resource Allocation System ( RAS ) is designed to be as fair and transparent as possible. We adopt the following principles, rules and cost rates.", "subpage_snippet": "", "source": "www.devon.gov.uk", "link": "https://www.devon.gov.uk/adult-social-care/assessments/care-support-needs-assessments/personal-budgets/estimated-budget-ras-principles-rules-and-rates/", "content": "Assessments Estimated budget - Resource Allocation System ( RAS ) principles, rules and rates Devon's Resource Allocation System ( RAS ) is designed to be as fair and transparent as possible. We adopt the following principles, rules and cost rates."}
+{"idx": 9, "title": "RAS | Proceedings of the ACM SIGOPS 28th Symposium on Operating Systems ...", "date": "", "ddg_snippet": "We take a two-level approach to scale resource allocation to all datacenters in a region, where a mixed-integer-programming solver continuously optimizes server-to-reservation assignments off the critical path, and a traditional container allocator does real-time placement of containers on servers in a reservation.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/abs/10.1145/3477132.3483578", "content": "We take a two-level approach to scale resource allocation to all datacenters in a region, where a mixed-integer-programming solver continuously optimizes server-to-reservation assignments off the critical path, and a traditional container allocator does real-time placement of containers on servers in a reservation."}
diff --git a/data/sampled_jsons/Revisiting_Simple_Regret_Fast_Rates_for_Returning_a_Good_Arm_Zhao_2023.jsonl b/data/sampled_jsons/Revisiting_Simple_Regret_Fast_Rates_for_Returning_a_Good_Arm_Zhao_2023.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1579aa59e45ef9e32f52f3107a36d9c6aa09d9fe
--- /dev/null
+++ b/data/sampled_jsons/Revisiting_Simple_Regret_Fast_Rates_for_Returning_a_Good_Arm_Zhao_2023.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Revisiting Simple Regret: Fast Rates for Returning a Good Arm", "date": "", "ddg_snippet": "Oct 30, 2022 · In this paper, we make significant progress on minimizing simple regret in both data-rich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms, and T is the number of samples.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2210.16913", "content": "Oct 30, 2022 · In this paper, we make significant progress on minimizing simple regret in both data-rich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms, and T is the number of samples."}
+{"idx": 1, "title": "Revisiting Simple Regret: Fast Rates for Returning a Good Arm", "date": "", "ddg_snippet": "In this paper, we make a significant progress on minimizing simple regret in both data-rich ($T\\ge n$) and data-poor regime ($T \\le n$) where $n$ is the number of arms and $T$ is the number of samples.", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v202/zhao23g.html", "content": "In this paper, we make a significant progress on minimizing simple regret in both data-rich ($T\\ge n$) and data-poor regime ($T \\le n$) where $n$ is the number of arms and $T$ is the number of samples."}
+{"idx": 2, "title": "Revisiting simple regret: fast rates for returning a good arm", "date": "", "ddg_snippet": "Jul 23, 2023 · In this paper, we make significant progress on minimizing simple regret in both datarich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms , and T is the number of samples.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3618408.3620180", "content": "Jul 23, 2023 · In this paper, we make significant progress on minimizing simple regret in both datarich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms , and T is the number of samples."}
+{"idx": 3, "title": "Yao Zhao", "date": "", "ddg_snippet": "Hello, I’m Yao Zhao , a 5th-year PhD student at the University of Arizona in computer science working with Dr. Kwang-Sung Jun. I mainly have fun exploring multi- armed bandits / RL / adaptive experimentation, and recently RLHF.", "subpage_snippet": "", "source": "meyaozhao.github.io", "link": "https://meyaozhao.github.io/", "content": "Hello, I’m Yao Zhao , a 5th-year PhD student at the University of Arizona in computer science working with Dr. Kwang-Sung Jun. I mainly have fun exploring multi- armed bandits / RL / adaptive experimentation, and recently RLHF."}
+{"idx": 4, "title": "ICML Poster Revisiting Simple Regret: Fast Rates for ...", "date": "", "ddg_snippet": "Abstract: Simple regret is a natural and parameter-free performance criterion for pure exploration in multi-armed bandits yet is less popular than the probability of missing the best arm or an ϵ ϵ - good arm , perhaps due to lack of easy ways to characterize it.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2023/poster/24916", "content": "Abstract: Simple regret is a natural and parameter-free performance criterion for pure exploration in multi-armed bandits yet is less popular than the probability of missing the best arm or an ϵ ϵ - good arm , perhaps due to lack of easy ways to characterize it."}
+{"idx": 5, "title": "Revisiting Simple Regret Minimization in Multi-Armed Bandits", "date": "", "ddg_snippet": "Oct 30, 2022 · Simple regret is a natural and parameter-free performance criterion for identifying a good arm in multi-armed bandits yet is less popular than the probability of missing the best arm or...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/364953384_Revisiting_Simple_Regret_Minimization_in_Multi-Armed_Bandits", "content": "Oct 30, 2022 · Simple regret is a natural and parameter-free performance criterion for identifying a good arm in multi-armed bandits yet is less popular than the probability of missing the best arm or..."}
+{"idx": 6, "title": "Revisiting Simple Regret: Fast Rates for Returning a Good Arm", "date": "", "ddg_snippet": "Apr 24, 2023 · In this paper, we make a significant progress on minimizing simple regret in both data-rich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms and T is the number of samples.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=qFk9EUzycd", "content": "Apr 24, 2023 · In this paper, we make a significant progress on minimizing simple regret in both data-rich (T ≥ n) and data-poor regime (T ≤ n) where n is the number of arms and T is the number of samples."}
+{"idx": 7, "title": "COLT 2023 Abstracts", "date": "", "ddg_snippet": "Abstract: We present a fast , differentially private algorithm for high-dimensional covariance-aware mean estimation with nearly optimal sample ...", "subpage_snippet": "", "source": "learningtheory.org", "link": "https://learningtheory.org/colt2023/abstracts.html", "content": "Abstract: We present a fast , differentially private algorithm for high-dimensional covariance-aware mean estimation with nearly optimal sample ..."}
+{"idx": 8, "title": "Csaba Szepesvári", "date": "", "ddg_snippet": "... Natural Policy Gradient: a Simple Efficient Policy Optimization Framework for ... Revisiting Simple Regret : Fast Rates for Returning a Good Arm .", "subpage_snippet": "", "source": "www.csauthors.net", "link": "https://www.csauthors.net/csaba-szepesvari/", "content": "... Natural Policy Gradient: a Simple Efficient Policy Optimization Framework for ... Revisiting Simple Regret : Fast Rates for Returning a Good Arm ."}
+{"idx": 9, "title": "GitHub - opendilab/awesome-exploration-rl: A curated list of", "date": "", "ddg_snippet": "... an intuitive feeling for ... For other detailed survey on exploration methods in RL, you can refer to Tianpei Yang et al and Susan Amin et al .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/opendilab/awesome-exploration-rl", "content": "... an intuitive feeling for ... For other detailed survey on exploration methods in RL, you can refer to Tianpei Yang et al and Susan Amin et al ."}
diff --git a/data/sampled_jsons/Roose_2024_'tangle_of_sloppy_tests'_Generative_AI_evaluation_year_2024.jsonl b/data/sampled_jsons/Roose_2024_'tangle_of_sloppy_tests'_Generative_AI_evaluation_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..13b2a8253bcdd27e140b1c616109bf5810659d54
--- /dev/null
+++ b/data/sampled_jsons/Roose_2024_'tangle_of_sloppy_tests'_Generative_AI_evaluation_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Position: Evaluating Generative AI Systems Is a Social ...", "date": "", "ddg_snippet": "by H Wallach · 2025 · Cited by 12 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2502.00561", "content": "by H Wallach · 2025 · Cited by 12 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ..."}
+{"idx": 1, "title": "Position: Evaluating Generative AI Systems is a Social Science ...", "date": "", "ddg_snippet": "The measurement tasks involved in evaluating generative AI (GenAI) systems are especially difficult, leading to what has been described as “a tangle of sloppy ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.00561v1", "content": "The measurement tasks involved in evaluating generative AI (GenAI) systems are especially difficult, leading to what has been described as “a tangle of sloppy ..."}
+{"idx": 2, "title": "Evaluating Generative AI Systems Is a Social Science ...", "date": "", "ddg_snippet": "by H Wallach · Cited by 12 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=1ZC4RNjqzU", "content": "by H Wallach · Cited by 12 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ..."}
+{"idx": 3, "title": "Position: Evaluating Generative AI Systems Is a Social ...", "date": "", "ddg_snippet": "The measurement tasks involved in evaluating generative AI (GenAI) systems lack sufficient scientific rigor, leading to what has been described as “a tangle of ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/40182", "content": "The measurement tasks involved in evaluating generative AI (GenAI) systems lack sufficient scientific rigor, leading to what has been described as “a tangle of ..."}
+{"idx": 4, "title": "Evaluating Generative AI Systems is a Social Science ...", "date": "", "ddg_snippet": "by H Wallach · Cited by 12 — “a tangle of sloppy tests [and] apples-to-oranges comparisons” ( Roose , 2024 ). In this position paper, we argue that the ML community would benefit from ...", "subpage_snippet": "", "source": "afedercooper.info", "link": "https://afedercooper.info/paper/wallach2024measurement.pdf", "content": "by H Wallach · Cited by 12 — “a tangle of sloppy tests [and] apples-to-oranges comparisons” ( Roose , 2024 ). In this position paper, we argue that the ML community would benefit from ..."}
+{"idx": 5, "title": "Position: Evaluating Generative AI Systems is a Social Science ...", "date": "", "ddg_snippet": "... generative AI (GenAI) systems are especially difficult, leading to what has been described as\"a tangle of sloppy tests [and] apples-to-oranges comparisons\"( ...", "subpage_snippet": "", "source": "www.consensus.app", "link": "https://www.consensus.app/papers/position-evaluating-generative-ai-systems-is-a-social-vaughan-corvi/6c05b8e5c587597da2560bb61f2cb2f9/", "content": "... generative AI (GenAI) systems are especially difficult, leading to what has been described as\"a tangle of sloppy tests [and] apples-to-oranges comparisons\"( ..."}
+{"idx": 6, "title": "Position: Evaluating Generative AI Systems is a Social ...", "date": "", "ddg_snippet": "31 Jan 2025 — ... tangle of sloppy tests [and] apples-to-oranges comparisons” ( Roose , 2024 ). In this position paper, we argue that the ML community would ...", "subpage_snippet": "", "source": "www.microsoft.com", "link": "https://www.microsoft.com/en-us/research/publication/position-evaluating-generative-ai-systems-is-a-social-science-measurement-challenge/?locale=ko-kr", "content": "31 Jan 2025 — ... tangle of sloppy tests [and] apples-to-oranges comparisons” ( Roose , 2024 ). In this position paper, we argue that the ML community would ..."}
+{"idx": 7, "title": "Revision History for Position: Evaluating Generative AI...", "date": "", "ddg_snippet": "23 Mar 2025 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/revisions?id=AJm6unbR3f", "content": "23 Mar 2025 — ... tangle of sloppy tests [and] apples-to-oranges comparisons\" ( Roose , 2024 ). In this position paper, we argue that the ML community would ..."}
+{"idx": 8, "title": "Microsoft New Future of Work Report 2024", "date": "", "ddg_snippet": "4 Dec 2024 — Evaluating General-Purpose AI with Psychometrics. •. To quote Kevin Roose in The New York Times, “ AI measurement is a mess—a tangle of sloppy ... 52 pages", "subpage_snippet": "", "source": "www.microsoft.com", "link": "https://www.microsoft.com/en-us/research/wp-content/uploads/2024/12/NFWReport2024_12.20.24.pdf", "content": "4 Dec 2024 — Evaluating General-Purpose AI with Psychometrics. •. To quote Kevin Roose in The New York Times, “ AI measurement is a mess—a tangle of sloppy ... 52 pages"}
+{"idx": 9, "title": "A.I. Has a Measurement Problem", "date": "", "ddg_snippet": "15 Apr 2024 — In short, A.I. measurement is a mess — a tangle of sloppy tests , apples-to-oranges comparisons and self-serving hype that has left users, ...", "subpage_snippet": "", "source": "www.nytimes.com", "link": "https://www.nytimes.com/2024/04/15/technology/ai-models-measurement.html", "content": "15 Apr 2024 — In short, A.I. measurement is a mess — a tangle of sloppy tests , apples-to-oranges comparisons and self-serving hype that has left users, ..."}
diff --git a/data/sampled_jsons/Rubin_1974_potential_response_variables_causal_inference_abstract.jsonl b/data/sampled_jsons/Rubin_1974_potential_response_variables_causal_inference_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..06a03176f2b9a1dd8b3e9e917ef239f7c39c23da
--- /dev/null
+++ b/data/sampled_jsons/Rubin_1974_potential_response_variables_causal_inference_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Rubin causal model - Wikipedia", "date": "", "ddg_snippet": "The Rubin causal model (RCM), also known as the Neyman- Rubin causal model, [1] is an approach to the statistical analysis of cause and effect based on the framework of potential outcomes, named after Donald Rubin .", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Rubin_causal_model", "content": "The Rubin causal model (RCM), also known as the Neyman- Rubin causal model, [1] is an approach to the statistical analysis of cause and effect based on the framework of potential outcomes, named after Donald Rubin ."}
+{"idx": 1, "title": "Rubin Causal Model | SpringerLink", "date": "", "ddg_snippet": "It appears that the first use of the formal concept of potential outcomes to define causal effects was Neyman ( 1923) in the context of randomization-based inference in randomized experiments, but this notation was not extended to nonrandomized studies until Rubin ( 1974 , 2010) provides some historical perspective.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/rwe/10.1007/978-3-662-69359-9_540", "content": "It appears that the first use of the formal concept of potential outcomes to define causal effects was Neyman ( 1923) in the context of randomization-based inference in randomized experiments, but this notation was not extended to nonrandomized studies until Rubin ( 1974 , 2010) provides some historical perspective."}
+{"idx": 2, "title": "Causal Inference Using Potential Outcomes: Design, Modeling, Decisions ...", "date": "", "ddg_snippet": "Abstract Causal effects are defined as comparisons of potential outcomes under different treatments on a common set of units. Observed values of the potential outcomes are revealed by the assignment mechanism—a probabilistic model for the treatment each unit receives as a function of covariates and potential outcomes. Fisher made tremendous contributions to causal inference through his work ...", "subpage_snippet": "", "source": "www.tandfonline.com", "link": "https://www.tandfonline.com/doi/abs/10.1198/016214504000001880", "content": "Abstract Causal effects are defined as comparisons of potential outcomes under different treatments on a common set of units. Observed values of the potential outcomes are revealed by the assignment mechanism—a probabilistic model for the treatment each unit receives as a function of covariates and potential outcomes. Fisher made tremendous contributions to causal inference through his work ..."}
+{"idx": 3, "title": "PDF Comparing Causal Frameworks: Potential Abstractions", "date": "", "ddg_snippet": "Toward the end of the twentieth century several frameworks arose for formalizing and an-alyzing problems of causal inference . One of these, associated with Rubin [37] and others (see [18]), takes the potential outcome—formalizing the effect of one variable on another—as a fundamental target of analysis. Causal assumptions in the Rubin causal model (RCM) are naturally encoded as algebraic ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2306.14351.pdf", "content": "Toward the end of the twentieth century several frameworks arose for formalizing and an-alyzing problems of causal inference . One of these, associated with Rubin [37] and others (see [18]), takes the potential outcome—formalizing the effect of one variable on another—as a fundamental target of analysis. Causal assumptions in the Rubin causal model (RCM) are naturally encoded as algebraic ..."}
+{"idx": 4, "title": "PDF Campbell's and Rubin's Perspectives on Causal Inference", "date": "", "ddg_snippet": "In this article, we provide an introduction to Donald Campbell's (Campbell, 1957; Shadish, Cook, & Campbell, 2002) and Donald Rubin's (Holland, 1986; Rubin , 1974 , 2005) perspectives on causal inference . Campbell's perspective has dominated thinking about causal inference in psychology, education, and some other behavioral sciences. Rubin's causal model (a.k.a., the potential outcomes ...", "subpage_snippet": "", "source": "jenni.uchicago.edu", "link": "http://jenni.uchicago.edu/Spencer_Conference/Papers+2010/West_Thoemmes_2010_Campbells+and+Rubins+Perspectives+on+Causal+Inference.pdf", "content": "In this article, we provide an introduction to Donald Campbell's (Campbell, 1957; Shadish, Cook, & Campbell, 2002) and Donald Rubin's (Holland, 1986; Rubin , 1974 , 2005) perspectives on causal inference . Campbell's perspective has dominated thinking about causal inference in psychology, education, and some other behavioral sciences. Rubin's causal model (a.k.a., the potential outcomes ..."}
+{"idx": 5, "title": "PDF Causal Inference - Harvard University", "date": "", "ddg_snippet": "Here, we use the commonly accepted statistical framework of causality that is based on the notion of potential outcomes. This framework is often called the Neyman- Rubin causal model because the framework rst appeared in Neyman (1923)'s analysis of randomized experiments and Rubin ( 1974 ) extended it to observational studies.", "subpage_snippet": "", "source": "imai.fas.harvard.edu", "link": "https://imai.fas.harvard.edu/teaching/files/causal.pdf", "content": "Here, we use the commonly accepted statistical framework of causality that is based on the notion of potential outcomes. This framework is often called the Neyman- Rubin causal model because the framework rst appeared in Neyman (1923)'s analysis of randomized experiments and Rubin ( 1974 ) extended it to observational studies."}
+{"idx": 6, "title": "2 Statistical Inference for Causal Effects, With Emphasis on ...", "date": "", "ddg_snippet": "We present here a framework for causal inference that is now commonly referred to as \" Rubin's Causal Model\" (RCM, Holland, 1986), for a series of articles written in the 1970s ( Rubin , 1974 , Rubin , 1975, Rubin , 1976a, Rubin , 1977, Rubin , 1978, Rubin , 1979a, Rubin , 1980). The framework has two essential parts and one optional part. The first part of the RCM defines causal effects through ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0169716107270026", "content": "We present here a framework for causal inference that is now commonly referred to as \" Rubin's Causal Model\" (RCM, Holland, 1986), for a series of articles written in the 1970s ( Rubin , 1974 , Rubin , 1975, Rubin , 1976a, Rubin , 1977, Rubin , 1978, Rubin , 1979a, Rubin , 1980). The framework has two essential parts and one optional part. The first part of the RCM defines causal effects through ..."}
+{"idx": 7, "title": "Causal Inference Using Potential Outcomes: Design, Modeling ... - JSTOR", "date": "", "ddg_snippet": "Donald B. Rubin Causal effects are defined as comparisons of potential outcomes under different treatments on a common set of units. Observed values of the potential outcomes are revealed by the assignment mechanism?a probabilistic model for the treatment each unit receives as a function of covariates and potential outcomes. Fisher made tremendous contributions to causal inference through his ...", "subpage_snippet": "", "source": "www.jstor.org", "link": "https://www.jstor.org/stable/pdf/27590541", "content": "Donald B. Rubin Causal effects are defined as comparisons of potential outcomes under different treatments on a common set of units. Observed values of the potential outcomes are revealed by the assignment mechanism?a probabilistic model for the treatment each unit receives as a function of covariates and potential outcomes. Fisher made tremendous contributions to causal inference through his ..."}
+{"idx": 8, "title": "PDF Lec4_ Rubin_potential_outcomes", "date": "", "ddg_snippet": "Methods for Causal Inference Lecture 4: Rubin's potential outcomes framework Ava Khamseh School of Informatics 2024-2025 Language of probability: Variables , evens, samples space, probability law Probability axioms, (conditional) total law of probability, independence, Bayes' rule", "subpage_snippet": "", "source": "opencourse.inf.ed.ac.uk", "link": "https://opencourse.inf.ed.ac.uk/sites/default/files/https/opencourse.inf.ed.ac.uk/mci/2025/lec4-rubinpotentialoutcomes.pdf", "content": "Methods for Causal Inference Lecture 4: Rubin's potential outcomes framework Ava Khamseh School of Informatics 2024-2025 Language of probability: Variables , evens, samples space, probability law Probability axioms, (conditional) total law of probability, independence, Bayes' rule"}
+{"idx": 9, "title": "PDF Teaching Causal Inference in Experiments and Observational Studies", "date": "", "ddg_snippet": "Key Words: potential outcomes, Rubin Causal Model (RCM), Fisherian inference , Neymanian inference , Bayesian inference , experiments, observational studies, instrumental variables , noncompliance, statistical education Inference for causal effects is a critical activity in many branches of science and public policy.", "subpage_snippet": "", "source": "www.statlit.org", "link": "http://www.statlit.org/pdf/1999RubinASA.pdf", "content": "Key Words: potential outcomes, Rubin Causal Model (RCM), Fisherian inference , Neymanian inference , Bayesian inference , experiments, observational studies, instrumental variables , noncompliance, statistical education Inference for causal effects is a critical activity in many branches of science and public policy."}
diff --git a/data/sampled_jsons/SimXRD_Table_1_AdvancedXRDAnalysis_Lee_2023_7_times_more_patterns_justification.jsonl b/data/sampled_jsons/SimXRD_Table_1_AdvancedXRDAnalysis_Lee_2023_7_times_more_patterns_justification.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c9d04a057b4e615fed0ec9bf9fcdc635cd28724f
--- /dev/null
+++ b/data/sampled_jsons/SimXRD_Table_1_AdvancedXRDAnalysis_Lee_2023_7_times_more_patterns_justification.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "SimXRD -4M: Big Simulated X-ray Diffraction Data and Crystal...", "date": "", "ddg_snippet": "AdvancedXRDAnalysis ( Lee et al., 2023 ). 29,569,650.Powder XRD patterns offer a one-dimensional representation of the three-dimensional diffraction pattern and are the most common experimental measurement.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2406.15469v2", "content": "AdvancedXRDAnalysis ( Lee et al., 2023 ). 29,569,650.Powder XRD patterns offer a one-dimensional representation of the three-dimensional diffraction pattern and are the most common experimental measurement."}
+{"idx": 1, "title": "(PDF) SIMXRD -4M: big simulated x-ray diffraction data and crystal...", "date": "", "ddg_snippet": "SimXRD comprises 4,065,346 simulated powder XRD patterns , representing 119,569 unique crystal structures under 33 simulated conditions that reflect real-world variations.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/389657996_SIMXRD-4M_BIG_SIMULATED_X-RAY_DIFFRACTION_DATA_AND_CRYSTAL_SYMMETRY_CLASSIFICATION_BENCHMARK", "content": "SimXRD comprises 4,065,346 simulated powder XRD patterns , representing 119,569 unique crystal structures under 33 simulated conditions that reflect real-world variations."}
+{"idx": 2, "title": "FAQ На море на автомобиле 2023 - обсуждение на форуме e 1 .ru...", "date": "", "ddg_snippet": "E. El_Camellero_. 20:13, 15.07. 2023 . От пользователя Ответственный за романтику. Туапсинский район сегодня.О. =Один в поле не воин=. 20:24, 15.07. 2023 . От пользователя JeeWell. Еще вопрос С Саратова лучше через Богучар или Волгоград?", "subpage_snippet": "", "source": "m.e1.ru", "link": "https://m.e1.ru/f/67/21427596/p/101", "content": "E. El_Camellero_. 20:13, 15.07. 2023 . От пользователя Ответственный за романтику. Туапсинский район сегодня.О. =Один в поле не воин=. 20:24, 15.07. 2023 . От пользователя JeeWell. Еще вопрос С Саратова лучше через Богучар или Волгоград?"}
+{"idx": 3, "title": "7 , 8 кл Русский язык (3-5 октября, 2023 ), ответы какие?", "date": "", "ddg_snippet": "03.10. 2023 (03.10. 2023 ) Помогай-ка. 1 . Напишите ответ в строке (без учета регистра). Прочитайте текст с пропущенными словами. В русском языке есть слово ( 1 ) со значением «жидкость, содержащаяся в клетках, тканях и полостях растительных и животных...", "subpage_snippet": "", "source": "shpargalka-reshalka.ru", "link": "https://shpargalka-reshalka.ru/1347/", "content": "03.10. 2023 (03.10. 2023 ) Помогай-ка. 1 . Напишите ответ в строке (без учета регистра). Прочитайте текст с пропущенными словами. В русском языке есть слово ( 1 ) со значением «жидкость, содержащаяся в клетках, тканях и полостях растительных и животных..."}
+{"idx": 4, "title": "D ATA", "date": "", "ddg_snippet": "AdvancedXRDAnalysis ( Lee et al., 2023 ) CrySTINet (Chen et al., 2024) CPICANN (Cao, 2024) SimXRD . XRD pattern Powder XRD patterns offer a one-dimensional representation of the three-dimensional diffraction pattern and are the most common experimental measurement.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=mkuB677eMM", "content": "AdvancedXRDAnalysis ( Lee et al., 2023 ) CrySTINet (Chen et al., 2024) CPICANN (Cao, 2024) SimXRD . XRD pattern Powder XRD patterns offer a one-dimensional representation of the three-dimensional diffraction pattern and are the most common experimental measurement."}
+{"idx": 5, "title": "Диспетчер Realtek HD для windows 10 нет в панели... | Medium", "date": "", "ddg_snippet": "No responses yet. More from Windows Tips.Husbands Are 7 Times More Likely to Leave Their Wives Over This 1 Thing.", "subpage_snippet": "", "source": "windows-tips.medium.com", "link": "https://windows-tips.medium.com/диспетчер-realtek-hd-для-windows-10-нет-в-панели-управления-что-делать-894438903ef", "content": "No responses yet. More from Windows Tips.Husbands Are 7 Times More Likely to Leave Their Wives Over This 1 Thing."}
+{"idx": 6, "title": "Oting Паладин'2025: фото, минусы и плюсы, цены и характеристики", "date": "", "ddg_snippet": "Официально Oting Паладин представили 28 ноября 2023 года на специальном мероприятии в Москве, причем состоялась премьера не только отдельно взятой модели, но и нового бренда в целом.", "subpage_snippet": "", "source": "ironhorse.ru", "link": "https://ironhorse.ru/oting/paladin/", "content": "Официально Oting Паладин представили 28 ноября 2023 года на специальном мероприятии в Москве, причем состоялась премьера не только отдельно взятой модели, но и нового бренда в целом."}
+{"idx": 7, "title": "Первый взгляд на Infinix Note 30 и Note 30 i : видимо, лучший...", "date": "", "ddg_snippet": "В прошлом году серия Infinix Note 12 Pro и Note 12 2023 доминировала на российском рынке по соотношению цена-возможности. Сложно было составить любую подборку смартфонов, чтобы не упомянуть про эти аппараты.", "subpage_snippet": "", "source": "mobile-review.com", "link": "https://mobile-review.com/all/reviews/smartphones/pervyj-vzglyad-na-infinix-note-30-i-note-30i-vidimo-luchshij-smartfon-leta-v-segmente-do-20-tysyach/", "content": "В прошлом году серия Infinix Note 12 Pro и Note 12 2023 доминировала на российском рынке по соотношению цена-возможности. Сложно было составить любую подборку смартфонов, чтобы не упомянуть про эти аппараты."}
+{"idx": 8, "title": "Скачать Geometry Dash 2.207 для Windows", "date": "", "ddg_snippet": "Geometry Dash Обновление 2.2 — двенадцатое крупное обновление, выпущенное 20 декабря 2023 года для ПК, Android и iOS. Обновление содержит главный уровень Dash, новые элементы управления камерой, эффекты постобработки для уровней и новый режим...", "subpage_snippet": "", "source": "geometriodash.ru", "link": "https://geometriodash.ru/skachat-na-windows", "content": "Geometry Dash Обновление 2.2 — двенадцатое крупное обновление, выпущенное 20 декабря 2023 года для ПК, Android и iOS. Обновление содержит главный уровень Dash, новые элементы управления камерой, эффекты постобработки для уровней и новый режим..."}
+{"idx": 9, "title": "Игры на PC торрент скачать бесплатно", "date": "", "ddg_snippet": "Игры 2023 года.", "subpage_snippet": "", "source": "utorrentgames.best", "link": "https://utorrentgames.best/", "content": "Игры 2023 года."}
diff --git a/data/sampled_jsons/Simple_yet_Effective_Incomplete_Multi-view_Clustering_Algorithm_2_convergence_termination_condition.jsonl b/data/sampled_jsons/Simple_yet_Effective_Incomplete_Multi-view_Clustering_Algorithm_2_convergence_termination_condition.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7ca59bcfa89d4bbcce985413dcc669913f732037
--- /dev/null
+++ b/data/sampled_jsons/Simple_yet_Effective_Incomplete_Multi-view_Clustering_Algorithm_2_convergence_termination_condition.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Automatic and Aligned Anchor Learning Strategy for Multi- ...", "date": "", "ddg_snippet": "by H Ma · Cited by 11 — clusters tends to average, we terminate the algorithm and obtain a more representative anchor set. The complete HBNC process ise elaborated in Algorithm 2 .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=TKRqWQVawP", "content": "by H Ma · Cited by 11 — clusters tends to average, we terminate the algorithm and obtain a more representative anchor set. The complete HBNC process ise elaborated in Algorithm 2 ."}
+{"idx": 1, "title": "DATA CLUSTERING: Algorithms and Applications", "date": "", "ddg_snippet": "... multi - view clustering , cluster en- sembles, and cluster validation ... Algorithm 2 illustrates these steps of optimizing Equation (2.4). © 2014 by ... 648 pages", "subpage_snippet": "", "source": "people.cs.vt.edu", "link": "https://people.cs.vt.edu/~reddy/papers/DCBOOK.pdf", "content": "... multi - view clustering , cluster en- sembles, and cluster validation ... Algorithm 2 illustrates these steps of optimizing Equation (2.4). © 2014 by ... 648 pages"}
+{"idx": 2, "title": "Boosting Spectral Clustering on Incomplete Data via ...", "date": "", "ddg_snippet": "by F Yu · 2023 · Cited by 3 — However, the KSL-Pp algorithm involves numerous hyper-parameters and entails a non-convex optimization process during the Z-update step (Line 5 in Algorithm 2 ), ... 21 pages", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/e5aa7171449b83f8b4eec1623eac9906-Paper-Conference.pdf", "content": "by F Yu · 2023 · Cited by 3 — However, the KSL-Pp algorithm involves numerous hyper-parameters and entails a non-convex optimization process during the Z-update step (Line 5 in Algorithm 2 ), ... 21 pages"}
+{"idx": 3, "title": "Multi-View Semi-Supervised Learning with Missing Views", "date": "", "ddg_snippet": "by B Quanz · 2012 · Cited by 2 — . Algorithm 2 Testing using the Feature Generation Network. Input: A set of ... here we propose a simple and efficient approach for the synthetic data.", "subpage_snippet": "", "source": "kuscholarworks.ku.edu", "link": "https://kuscholarworks.ku.edu/server/api/core/bitstreams/662896a1-9b5c-4624-8d1e-23f7cd5dfe9a/content", "content": "by B Quanz · 2012 · Cited by 2 — . Algorithm 2 Testing using the Feature Generation Network. Input: A set of ... here we propose a simple and efficient approach for the synthetic data."}
+{"idx": 4, "title": "A Systematic Survey of Sparse Clustering", "date": "", "ddg_snippet": "by JBM Benjamin · 2025 — likelihood approach to clustering can be used [1], such as the expectation and maximization (EM) algorithm [2 ], [3]. The nonparametric ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/iel8/6287639/10820123/11015960.pdf", "content": "by JBM Benjamin · 2025 — likelihood approach to clustering can be used [1], such as the expectation and maximization (EM) algorithm [2 ], [3]. The nonparametric ..."}
+{"idx": 5, "title": "An effective imputation approach for handling missing data ...", "date": "", "ddg_snippet": "by K Sethia · 2025 · Cited by 1 — The proposed MDI algorithms LI-IIFCM and LI-IIFCM-σ, are also discussed in this section as Proposed Algorithm 1 and Proposed Algorithm 2 , ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s10791-025-09639-6", "content": "by K Sethia · 2025 · Cited by 1 — The proposed MDI algorithms LI-IIFCM and LI-IIFCM-σ, are also discussed in this section as Proposed Algorithm 1 and Proposed Algorithm 2 , ..."}
+{"idx": 6, "title": "Multiview Spectral Clustering With Bipartite Graph - Wei Xia", "date": "", "ddg_snippet": "by H Yang · 2022 · Cited by 84 — The whole algorithm is summarized in Algorithm 2 . C. Convergence Analysis. The convergence analysis of Algorithm 1 and 2 are quite similar, for simplicity ... 15 pages", "subpage_snippet": "", "source": "xdweixia.github.io", "link": "https://xdweixia.github.io/assets/pdf/2022-TIP.pdf", "content": "by H Yang · 2022 · Cited by 84 — The whole algorithm is summarized in Algorithm 2 . C. Convergence Analysis. The convergence analysis of Algorithm 1 and 2 are quite similar, for simplicity ... 15 pages"}
+{"idx": 7, "title": "Optimization of Density Peak Clustering Algorithm Based ...", "date": "", "ddg_snippet": "by H Huang · 2023 · Cited by 3 — The BWDPC algorithm ( Algorithm 2 ) is as follows. Algorithm 2 : BWDPC algorithm. Input: Experimental Dataset X = {x1, x2, …, xn} Output: ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11154350/", "content": "by H Huang · 2023 · Cited by 3 — The BWDPC algorithm ( Algorithm 2 ) is as follows. Algorithm 2 : BWDPC algorithm. Input: Experimental Dataset X = {x1, x2, …, xn} Output: ..."}
+{"idx": 8, "title": "Separated Cluster - an overview", "date": "", "ddg_snippet": "... simple and efficient algorithms for accurately exploring shared information in multiview data. ... The S-link algorithm is illustrated in Algorithm 2 . Sign in to ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/topics/computer-science/separated-cluster", "content": "... simple and efficient algorithms for accurately exploring shared information in multiview data. ... The S-link algorithm is illustrated in Algorithm 2 . Sign in to ..."}
+{"idx": 9, "title": "SimplePractice", "date": "", "ddg_snippet": "We would like to show you a description here but the site won’t allow us.", "subpage_snippet": "", "source": "secure.simplepractice.com", "link": "https://secure.simplepractice.com/users/sign_in", "content": "We would like to show you a description here but the site won’t allow us."}
diff --git a/data/sampled_jsons/Song_et_al._2021_Score-based_Generative_Models_abstract.jsonl b/data/sampled_jsons/Song_et_al._2021_Score-based_Generative_Models_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9e8fa789e7e89e5189bf6cd32f7bcb556aa4e588
--- /dev/null
+++ b/data/sampled_jsons/Song_et_al._2021_Score-based_Generative_Models_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Score-Based Generative Modeling through Stochastic ...", "date": "", "ddg_snippet": "by Y Song · 2020 · Cited by 8744 — Abstract:Creating noise from data is easy; creating data from noise is generative modeling . We present a stochastic differential equation ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2011.13456", "content": "by Y Song · 2020 · Cited by 8744 — Abstract:Creating noise from data is easy; creating data from noise is generative modeling . We present a stochastic differential equation ..."}
+{"idx": 1, "title": "Score-based Generative Modeling in Latent Space", "date": "", "ddg_snippet": "by A Vahdat · Cited by 790 — Here, we propose the Latent Score - based Generative . Model (LSGM), a novel approach that trains SGMs in a latent space, relying on the variational autoencoder ... 16 pages", "subpage_snippet": "", "source": "proceedings.nips.cc", "link": "https://proceedings.nips.cc/paper/2021/file/5dca4c6b9e244d24a30b4c45601d9720-Paper.pdf", "content": "by A Vahdat · Cited by 790 — Here, we propose the Latent Score - based Generative . Model (LSGM), a novel approach that trains SGMs in a latent space, relying on the variational autoencoder ... 16 pages"}
+{"idx": 2, "title": "SCORE-BASED GENERATIVE MODELING THROUGH ...", "date": "", "ddg_snippet": "by Y Song · Cited by 8744 — Score matching with Langevin dynamics (SMLD) ( Song & Ermon,. 2019) estimates the score (i.e., the gradient of the log probability density with respect to data) ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/ef0eadbe07115b0853e964f17aa09d811cd490f1.pdf?ref=news-tutorials-ai-research", "content": "by Y Song · Cited by 8744 — Score matching with Langevin dynamics (SMLD) ( Song & Ermon,. 2019) estimates the score (i.e., the gradient of the log probability density with respect to data) ..."}
+{"idx": 3, "title": "Official code for Score-Based Generative Modeling through ...", "date": "", "ddg_snippet": "We propose a unified framework that generalizes and improves previous work on score-based generative models through the lens of stochastic differential ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/yang-song/score_sde", "content": "We propose a unified framework that generalizes and improves previous work on score-based generative models through the lens of stochastic differential ..."}
+{"idx": 4, "title": "[2110.00473] Score-Based Generative Classifiers", "date": "", "ddg_snippet": "by RS Zimmermann · 2021 · Cited by 87 — Our work highlights that score-based generative models are closing the gap in classification accuracy compared to standard discriminative models.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2110.00473", "content": "by RS Zimmermann · 2021 · Cited by 87 — Our work highlights that score-based generative models are closing the gap in classification accuracy compared to standard discriminative models."}
+{"idx": 5, "title": "Score-based generative modeling secretly minimizes the ...", "date": "", "ddg_snippet": "by D Kwon · 2022 · Cited by 73 — Abstract. Score-based generative models are shown to achieve remarkable empirical performances in various applications such as image generation ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3600270.3601739", "content": "by D Kwon · 2022 · Cited by 73 — Abstract. Score-based generative models are shown to achieve remarkable empirical performances in various applications such as image generation ..."}
+{"idx": 6, "title": "Diffusion Schrödinger Bridge with Applications to Score- ...", "date": "", "ddg_snippet": "by V De Bortoli · 2021 · Cited by 635 — The first DSB iteration recovers the methodology proposed by Song et al . ( 2021 ), with the flexibility of using shorter time intervals, as subsequent DSB ...", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper/2021/hash/940392f5f32a7ade1cc201767cf83e31-Abstract.html", "content": "by V De Bortoli · 2021 · Cited by 635 — The first DSB iteration recovers the methodology proposed by Song et al . ( 2021 ), with the flexibility of using shorter time intervals, as subsequent DSB ..."}
+{"idx": 7, "title": "Convergence of score-based generative modeling for general ...", "date": "", "ddg_snippet": "by H Lee · 2023 · Cited by 181 — Abstract. Score-based generative modeling (SGM ) has grown to be a hugely successful method for learning to generate samples from complex data distributions ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v201/lee23a/lee23a.pdf", "content": "by H Lee · 2023 · Cited by 181 — Abstract. Score-based generative modeling (SGM ) has grown to be a hugely successful method for learning to generate samples from complex data distributions ..."}
+{"idx": 8, "title": "Score-based generative models: Score matching", "date": "", "ddg_snippet": "Let us start with the following quote from the abstract of ( Song et al ., 2020):. \"Creating noise from data is easy; creating data from noise is generative ...", "subpage_snippet": "", "source": "jmtomczak.github.io", "link": "https://jmtomczak.github.io/blog/16/16_score_matching.html", "content": "Let us start with the following quote from the abstract of ( Song et al ., 2020):. \"Creating noise from data is easy; creating data from noise is generative ..."}
+{"idx": 9, "title": "Score-based Generative Modeling in Latent Space", "date": "", "ddg_snippet": "by A Vahdat · Cited by 790 — Abstract: Score-based generative models (SGMs ) have recently demonstrated impressive results in terms of both sample quality and ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=P9TYG0j-wtG", "content": "by A Vahdat · Cited by 790 — Abstract: Score-based generative models (SGMs ) have recently demonstrated impressive results in terms of both sample quality and ..."}
diff --git a/data/sampled_jsons/Spring_Li_et_al._2024_sharding_blockchain_Spring_state_placement_abstract_year_2024.jsonl b/data/sampled_jsons/Spring_Li_et_al._2024_sharding_blockchain_Spring_state_placement_abstract_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..48e1d7b69eef5f5583c4cf27cf1da16d6c6e5888
--- /dev/null
+++ b/data/sampled_jsons/Spring_Li_et_al._2024_sharding_blockchain_Spring_state_placement_abstract_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Improving the Throughput of Sharding Blockchain via Deep ...", "date": "", "ddg_snippet": "13 May 2024 — In this paper, we present SPRING , the first deep-reinforcement-learning(DRL)-based sharding framework for state placement.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1145/3589334.3645386", "content": "13 May 2024 — In this paper, we present SPRING , the first deep-reinforcement-learning(DRL)-based sharding framework for state placement."}
+{"idx": 1, "title": "Improving the Throughput of Sharding Blockchain via Deep ...", "date": "", "ddg_snippet": "by P Li · 2024 · Cited by 21 — SPRING is a deep-reinforcement-learning framework for state placement in sharding blockchains, aiming to reduce cross-shard transactions and ...", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/WWW_Spring_camera_ready.pdf", "content": "by P Li · 2024 · Cited by 21 — SPRING is a deep-reinforcement-learning framework for state placement in sharding blockchains, aiming to reduce cross-shard transactions and ..."}
+{"idx": 2, "title": "Improving the Throughput of Sharding Blockchain via Deep ...", "date": "", "ddg_snippet": "by P Li · Cited by 21 — In this paper, we present Spring , the first deep- reinforcement-learning(DRL)-based sharding framework for state placement . Spring formulates the state ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=8oczaP1YKD", "content": "by P Li · Cited by 21 — In this paper, we present Spring , the first deep- reinforcement-learning(DRL)-based sharding framework for state placement . Spring formulates the state ..."}
+{"idx": 3, "title": "Building blocks of sharding blockchain systems: Concepts, ...", "date": "", "ddg_snippet": "by Y Liu · 2022 · Cited by 147 — We conduct a systematic analysis of existing sharding blockchain systems and derive a conceptual decomposition of their architecture into functional components.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S1574013722000478", "content": "by Y Liu · 2022 · Cited by 147 — We conduct a systematic analysis of existing sharding blockchain systems and derive a conceptual decomposition of their architecture into functional components."}
+{"idx": 4, "title": "Enhancing Sharding Blockchain via Deep Reinforcement ...", "date": "", "ddg_snippet": "22 Apr 2025 — 2024. SPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement. In Proceedings of the ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1145/3696410.3714926", "content": "22 Apr 2025 — 2024. SPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement. In Proceedings of the ..."}
+{"idx": 5, "title": "ContribChain: A Stress-Balanced Blockchain Sharding ...", "date": "", "ddg_snippet": "by X Huang · 2025 — ContribChain is a blockchain sharding protocol that automatically considers node contributions to achieve stress balance, addressing the issue ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2505.06899", "content": "by X Huang · 2025 — ContribChain is a blockchain sharding protocol that automatically considers node contributions to achieve stress balance, addressing the issue ..."}
+{"idx": 6, "title": "Enhancing Sharding Blockchain via Deep Reinforcement ...", "date": "", "ddg_snippet": "by M Song · 2025 · Cited by 2 — Spring shows a higher CSTX ratio compared to AERO, primarily because its approach focuses more on optimizing state placement rather than account migration. 11 pages", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/WWW_AERO_camera_ready.pdf", "content": "by M Song · 2025 · Cited by 2 — Spring shows a higher CSTX ratio compared to AERO, primarily because its approach focuses more on optimizing state placement rather than account migration. 11 pages"}
+{"idx": 7, "title": "A new trust-based, DRL-driven framework for blockchain ...", "date": "", "ddg_snippet": "by Z Zhang · 2024 · Cited by 20 — This paper introduces a Trust-based and DRL-driven (TbDd) framework, crafted to counter collusion attack risks and dynamically adjust node allocation.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S1389128624001750", "content": "by Z Zhang · 2024 · Cited by 20 — This paper introduces a Trust-based and DRL-driven (TbDd) framework, crafted to counter collusion attack risks and dynamically adjust node allocation."}
+{"idx": 8, "title": "BlockEmulator: An Emulator Enabling to Test Blockchain ...", "date": "", "ddg_snippet": "by H Huang · 2025 · Cited by 34 — We developed BlockEmulator, which is designed as an experimental platform, particularly for emulating blockchain sharding mechanisms.", "subpage_snippet": "", "source": "www.computer.org", "link": "https://www.computer.org/csdl/journal/sc/2025/02/10908689/24MWrrwc2Fa", "content": "by H Huang · 2025 · Cited by 34 — We developed BlockEmulator, which is designed as an experimental platform, particularly for emulating blockchain sharding mechanisms."}
+{"idx": 9, "title": "A self evolving high performance sharded consortium ...", "date": "", "ddg_snippet": "by X Zhu · 2025 — To overcome this, we present a self-evolving high-performance sharded consortium blockchain (EvoChain) designed for 6G networks. EvoChain utilizes subjective ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12316968/", "content": "by X Zhu · 2025 — To overcome this, we present a self-evolving high-performance sharded consortium blockchain (EvoChain) designed for 6G networks. EvoChain utilizes subjective ..."}
diff --git a/data/sampled_jsons/Statistical_Collusion_Rs(k)_sqrt_log.jsonl b/data/sampled_jsons/Statistical_Collusion_Rs(k)_sqrt_log.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ea11d57c5aa4b3fac426d9d880a9aef5bbb5e358
--- /dev/null
+++ b/data/sampled_jsons/Statistical_Collusion_Rs(k)_sqrt_log.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Dan Boneh", "date": "", "ddg_snippet": "Many existing folding protocols rely on the discrete- log based Pedersen commitment scheme, and are therefore not post-quantum secure and require a ...", "subpage_snippet": "", "source": "www.iacr.org", "link": "https://www.iacr.org/cryptodb/data/author.php?authorkey=230", "content": "Many existing folding protocols rely on the discrete- log based Pedersen commitment scheme, and are therefore not post-quantum secure and require a ..."}
+{"idx": 1, "title": "Ben C Davis", "date": "", "ddg_snippet": "... and instances of that problem are key. ... Induction is a chain of assumption that logically flow from one to another, starting with a base case.", "subpage_snippet": "", "source": "bencdavis.com", "link": "https://bencdavis.com/reading/adm", "content": "... and instances of that problem are key. ... Induction is a chain of assumption that logically flow from one to another, starting with a base case."}
+{"idx": 2, "title": "(PDF) TMVA-Toolkit for Multivariate Data Analysis - Academia.edu Cs-pbft: a comprehensive scoring-based Practical Byzantine ... H.Schulze, C.Luders - Theory and Applications of OFDM and CDMA BOOK PRACTICAL HANDBOOK OF GENETIC ALGORITHMS.pdf Digital Media Processing: Dsp Algorithms Using C [PDF ... Digital Media Processing: DSP Algorithms Using C - PDF Free ... Full text of \"Queen's Journal 1946\" - Archive.org", "date": "", "ddg_snippet": "In high-energy physics, with the search for ever smaller signals in ever larger data sets, it has become essential to extract a maximum of the available information from the data. Multivariate classification methods based on machine learning May 14, 2025 · Practical Byzantine Fault Tolerance (PBFT) is one of the most important consensus algorithms in distributed systems, which can effectively respond to the threat of malicious nodes. However, PBFT still has shortcomings in terms of arbitrary selection of primary nodes, high communication overhead, and lack of reward and punishment mechanisms. To address this problem, in this paper, we propose a ... The concept of log -likelihood ratios is well established in probability theory. However, their usefulness for channel coding and their intuitively amazing visualization has been established in great parts by Hagenauer and coworkers (Hagenauer 1988, 1995; Hagenauer and Hoeher 1989; Hagenauer et al. 1996). The Practical Handbook of Genetic Algorithms serves as an updated resource for practitioners and researchers in the field of Genetic Algorithms (GAs). It revisits foundational concepts while incorporating recent advancements and applications in the domain. This second edition features contributions across various chapters, each addressing specific challenges and techniques, including fuzzy ... Digital Media Processing: Dsp Algorithms Using C [PDF] [7b4pvl45g2e0]. Multimedia processing demands efficient programming in order to optimize functionality. Data, image, audio, and video pr... Digital Media Processing DSP Algorithms Using CHazarathaiah MalepatiAMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YO... Full text of \"Queen's Journal 1946\" See other formats ©brarg KINGSTON. ONTARIO Published Weekly by the Alma Mater Society of Queen's University VOL. LXXIII KINGSTON, ONTARIO, FRIDAY, APRIL 26. 1946 No. 1 PREMIER IN QUEEN'S HONOUR LIST First Building Of Expansion Project Started I^igging Operations £or the Gordon Hall Wing A strain on chemistry facilities which have heen taxed to the limit ...", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/62101608/TMVA_Toolkit_for_Multivariate_Data_Analysis", "content": "In high-energy physics, with the search for ever smaller signals in ever larger data sets, it has become essential to extract a maximum of the available information from the data. Multivariate classification methods based on machine learning May 14, 2025 · Practical Byzantine Fault Tolerance (PBFT) is one of the most important consensus algorithms in distributed systems, which can effectively respond to the threat of malicious nodes. However, PBFT still has shortcomings in terms of arbitrary selection of primary nodes, high communication overhead, and lack of reward and punishment mechanisms. To address this problem, in this paper, we propose a ... The concept of log -likelihood ratios is well established in probability theory. However, their usefulness for channel coding and their intuitively amazing visualization has been established in great parts by Hagenauer and coworkers (Hagenauer 1988, 1995; Hagenauer and Hoeher 1989; Hagenauer et al. 1996). The Practical Handbook of Genetic Algorithms serves as an updated resource for practitioners and researchers in the field of Genetic Algorithms (GAs). It revisits foundational concepts while incorporating recent advancements and applications in the domain. This second edition features contributions across various chapters, each addressing specific challenges and techniques, including fuzzy ... Digital Media Processing: Dsp Algorithms Using C [PDF] [7b4pvl45g2e0]. Multimedia processing demands efficient programming in order to optimize functionality. Data, image, audio, and video pr... Digital Media Processing DSP Algorithms Using CHazarathaiah MalepatiAMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YO... Full text of \"Queen's Journal 1946\" See other formats ©brarg KINGSTON. ONTARIO Published Weekly by the Alma Mater Society of Queen's University VOL. LXXIII KINGSTON, ONTARIO, FRIDAY, APRIL 26. 1946 No. 1 PREMIER IN QUEEN'S HONOUR LIST First Building Of Expansion Project Started I^igging Operations £or the Gordon Hall Wing A strain on chemistry facilities which have heen taxed to the limit ..."}
+{"idx": 3, "title": "H.Schulze, C.Luders - Theory and Applications of OFDM and CDMA", "date": "", "ddg_snippet": "The concept of log -likelihood ratios is well established in probability theory. However, their usefulness for channel coding and their intuitively amazing visualization has been established in great parts by Hagenauer and coworkers (Hagenauer 1988, 1995; Hagenauer and Hoeher 1989; Hagenauer et al. 1996).", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/36459811/H_Schulze_C_Luders_Theory_and_Applications_of_OFDM_and_CDMA", "content": "The concept of log -likelihood ratios is well established in probability theory. However, their usefulness for channel coding and their intuitively amazing visualization has been established in great parts by Hagenauer and coworkers (Hagenauer 1988, 1995; Hagenauer and Hoeher 1989; Hagenauer et al. 1996)."}
+{"idx": 4, "title": "IACR News", "date": "", "ddg_snippet": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ...", "subpage_snippet": "", "source": "www.iacr.org", "link": "https://www.iacr.org/news/index.php?page=23&archive=1", "content": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ..."}
+{"idx": 5, "title": "IACR News", "date": "", "ddg_snippet": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ...", "subpage_snippet": "", "source": "www.iacr.org", "link": "https://www.iacr.org/news/index.php?page=13&archive=1", "content": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ..."}
+{"idx": 6, "title": "IACR News", "date": "", "ddg_snippet": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ...", "subpage_snippet": "", "source": "www.iacr.org", "link": "https://www.iacr.org/news/index.php?page=21&archive=1", "content": "Our experiments show that the latency of our $64$ bit multiplication is $\\approx 6$ times faster than Kim and slightly faster than TFHE- rs , while ..."}
+{"idx": 7, "title": "IACR News", "date": "", "ddg_snippet": "We analyze Kaneko's bound to prove that, away from the $j$-invariant $0$, edges of multiplicity at least three can occur in the supersingular $\\ell ...", "subpage_snippet": "", "source": "www.iacr.org", "link": "https://www.iacr.org/news/index.php?page=13", "content": "We analyze Kaneko's bound to prove that, away from the $j$-invariant $0$, edges of multiplicity at least three can occur in the supersingular $\\ell ..."}
+{"idx": 8, "title": "NeurIPS 2024 Spotlight Posters", "date": "", "ddg_snippet": "... train expressive (conditional) diffusion models, enabling us to sample neural activity with realistic single-neuron and population spiking statistics .", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2024/events/spotlight-posters-2024", "content": "... train expressive (conditional) diffusion models, enabling us to sample neural activity with realistic single-neuron and population spiking statistics ."}
+{"idx": 9, "title": "Detection and prediction of insider threats to cyber security:", "date": "", "ddg_snippet": "... key challenges facing the insider threat detection and prediction system include unbounded patterns, uneven time lags between activities, data ...", "subpage_snippet": "", "source": "bdataanalytics.biomedcentral.com", "link": "https://bdataanalytics.biomedcentral.com/articles/10.1186/s41044-016-0006-0", "content": "... key challenges facing the insider threat detection and prediction system include unbounded patterns, uneven time lags between activities, data ..."}
diff --git "a/data/sampled_jsons/Statistical_Collusion_by_Collectives_on_Learning_Platforms_Theorem_3.3_R\316\264(n).jsonl" "b/data/sampled_jsons/Statistical_Collusion_by_Collectives_on_Learning_Platforms_Theorem_3.3_R\316\264(n).jsonl"
new file mode 100644
index 0000000000000000000000000000000000000000..a0476b8ac27e7623c4f98e4dd7738c9e105c3b97
--- /dev/null
+++ "b/data/sampled_jsons/Statistical_Collusion_by_Collectives_on_Learning_Platforms_Theorem_3.3_R\316\264(n).jsonl"
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Coopetition - Wikipedia", "date": "", "ddg_snippet": "... received a patent titled \"Method for Creating Coopertition\" (spelled as one word, with no hyphen), which involves giving FIRST Robotics teams some ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Coopetition", "content": "... received a patent titled \"Method for Creating Coopertition\" (spelled as one word, with no hyphen), which involves giving FIRST Robotics teams some ..."}
+{"idx": 1, "title": "Regulation of Algorithmic Collusion", "date": "", "ddg_snippet": "... be checked from data logged by the algorithm while deployed to prove statistically that the algorithm is not colluding under reasonable assumptions.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2401.15794v2", "content": "... be checked from data logged by the algorithm while deployed to prove statistically that the algorithm is not colluding under reasonable assumptions."}
+{"idx": 2, "title": "Federated Hypergraph Learning with Local Differential Privacy:", "date": "", "ddg_snippet": "In [ 16 ] , authors introduced HyperSAGE, an inductive framework that generalizes representation learning on hypergraphs using a two-level neural ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2408.05160v3", "content": "In [ 16 ] , authors introduced HyperSAGE, an inductive framework that generalizes representation learning on hypergraphs using a two-level neural ..."}
+{"idx": 3, "title": "Secret Collusion among AI Agents: Multi-Agent Deception via", "date": "", "ddg_snippet": "... of these applications are classified as “high-risk” by legislators due to their potential impact on human lives and use of sensitive data ( 35 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2402.07510v4", "content": "... of these applications are classified as “high-risk” by legislators due to their potential impact on human lives and use of sensitive data ( 35 ..."}
+{"idx": 4, "title": "Algorithmic Collusion in Auctions: Evidence from Controlled", "date": "", "ddg_snippet": "... on constant valuations with only two or three bidders, used relatively simple exploration mechanisms, and did not thoroughly parse out how algorithmic ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2306.09437v2", "content": "... on constant valuations with only two or three bidders, used relatively simple exploration mechanisms, and did not thoroughly parse out how algorithmic ..."}
+{"idx": 5, "title": "Algorithmic Collective Action with Multiple Collectives", "date": "", "ddg_snippet": "For each regime, we derive lower bounds on the per- collective and global success, revealing interesting trade-offs driven by the interplay of the ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.19149v1", "content": "For each regime, we derive lower bounds on the per- collective and global success, revealing interesting trade-offs driven by the interplay of the ..."}
+{"idx": 6, "title": "Montréal | Freakonometrics", "date": "", "ddg_snippet": "Next week, we organize the first Montréal – Guanajuato Workshop on Probability and Machine Learning , at CIMAT, Centro de Investigación en ...", "subpage_snippet": "", "source": "freakonometrics.hypotheses.org", "link": "https://freakonometrics.hypotheses.org/tag/montreal", "content": "Next week, we organize the first Montréal – Guanajuato Workshop on Probability and Machine Learning , at CIMAT, Centro de Investigación en ..."}
+{"idx": 7, "title": "mutatis mutandis | Freakonometrics", "date": "", "ddg_snippet": "We propose here a mutatis mutandis version of the CATE, which will be done simply in dimension one by saying that the CATE must be computed relative ...", "subpage_snippet": "", "source": "freakonometrics.hypotheses.org", "link": "https://freakonometrics.hypotheses.org/tag/mutatis-mutandis", "content": "We propose here a mutatis mutandis version of the CATE, which will be done simply in dimension one by saying that the CATE must be computed relative ..."}
+{"idx": 8, "title": "Publications - Chair of Decision Science & Systems", "date": "", "ddg_snippet": "By clicking on „enable search“ you ... In Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI) , pages 9537–9545, 2024.", "subpage_snippet": "", "source": "www.cs.cit.tum.de", "link": "https://www.cs.cit.tum.de/en/dss/publications/", "content": "By clicking on „enable search“ you ... In Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI) , pages 9537–9545, 2024."}
+{"idx": 9, "title": "The Knowledge Complexity of Interactive Proof Systems | SIAM", "date": "", "ddg_snippet": "... residuosity and 'quadratic nonresiduosity. ... Rackoff, How to simultaneously exchange a secret bit by flipping a symmetrically-biased coin, Proc.", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/abs/10.1137/0218012", "content": "... residuosity and 'quadratic nonresiduosity. ... Rackoff, How to simultaneously exchange a secret bit by flipping a symmetrically-biased coin, Proc."}
diff --git a/data/sampled_jsons/Symmetric_Cross_Entropy_SCE_=_alphaCE_+_betaRCE_formula.jsonl b/data/sampled_jsons/Symmetric_Cross_Entropy_SCE_=_alphaCE_+_betaRCE_formula.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..42968b618f13169d5180154d7147774de50c4e1f
--- /dev/null
+++ b/data/sampled_jsons/Symmetric_Cross_Entropy_SCE_=_alphaCE_+_betaRCE_formula.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Symmetric Cross Entropy for Robust Learning with Noisy Labels", "date": "", "ddg_snippet": "Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy ( RCE ). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content_ICCV_2019/papers/Wang_Symmetric_Cross_Entropy_for_Robust_Learning_With_Noisy_Labels_ICCV_2019_paper.pdf", "content": "Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy ( RCE ). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels."}
+{"idx": 1, "title": "Symmetric Cross Entropy for Robust Learning with Noisy Labels", "date": "", "ddg_snippet": "Aug 16, 2019 · Inspired by the symmetric KL-divergence, we propose the approach of \\textbf {Symmetric cross entropy Learning} (SL), boosting CE symmetrically with a noise robust counterpart Reverse Cross Entropy ( RCE ). Our proposed SL approach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy labels.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1908.06112", "content": "Aug 16, 2019 · Inspired by the symmetric KL-divergence, we propose the approach of \\textbf {Symmetric cross entropy Learning} (SL), boosting CE symmetrically with a noise robust counterpart Reverse Cross Entropy ( RCE ). Our proposed SL approach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy labels."}
+{"idx": 2, "title": "【SCE 损失】Symmetric Cross Entropy for Robust Learning with ... Hierarchical symmetric cross entropy for distant supervised ... Symmetric Cross Entropy for Robust Learning with Noisy Labels symmetric_cross_entropy_for_noisy_labels/README.md ... - GitHub Why don't we use a symmetric cross-entropy loss? Symmetric Cross Entropy for Robust Learning with Noisy Labels Hierarchical symmetric cross entropy for distant supervised relation Hierarchical symmetric cross entropy for distant supervised relation Symmetric Cross Entropy for Robust Learning with Noisy Labels Hierarchical symmetric cross entropy for distant supervised relation Hierarchical symmetric cross entropy for distant supervised relation A. Proof for Theorem 1 - CVF Open Access", "date": "", "ddg_snippet": "去杭州之前就想专门整理一下在有噪声样本下的分类损失，在这之前，先就较经典的一篇分析一下各种推导原理，剩下就简单一点只看损失部分～ See full list on zhuanlan.zhihu.com 在分类任务上，最普遍的损失函数是 Cross Entropy，即交叉熵损失： 该损失可以直观理解成努力提高样本对应标签类别的预测概率值。但是当标签中存在噪声和不准确时，这个严格惩罚预测值和标签靠拢的函数就无法调整这种噪声带了的巨大精确度下降： 可以看到图片 (b) 为带噪声标签情况下的交叉损失精度，大致有百分之十的精度下降，而那些本就不容易学习的类别精度更糟糕了。 2016 年 Label Smoothing Regularization 被提出，如今引用量已经一万五多了，文章信息： Rethinking the Inception Architecture for Computer Vision目标是提出一种正则项来缓解这种噪声带来的过拟合影响，理解起来很简单，就是重新改写原来严格的 0/1 one-hot 标签，公式如下： \\epsilon 为平滑项的参数， u(k) 一般为 1/K，于是可以写作： See full list on zhuanlan.zhihu.com 文章认为我们的模型本身就具有正确判断样本类别的能力，在噪声较多的情况下或许甚至比正确标签还准确，因此完全可以加入一个以模型预测为基点的损失部分。受启发与对称 KL： 文章提出了对称的交叉熵损失，即： 损失函数很好理解啦，就是将标注标签和预测值反过来。而由于标注标签是 one-hot 的，大量标签是 0，因此为了计算 log 0，将此部分取了一个常数 A < 0，即是一个惩罚项。另外为了进一步提高鲁棒性和自由度，这个函数可以额外加入两个超参数： 可以看一下实现代码： See full list on zhuanlan.zhihu.com 首先证明鲁棒性，即 noise-tolerant，作者认为如果在干净样本条件下和噪声样本条件下训练出来的最优模型 f^{*} 有着同样的分布概率，那说明这个损失函数对噪声是鲁棒的。 我们只需要考虑新提出的 RCE 部分，先定义两个期望损失： \\eta 为噪声存在的概率，之后推导公式为： 我加了自己理解的部分标注 于是最优解的差就可以推导为： 因此，想满足条件，就是使上面参数大于 0，于是： See full list on zhuanlan.zhihu.com 在 CIFAR-10 数据集上采用 60% 对称噪声的实验结果在 CIFAR-10 数据集上采用 40% 对称噪声的实验结果：置信度得到了大幅提升；类别正确数亦然在 CIFAR-10 数据集上采用 60% 对称噪声的可视化结果在 CIFAR-10 数据集上采用 60% 对称噪声：对 A 和 alpha 的消融实验在 CIFAR-10 数据集上采用 60% 对称噪声：与其他损失函数的精度比较丰富的实验结果～ See full list on zhuanlan.zhihu.com Sep 3, 2024 · The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels. Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy ( RCE ). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels. @inproceedings {wang2019symmetric, title= {Symmetric cross entropy for robust learning with noisy labels}, author= {Wang, Yisen and Ma, Xingjun and Chen, Zaiyi and Luo, Yuan and Yi, Jinfeng and Bailey, James}, booktitle= {IEEE International Conference on Computer Vision}, year= {2019} } Mar 6, 2018 · The minimum value that the cross-entropy of ℍ [𝑝,𝑞] can have is when 𝑞=𝑝 which is ℍ [𝑝,𝑝], simple the entropy of the distribution 𝑝. While evaluating different built models say 𝑞 and 𝑞', we often need to compare different them, and cross-entropy can be used here. What is symmetric cross entropy learning? Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy (RCE). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels. What is a symmetric cross entropy loss function? The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels. Since the introduction of RCE makes SCE noise-tolerant, this section theoretically analyzes and derives the loss function \\ (l_ { rce }\\). What is symmetric cross entropy loss in DSRE? Since the Cross Entropy (CE) loss function is easily affected by errors caused by noisy labels, this paper introduces the Symmetric Cross Entropy (SCE) into the DSRE task. It serves as the Classification Symmetric Cross Entropy loss \\ (l_ {cla}\\) to guide the optimization direction from within the extractor. \\ (l_ {cla}\\) is defined as: What is generalized cross entropy (GCE) loss? Generalized Cross Entropy (GCE) loss applies a Box-Cox transformation to prob-abilities (power law function of probability with exponent q) and can behave like a weighted MAE. Label Smoothing Regularization (LSR) [21, 17] is another technique using soft labels in place of one-hot labels to alleviate overfitting to noisy labels. How effective is consistent symmetric cross entropy in hscere? The Classification Symmetric Cross Entropy can effectively improve the noise resistance within the relation extractor. Table 3 displays the experimental results assessing the effectiveness of Consistent Symmetric Cross Entropy in HSCERE. What is hierarchical symmetric cross entropy (hsce) loss? We provide a joint loss , namely Hierarchical Symmetric Cross Entropy (HSCE) loss. HSCE consists of two components: Classification Symmetric Cross Entropy from within each relation extractor and Consistency Symmetric Cross Entropy between two relation extractors. A. Proof for Theorem 1 Theorem 1. In a multi-class classification problem, ` rce is noise tolerant under symmetric or uniform label noise if noise rate < 1 1", "subpage_snippet": "", "source": "zhuanlan.zhihu.com", "link": "https://zhuanlan.zhihu.com/p/420827592", "content": "去杭州之前就想专门整理一下在有噪声样本下的分类损失，在这之前，先就较经典的一篇分析一下各种推导原理，剩下就简单一点只看损失部分～ See full list on zhuanlan.zhihu.com 在分类任务上，最普遍的损失函数是 Cross Entropy，即交叉熵损失： 该损失可以直观理解成努力提高样本对应标签类别的预测概率值。但是当标签中存在噪声和不准确时，这个严格惩罚预测值和标签靠拢的函数就无法调整这种噪声带了的巨大精确度下降： 可以看到图片 (b) 为带噪声标签情况下的交叉损失精度，大致有百分之十的精度下降，而那些本就不容易学习的类别精度更糟糕了。 2016 年 Label Smoothing Regularization 被提出，如今引用量已经一万五多了，文章信息： Rethinking the Inception Architecture for Computer Vision目标是提出一种正则项来缓解这种噪声带来的过拟合影响，理解起来很简单，就是重新改写原来严格的 0/1 one-hot 标签，公式如下： \\epsilon 为平滑项的参数， u(k) 一般为 1/K，于是可以写作： See full list on zhuanlan.zhihu.com 文章认为我们的模型本身就具有正确判断样本类别的能力，在噪声较多的情况下或许甚至比正确标签还准确，因此完全可以加入一个以模型预测为基点的损失部分。受启发与对称 KL： 文章提出了对称的交叉熵损失，即： 损失函数很好理解啦，就是将标注标签和预测值反过来。而由于标注标签是 one-hot 的，大量标签是 0，因此为了计算 log 0，将此部分取了一个常数 A < 0，即是一个惩罚项。另外为了进一步提高鲁棒性和自由度，这个函数可以额外加入两个超参数： 可以看一下实现代码： See full list on zhuanlan.zhihu.com 首先证明鲁棒性，即 noise-tolerant，作者认为如果在干净样本条件下和噪声样本条件下训练出来的最优模型 f^{*} 有着同样的分布概率，那说明这个损失函数对噪声是鲁棒的。 我们只需要考虑新提出的 RCE 部分，先定义两个期望损失： \\eta 为噪声存在的概率，之后推导公式为： 我加了自己理解的部分标注 于是最优解的差就可以推导为： 因此，想满足条件，就是使上面参数大于 0，于是： See full list on zhuanlan.zhihu.com 在 CIFAR-10 数据集上采用 60% 对称噪声的实验结果在 CIFAR-10 数据集上采用 40% 对称噪声的实验结果：置信度得到了大幅提升；类别正确数亦然在 CIFAR-10 数据集上采用 60% 对称噪声的可视化结果在 CIFAR-10 数据集上采用 60% 对称噪声：对 A 和 alpha 的消融实验在 CIFAR-10 数据集上采用 60% 对称噪声：与其他损失函数的精度比较丰富的实验结果～ See full list on zhuanlan.zhihu.com Sep 3, 2024 · The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels. Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy ( RCE ). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels. @inproceedings {wang2019symmetric, title= {Symmetric cross entropy for robust learning with noisy labels}, author= {Wang, Yisen and Ma, Xingjun and Chen, Zaiyi and Luo, Yuan and Yi, Jinfeng and Bailey, James}, booktitle= {IEEE International Conference on Computer Vision}, year= {2019} } Mar 6, 2018 · The minimum value that the cross-entropy of ℍ [𝑝,𝑞] can have is when 𝑞=𝑝 which is ℍ [𝑝,𝑝], simple the entropy of the distribution 𝑝. While evaluating different built models say 𝑞 and 𝑞', we often need to compare different them, and cross-entropy can be used here. What is symmetric cross entropy learning? Inspired by the symmetric KL-divergence, we pro-pose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust coun-terpart Reverse Cross Entropy (RCE). Our proposed SL ap-proach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy la-bels. What is a symmetric cross entropy loss function? The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels. Since the introduction of RCE makes SCE noise-tolerant, this section theoretically analyzes and derives the loss function \\ (l_ { rce }\\). What is symmetric cross entropy loss in DSRE? Since the Cross Entropy (CE) loss function is easily affected by errors caused by noisy labels, this paper introduces the Symmetric Cross Entropy (SCE) into the DSRE task. It serves as the Classification Symmetric Cross Entropy loss \\ (l_ {cla}\\) to guide the optimization direction from within the extractor. \\ (l_ {cla}\\) is defined as: What is generalized cross entropy (GCE) loss? Generalized Cross Entropy (GCE) loss applies a Box-Cox transformation to prob-abilities (power law function of probability with exponent q) and can behave like a weighted MAE. Label Smoothing Regularization (LSR) [21, 17] is another technique using soft labels in place of one-hot labels to alleviate overfitting to noisy labels. How effective is consistent symmetric cross entropy in hscere? The Classification Symmetric Cross Entropy can effectively improve the noise resistance within the relation extractor. Table 3 displays the experimental results assessing the effectiveness of Consistent Symmetric Cross Entropy in HSCERE. What is hierarchical symmetric cross entropy (hsce) loss? We provide a joint loss , namely Hierarchical Symmetric Cross Entropy (HSCE) loss. HSCE consists of two components: Classification Symmetric Cross Entropy from within each relation extractor and Consistency Symmetric Cross Entropy between two relation extractors. A. Proof for Theorem 1 Theorem 1. In a multi-class classification problem, ` rce is noise tolerant under symmetric or uniform label noise if noise rate < 1 1"}
+{"idx": 3, "title": "Hierarchical symmetric cross entropy for distant supervised ...", "date": "", "ddg_snippet": "Sep 3, 2024 · The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s10489-024-05798-z", "content": "Sep 3, 2024 · The Symmetric Cross Entropy ( SCE ) loss function combines the Cross Entropy ( CE ) and Reverse Cross Entropy ( RCE ) to improve the robustness of the model under noisy labels."}
+{"idx": 4, "title": "symmetric_cross_entropy_for_noisy_labels/README.md ... - GitHub", "date": "", "ddg_snippet": "@inproceedings {wang2019symmetric, title= {Symmetric cross entropy for robust learning with noisy labels}, author= {Wang, Yisen and Ma, Xingjun and Chen, Zaiyi and Luo, Yuan and Yi, Jinfeng and Bailey, James}, booktitle= {IEEE International Conference on Computer Vision}, year= {2019} }", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/YisenWang/symmetric_cross_entropy_for_noisy_labels/blob/master/README.md", "content": "@inproceedings {wang2019symmetric, title= {Symmetric cross entropy for robust learning with noisy labels}, author= {Wang, Yisen and Ma, Xingjun and Chen, Zaiyi and Luo, Yuan and Yi, Jinfeng and Bailey, James}, booktitle= {IEEE International Conference on Computer Vision}, year= {2019} }"}
+{"idx": 5, "title": "Why don't we use a symmetric cross-entropy loss?", "date": "", "ddg_snippet": "Mar 6, 2018 · The minimum value that the cross-entropy of ℍ [𝑝,𝑞] can have is when 𝑞=𝑝 which is ℍ [𝑝,𝑝], simple the entropy of the distribution 𝑝. While evaluating different built models say 𝑞 and 𝑞', we often need to compare different them, and cross-entropy can be used here.", "subpage_snippet": "", "source": "stats.stackexchange.com", "link": "https://stats.stackexchange.com/questions/331942/why-dont-we-use-a-symmetric-cross-entropy-loss", "content": "Mar 6, 2018 · The minimum value that the cross-entropy of ℍ [𝑝,𝑞] can have is when 𝑞=𝑝 which is ℍ [𝑝,𝑝], simple the entropy of the distribution 𝑝. While evaluating different built models say 𝑞 and 𝑞', we often need to compare different them, and cross-entropy can be used here."}
+{"idx": 6, "title": "Symmetric Reinforcement Learning Loss for Robust ...", "date": "", "ddg_snippet": "27 May 2024 — Therefore, SCE combines CE and RCE losses (Equation 6),. Report issue for preceding element. L sce = α L ce + β L rce subscript L sce 𝛼 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2405.17618v1", "content": "27 May 2024 — Therefore, SCE combines CE and RCE losses (Equation 6),. Report issue for preceding element. L sce = α L ce + β L rce subscript L sce 𝛼 ..."}
+{"idx": 7, "title": "Can Cross Entropy Loss Be Robust to Label Noise?", "date": "", "ddg_snippet": "by L Feng · Cited by 220 — Symmetric Cross Entropy (SCE ) [Wang et al., 2019] combines CCE and Reverse Cross Entropy (RCE, which is equivalent to MAE) by tuning the regularization ... 7 pages", "subpage_snippet": "", "source": "www.ijcai.org", "link": "https://www.ijcai.org/proceedings/2020/0305.pdf", "content": "by L Feng · Cited by 220 — Symmetric Cross Entropy (SCE ) [Wang et al., 2019] combines CCE and Reverse Cross Entropy (RCE, which is equivalent to MAE) by tuning the regularization ... 7 pages"}
+{"idx": 8, "title": "Symmetric Reinforcement Learning Loss for Robust ...", "date": "", "ddg_snippet": "Symmetric Cross Entropy (SCE ) (Wang et al., 2019) is designed for noisy classification datasets. Cross Entropy (CE) loss (Equation 4) performs effectively when ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/44897", "content": "Symmetric Cross Entropy (SCE ) (Wang et al., 2019) is designed for noisy classification datasets. Cross Entropy (CE) loss (Equation 4) performs effectively when ..."}
+{"idx": 9, "title": "Asymmetric Loss Functions for Learning with Noisy Labels", "date": "", "ddg_snippet": "by X Zhou · 2021 · Cited by 91 — Inspired by the symmetric KL-divergence, the symmetric cross entropy ( SCE ) (Wang et al., 2019b) was proposed to combine CE with a noise tolerance term, namely ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "http://proceedings.mlr.press/v139/zhou21f/zhou21f.pdf", "content": "by X Zhou · 2021 · Cited by 91 — Inspired by the symmetric KL-divergence, the symmetric cross entropy ( SCE ) (Wang et al., 2019b) was proposed to combine CE with a noise tolerance term, namely ..."}
diff --git a/data/sampled_jsons/TOP-ERL_Transformer-based_Off-Policy_Episodic_Reinforcement_Learning_random_segment_length.jsonl b/data/sampled_jsons/TOP-ERL_Transformer-based_Off-Policy_Episodic_Reinforcement_Learning_random_segment_length.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a4caa4e848ebb418911f970e0372008a4ee9d115
--- /dev/null
+++ b/data/sampled_jsons/TOP-ERL_Transformer-based_Off-Policy_Episodic_Reinforcement_Learning_random_segment_length.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "TOP-ERL: Transformer-based Off-Policy Episodic ...", "date": "", "ddg_snippet": "by G Li · Cited by 7 — The approach retains the main architecture of TOP - ERL but reduces the segment length to one in critic updates. In this case, the transformer critic ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=N4NhVN30ph", "content": "by G Li · Cited by 7 — The approach retains the main architecture of TOP - ERL but reduces the segment length to one in critic updates. In this case, the transformer critic ..."}
+{"idx": 1, "title": "TOP-ERL: Transformer-based Off-Policy Episodic Reinforcement ...", "date": "", "ddg_snippet": "Abstract—This work introduces Transformer - based Off - Policy . Episodic Reinforcement Learning ( TOP - ERL ), a novel algorithm that enables off - policy updates in ...", "subpage_snippet": "", "source": "rss25-roboreps.github.io", "link": "https://rss25-roboreps.github.io/papers/1_TOP_ERL_Transformer_based_Of.pdf", "content": "Abstract—This work introduces Transformer - based Off - Policy . Episodic Reinforcement Learning ( TOP - ERL ), a novel algorithm that enables off - policy updates in ..."}
+{"idx": 2, "title": "transformer-based off-policy episodic reinforcement learning", "date": "", "ddg_snippet": "by G Li · 2024 · Cited by 7 — The results indicate that the random segment length has the most significant effect on TOP-ERL's performance . When using fixed 25 segments ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2410.09536", "content": "by G Li · 2024 · Cited by 7 — The results indicate that the random segment length has the most significant effect on TOP-ERL's performance . When using fixed 25 segments ..."}
+{"idx": 3, "title": "Transformer-based Off-Policy Episodic Reinforcement Learning", "date": "", "ddg_snippet": "12 Oct 2024 — The random segment length had the most significant effect on TOP - ERL's performance, as using fixed 25 segments caused the success rate to ...", "subpage_snippet": "", "source": "liner.com", "link": "https://liner.com/review/toperl-transformerbased-offpolicy-episodic-reinforcement-learning", "content": "12 Oct 2024 — The random segment length had the most significant effect on TOP - ERL's performance, as using fixed 25 segments caused the success rate to ..."}
+{"idx": 4, "title": "Transformer-based Off-Policy Episodic Reinforcement Learning", "date": "", "ddg_snippet": "This page provides the most accurate and concise summary worldwide for the paper titled TOP - ERL : Transformer - based Off - Policy Episodic Reinforcement Learning .", "subpage_snippet": "", "source": "www.themoonlight.io", "link": "https://www.themoonlight.io/en/review/top-erl-transformer-based-off-policy-episodic-reinforcement-learning", "content": "This page provides the most accurate and concise summary worldwide for the paper titled TOP - ERL : Transformer - based Off - Policy Episodic Reinforcement Learning ."}
+{"idx": 5, "title": "A Transformer-based Soft Actor-Critic with N-Step Returns", "date": "", "ddg_snippet": "5 Mar 2025 — Transformer - based Off - Policy Episodic RL ( TOP - ERL ) (Li et al., 2024a) introduces architectural innovation through a transformer - based critic ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.03660v1", "content": "5 Mar 2025 — Transformer - based Off - Policy Episodic RL ( TOP - ERL ) (Li et al., 2024a) introduces architectural innovation through a transformer - based critic ..."}
+{"idx": 6, "title": "Reinforcement Learning with Action Chunking", "date": "", "ddg_snippet": "We present Q-chunking, a simple yet effective recipe for improving reinforcement learning (RL) algorithms for long-horizon, sparse-reward tasks.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/fe95604beeb8813dfe73e2988998dacf80355101.pdf", "content": "We present Q-chunking, a simple yet effective recipe for improving reinforcement learning (RL) algorithms for long-horizon, sparse-reward tasks."}
+{"idx": 7, "title": "Downloads", "date": "", "ddg_snippet": "3D Segmenter : 3D Transformer based Semantic ... A Generalized Projected Bellman Error for Off - policy Value Estimation in Reinforcement Learning", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/Downloads/2023", "content": "3D Segmenter : 3D Transformer based Semantic ... A Generalized Projected Bellman Error for Off - policy Value Estimation in Reinforcement Learning"}
+{"idx": 8, "title": "gender Archives - Political Studies Review: our blog", "date": "", "ddg_snippet": "... segment on Last Week Tonight about Net Neutrality in which he explains the issue in more depth and provides the background that citizens would need to ...", "subpage_snippet": "", "source": "psr.brunel.ac.uk", "link": "https://psr.brunel.ac.uk/tag/gender/", "content": "... segment on Last Week Tonight about Net Neutrality in which he explains the issue in more depth and provides the background that citizens would need to ..."}
+{"idx": 9, "title": "Week 6 (35-42 Days) – International Working Dog", "date": "", "ddg_snippet": "... dog research into practice: development of science- based ... Transformative Power of Freedom of Choice in Changing Behavior - Presented by Marty G.", "subpage_snippet": "", "source": "www.iwdba.org", "link": "https://www.iwdba.org/working-dog-resources/week-6-35-42-days/", "content": "... dog research into practice: development of science- based ... Transformative Power of Freedom of Choice in Changing Behavior - Presented by Marty G."}
diff --git a/data/sampled_jsons/Taming_Knowledge_Conflicts_in_Language_Models_Section_2.1_knowledge_conflict_types_year_2024.jsonl b/data/sampled_jsons/Taming_Knowledge_Conflicts_in_Language_Models_Section_2.1_knowledge_conflict_types_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ab3c79921cea66d5f2d0197b198cb35ac3ad788f
--- /dev/null
+++ b/data/sampled_jsons/Taming_Knowledge_Conflicts_in_Language_Models_Section_2.1_knowledge_conflict_types_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "Jun 9, 2025 · Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this conflict to the interplay between “memory heads” and “context heads”, attention heads assumed to promote either memory or context exclusively. In this study, we go beyond this fundamental assumption by uncovering a critical phenomenon we term ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.10996v2", "content": "Jun 9, 2025 · Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this conflict to the interplay between “memory heads” and “context heads”, attention heads assumed to promote either memory or context exclusively. In this study, we go beyond this fundamental assumption by uncovering a critical phenomenon we term ..."}
+{"idx": 1, "title": "Resolving Knowledge Conflicts in Large Language Models", "date": "", "ddg_snippet": "Sep 21, 2023 · Large language models (LLMs) often encounter knowledge conflicts , scenarios where discrepancy arises between the internal parametric knowledge of LLMs and non-parametric information provided in the prompt context. In this work we ask what are the desiderata for LLMs when a knowledge conflict arises and whether existing LLMs fulfill them.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=S7T0slMrTD", "content": "Sep 21, 2023 · Large language models (LLMs) often encounter knowledge conflicts , scenarios where discrepancy arises between the internal parametric knowledge of LLMs and non-parametric information provided in the prompt context. In this work we ask what are the desiderata for LLMs when a knowledge conflict arises and whether existing LLMs fulfill them."}
+{"idx": 2, "title": "Untangle the KNOT: Interweaving Conflicting Knowledge and ...", "date": "", "ddg_snippet": "In this section , we provide definitions for the prob-ing task designed to resolve knowledge conflicts , and formulate 3 levels of reasoning skills involved in dealing with conflicting knowledge .", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2024.lrec-main.1493.pdf", "content": "In this section , we provide definitions for the prob-ing task designed to resolve knowledge conflicts , and formulate 3 levels of reasoning skills involved in dealing with conflicting knowledge ."}
+{"idx": 3, "title": "Tug-of-War Between Knowledge: Exploring and Resolving ...", "date": "", "ddg_snippet": "To solve the challenge of knowledge conflicts , we propose a method called Conflict -Disentangle Contrastive Decoding (CD2) to better calibrate the model ’s confidence. Experimental results demonstrate that our CD2 can effectively resolve knowledge conflicts in RALMs.", "subpage_snippet": "", "source": "www.lrec-conf.org", "link": "http://www.lrec-conf.org/proceedings/lrec-coling-2024/pdf/2024.main-1.1466.pdf", "content": "To solve the challenge of knowledge conflicts , we propose a method called Conflict -Disentangle Contrastive Decoding (CD2) to better calibrate the model ’s confidence. Experimental results demonstrate that our CD2 can effectively resolve knowledge conflicts in RALMs."}
+{"idx": 4, "title": "Resolving Knowledge Conflicts in Large Language Models", "date": "", "ddg_snippet": "Abstract Large language models (LLMs) often encounter knowledge conflicts , scenar-ios where discrepancy arises between the internal parametric knowledge of LLMs and non-parametric information provided in the prompt context. In this work we ask what are the desiderata for LLMs when a knowledge conflict arises and whether existing LLMs fulfill them.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2310.00935", "content": "Abstract Large language models (LLMs) often encounter knowledge conflicts , scenar-ios where discrepancy arises between the internal parametric knowledge of LLMs and non-parametric information provided in the prompt context. In this work we ask what are the desiderata for LLMs when a knowledge conflict arises and whether existing LLMs fulfill them."}
+{"idx": 5, "title": "大语言模型知识冲突的综述_knowledge conflicts for llms: a survey-C...", "date": "", "ddg_snippet": "Aug 7, 2024 · 2.4 内部记忆冲突（Intra-Memory Conflict ） 随着LLMs的发展，LLMs被广泛应用于知识密集型问答系统（Gao等人，2023b；Yu等人，2022；Petroni等人，2019；Chen等人，2023c）。 有效部署LLMs的一个关键方面是确保它们对具有相似含义或意图的各种表达产生一致的输出。", "subpage_snippet": "", "source": "blog.csdn.net", "link": "https://blog.csdn.net/m0_59164304/article/details/141001716", "content": "Aug 7, 2024 · 2.4 内部记忆冲突（Intra-Memory Conflict ） 随着LLMs的发展，LLMs被广泛应用于知识密集型问答系统（Gao等人，2023b；Yu等人，2022；Petroni等人，2019；Chen等人，2023c）。 有效部署LLMs的一个关键方面是确保它们对具有相似含义或意图的各种表达产生一致的输出。"}
+{"idx": 6, "title": "ConflictBank: A Benchmark for Evaluating the Influence of ...", "date": "", "ddg_snippet": "Aug 21, 2024 · Large language models (LLMs) have achieved impressive advancements across numerous disciplines, yet the critical issue of knowledge conflicts , a major source of hallucinations, has rarely been ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/383308137_ConflictBank_A_Benchmark_for_Evaluating_the_Influence_of_Knowledge_Conflicts_in_LLM", "content": "Aug 21, 2024 · Large language models (LLMs) have achieved impressive advancements across numerous disciplines, yet the critical issue of knowledge conflicts , a major source of hallucinations, has rarely been ..."}
+{"idx": 7, "title": "Taming Knowledge Conflicts in Language Models | OpenReview", "date": "", "ddg_snippet": "Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this conflict to the interplay between \"memory heads\" and \"context heads\", attention heads assumed to promote either memory or context...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0cEZyhHEks&referrer=[the+profile+of+Hanghang+Tong](/profile?id=~Hanghang_Tong2)", "content": "Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this conflict to the interplay between \"memory heads\" and \"context heads\", attention heads assumed to promote either memory or context..."}
+{"idx": 8, "title": "Addressing Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "#Understanding Knowledge Conflicts . Language models use two types of knowledge : Internal Memory and External Context. Internal memory is the knowledge that the model has learned during training. This is often factual but can become outdated.", "subpage_snippet": "", "source": "scisimple.com", "link": "https://scisimple.com/en/articles/2025-09-03-addressing-knowledge-conflicts-in-language-models--a9pwq1z", "content": "#Understanding Knowledge Conflicts . Language models use two types of knowledge : Internal Memory and External Context. Internal memory is the knowledge that the model has learned during training. This is often factual but can become outdated."}
+{"idx": 9, "title": "Taming Knowledge Conflicts in", "date": "", "ddg_snippet": "Taming Knowledge Conflicts in Language Models . Gaotang Li†, Yuzhong Chen‡, Hanghang Tong†. †University of Illinois Urbana-Champaign ‡ VISA Research Contact: gaotang3@illinois.edu. Paper. Knowledge Conflict : Parametric Memory vs. Contextual.", "subpage_snippet": "", "source": "actionable-interpretability.github.io", "link": "https://actionable-interpretability.github.io/posters/ICML_25_JuICE_Poster+-+Gaotang+Li.pdf", "content": "Taming Knowledge Conflicts in Language Models . Gaotang Li†, Yuzhong Chen‡, Hanghang Tong†. †University of Illinois Urbana-Champaign ‡ VISA Research Contact: gaotang3@illinois.edu. Paper. Knowledge Conflict : Parametric Memory vs. Contextual."}
diff --git a/data/sampled_jsons/TarFlow_score_function_gradient_log_density_denoising_normalizing_flows_year_2024.jsonl b/data/sampled_jsons/TarFlow_score_function_gradient_log_density_denoising_normalizing_flows_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e5932cd57250594a37519cb1b433558ef1cbf383
--- /dev/null
+++ b/data/sampled_jsons/TarFlow_score_function_gradient_log_density_denoising_normalizing_flows_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2412.06329] Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2412.06329", "content": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years."}
+{"idx": 1, "title": "Normalizing Flows are Capable Generative Models - Apple Machine...", "date": "", "ddg_snippet": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years.", "subpage_snippet": "", "source": "machinelearning.apple.com", "link": "https://machinelearning.apple.com/research/normalizing-flows", "content": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years."}
+{"idx": 2, "title": "Transformer Autoregressive Flow ( TARFlow )", "date": "", "ddg_snippet": "TARFlow combines autoregressive Transformers with normalizing flows to enable efficient, expressive density modeling for high-dimensional image generation.", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/topics/transformer-autoregressive-flow-tarflow", "content": "TARFlow combines autoregressive Transformers with normalizing flows to enable efficient, expressive density modeling for high-dimensional image generation."}
+{"idx": 3, "title": "Unlocking the Power of Normalizing Flows in Generative... - Glcnd.io", "date": "", "ddg_snippet": "Unveiling TarFlow : Advancing Normalizing Flows in Generative Modeling.The Importance of Normalizing Flows . NFs offer an elegant solution to the problem of learnable density estimation by transforming a simple distribution, such as a Gaussian, into a more complex one.", "subpage_snippet": "", "source": "glcnd.io", "link": "https://glcnd.io/unlocking-the-power-of-normalizing-flows-in-generative-modeling/", "content": "Unveiling TarFlow : Advancing Normalizing Flows in Generative Modeling.The Importance of Normalizing Flows . NFs offer an elegant solution to the problem of learnable density estimation by transforming a simple distribution, such as a Gaussian, into a more complex one."}
+{"idx": 4, "title": "\" Normalizing Flows are Capable Generative Models\"", "date": "", "ddg_snippet": "→ TARFlow reimagines Normalizing Flows using a stack of autoregressive Transformer blocks that process image patches.", "subpage_snippet": "", "source": "www.rohan-paul.com", "link": "https://www.rohan-paul.com/p/normalizing-flows-are-capable-generative", "content": "→ TARFlow reimagines Normalizing Flows using a stack of autoregressive Transformer blocks that process image patches."}
+{"idx": 5, "title": "(PDF) TrajFlow: A Generative Framework for Occupancy Density ...", "date": "", "ddg_snippet": "Density Estimation Using Normalizing Flows .into the inverse flow Ψ−1. ψto produce synthetic samples from the data distribution. gradients can also be computed by solving another ordinary. differential equation initial value problem [32].", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/388401805_TrajFlow_A_Generative_Framework_for_Occupancy_Density_Estimation_Using_Normalizing_Flows", "content": "Density Estimation Using Normalizing Flows .into the inverse flow Ψ−1. ψto produce synthetic samples from the data distribution. gradients can also be computed by solving another ordinary. differential equation initial value problem [32]."}
+{"idx": 6, "title": "Normalizing Flows as Generative Models - Kifinity", "date": "", "ddg_snippet": "TarFlow is a scalable Transformer-based architecture for Normalizing Flows that models image pixels directly. It incorporates techniques like noise augmentation, denoising , and guidance to improve sample quality.", "subpage_snippet": "", "source": "www.kifinity.com", "link": "https://www.kifinity.com/post/normalizing-flows-machinelearning-apple-e63648a2", "content": "TarFlow is a scalable Transformer-based architecture for Normalizing Flows that models image pixels directly. It incorporates techniques like noise augmentation, denoising , and guidance to improve sample quality."}
+{"idx": 7, "title": "Paper page - Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers/2412.06329", "content": "Normalizing Flows (NFs) are likelihood-based models for continuous inputs. They have demonstrated promising results on both density estimation and generative modeling tasks, but have received relatively little attention in recent years."}
+{"idx": 8, "title": "GitHub - TheProParadox/ tarflow _jax: Implemented Tarflow in JAX", "date": "", "ddg_snippet": "Score -based denoising step.Description of transformer_ flow .py. Implemented a Transformer-based normalizing flow model using Equinox and JAX, designed for image processing tasks with patch-based autoregression.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/TheProParadox/tarflow_jax", "content": "Score -based denoising step.Description of transformer_ flow .py. Implemented a Transformer-based normalizing flow model using Equinox and JAX, designed for image processing tasks with patch-based autoregression."}
+{"idx": 9, "title": "A Fokker-Planck-Based Loss Function that Bridges Dynamics with...", "date": "", "ddg_snippet": "It simultaneously estimates normalized density , energy, and score functions from both empirical data and dynamics.This formulation allows the neural network to learn both the score function ( gradient of log density ) and the dynamics simultaneously.", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/fokker-planck-based-loss-function-that-bridges", "content": "It simultaneously estimates normalized density , energy, and score functions from both empirical data and dynamics.This formulation allows the neural network to learn both the score function ( gradient of log density ) and the dynamics simultaneously."}
diff --git a/data/sampled_jsons/Temporally-Correlated_Episodic_RL_(TCE)_abstract_year_2024.jsonl b/data/sampled_jsons/Temporally-Correlated_Episodic_RL_(TCE)_abstract_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f0cf1b9ce536cf3cc2dde20d39a651d9df20e0d6
--- /dev/null
+++ b/data/sampled_jsons/Temporally-Correlated_Episodic_RL_(TCE)_abstract_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "step-based policy updates for temporally-correlated episodic ...", "date": "", "ddg_snippet": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ... 33 pages", "subpage_snippet": "", "source": "proceedings.iclr.cc", "link": "https://proceedings.iclr.cc/paper_files/paper/2024/file/52da50b1ef221e4b1793e3bf44dd973d-Paper-Conference.pdf", "content": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ... 33 pages"}
+{"idx": 1, "title": "transformer-based off-policy episodic reinforcement learning", "date": "", "ddg_snippet": "by G Li · 2024 · Cited by 7 — TCE Temporally-Correlated Episodic RL (TCE) (Li et al., 2024) is an innovative ERL algorithm that leverages step-level information in ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2410.09536", "content": "by G Li · 2024 · Cited by 7 — TCE Temporally-Correlated Episodic RL (TCE) (Li et al., 2024) is an innovative ERL algorithm that leverages step-level information in ..."}
+{"idx": 2, "title": "TOP-ERL: Transformer-based Off-Policy Episodic Reinforcement ...", "date": "", "ddg_snippet": "g) TCE: Temporally-Correlated Episodic RL (TCE ) [24] is an innovative ERL algorithm that leverages step-level in- formation in episodic policy updates ...", "subpage_snippet": "", "source": "rss25-roboreps.github.io", "link": "https://rss25-roboreps.github.io/papers/1_TOP_ERL_Transformer_based_Of.pdf", "content": "g) TCE: Temporally-Correlated Episodic RL (TCE ) [24] is an innovative ERL algorithm that leverages step-level in- formation in episodic policy updates ..."}
+{"idx": 3, "title": "Open the Black Box: Step-based Policy Updates for ...", "date": "", "ddg_snippet": "by G Li · 2024 · Cited by 11 — We introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy updates.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2401.11437", "content": "by G Li · 2024 · Cited by 11 — We introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy updates."}
+{"idx": 4, "title": "Open the Black Box: Step-based Policy Updates for...", "date": "", "ddg_snippet": "by G Li · Cited by 11 — The paper proposes a framework ( Temporally - Correlated Episodic RL ( TCE )) for reinforcement learning ( RL )-based robot trajectory estimation.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=mnipav175N", "content": "by G Li · Cited by 11 — The paper proposes a framework ( Temporally - Correlated Episodic RL ( TCE )) for reinforcement learning ( RL )-based robot trajectory estimation."}
+{"idx": 5, "title": "Daily Papers", "date": "", "ddg_snippet": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers?q=Temporally-Correlated+Episodic+RL", "content": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ..."}
+{"idx": 6, "title": "Latent Space Exploration and Trajectory Space Update in ...", "date": "", "ddg_snippet": "by G Li — Our proposed method, Temporally - Correlated Episodic . RL ( TCE ), moves beyond the traditional approach of treating an entire trajectory as a single data point.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=e8dcuniLcA", "content": "by G Li — Our proposed method, Temporally - Correlated Episodic . RL ( TCE ), moves beyond the traditional approach of treating an entire trajectory as a single data point."}
+{"idx": 7, "title": "Daily Papers", "date": "", "ddg_snippet": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/papers?q=Temporal-aware+Hierarchical+Cognitive+Reinforcement+Learning", "content": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ..."}
+{"idx": 8, "title": "Track: Poster Session 1", "date": "", "ddg_snippet": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ...", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2024/session/19806", "content": "In this work, we introduce a novel ERL algorithm, Temporally - Correlated Episodic RL ( TCE ), which effectively utilizes step information in episodic policy ..."}
+{"idx": 9, "title": "Category learning in a recurrent neural network with ...", "date": "", "ddg_snippet": "by Y Zhang · 2022 · Cited by 2 — Our results suggest that the neurons in the recurrent neural network have the ability to form categorical representations through deep reinforcement learning.", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC9640766/", "content": "by Y Zhang · 2022 · Cited by 2 — Our results suggest that the neurons in the recurrent neural network have the ability to form categorical representations through deep reinforcement learning."}
diff --git a/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_Sampling_Demons_arxiv2410.05760_appendix_reward_estimation_r.jsonl b/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_Sampling_Demons_arxiv2410.05760_appendix_reward_estimation_r.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..40512c38f4d5805b75992229629338338c12a47b
--- /dev/null
+++ b/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_Sampling_Demons_arxiv2410.05760_appendix_reward_estimation_r.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2410.05760] Training-free Diffusion Model Alignment with ...", "date": "", "ddg_snippet": "Oct 8, 2024 · Aligning diffusion models with user preferences has been a key challenge. Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.05760", "content": "Oct 8, 2024 · Aligning diffusion models with user preferences has been a key challenge. Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ..."}
+{"idx": 1, "title": "GitHub - catchfree1225/demonsampling: [ICLR'25] Official ...", "date": "", "ddg_snippet": "Official implementation of ICLR 2025 \" Sampling Demon \" ( arXiv:2410.05760 ). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models. By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/catchfree1225/demonsampling", "content": "Official implementation of ICLR 2025 \" Sampling Demon \" ( arXiv:2410.05760 ). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models. By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ..."}
+{"idx": 2, "title": "Training-free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Text-to-image diffusion model alignment is critical for improving the alignment between the generated images and human preferences. While training -based methods are constrained by high computational costs and dataset requirements, training-free alignment methods remain underexplored and are often limited by inaccurate guidance.", "subpage_snippet": "", "source": "axi.lims.ac.uk", "link": "https://axi.lims.ac.uk/paper/2410.05760", "content": "Text-to-image diffusion model alignment is critical for improving the alignment between the generated images and human preferences. While training -based methods are constrained by high computational costs and dataset requirements, training-free alignment methods remain underexplored and are often limited by inaccurate guidance."}
+{"idx": 3, "title": "Training-Free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ...", "subpage_snippet": "", "source": "proceedings.iclr.cc", "link": "https://proceedings.iclr.cc/paper_files/paper/2025/hash/eeab2e00835c71d64458ad1821e05664-Abstract-Conference.html", "content": "Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ..."}
+{"idx": 4, "title": "Training-free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "The paper introduces \" Sampling Demons ,\" a training-free approach that addresses these limitations by enabling diffusion model alignment with arbitrary, non-differentiable reward functions during inference.", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/overview/2410.05760v2", "content": "The paper introduces \" Sampling Demons ,\" a training-free approach that addresses these limitations by enabling diffusion model alignment with arbitrary, non-differentiable reward functions during inference."}
+{"idx": 5, "title": "Training-free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model retraining.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.05760v1", "content": "Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model retraining."}
+{"idx": 6, "title": "training-free diffusion model alignment with sampling ...", "date": "", "ddg_snippet": "by PH Yeh · 2024 · Cited by 5 — Aligning diffusion models with user preferences has been a key challenge . Existing methods for aligning diffusion models either require ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2410.05760?", "content": "by PH Yeh · 2024 · Cited by 5 — Aligning diffusion models with user preferences has been a key challenge . Existing methods for aligning diffusion models either require ..."}
+{"idx": 7, "title": "Flow-GRPO: Training Flow Matching Models via Online RL", "date": "", "ddg_snippet": "We find that online RL for flow matching models can use fewer denoising steps for training sample generation, significantly accelerating training .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.05470v1", "content": "We find that online RL for flow matching models can use fewer denoising steps for training sample generation, significantly accelerating training ."}
+{"idx": 8, "title": "Diffusion-Sharpening: Fine-tuning Diffusion Models with", "date": "", "ddg_snippet": "During training , we sample multiple trajectories and compute rewards through path integration, guiding the model to optimize towards the best ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.12146v1", "content": "During training , we sample multiple trajectories and compute rewards through path integration, guiding the model to optimize towards the best ..."}
+{"idx": 9, "title": "Published as a conference paper at ICLR 2025 - OpenReview", "date": "", "ddg_snippet": "This paper introduces a training-free method to improve diffusion models for text-to-image generation, enhancing aesthetic scores without additional training .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=tfemquulED", "content": "This paper introduces a training-free method to improve diffusion models for text-to-image generation, enhancing aesthetic scores without additional training ."}
diff --git a/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_with_Sampling_Demons_appendix_supplementary_Section_J.jsonl b/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_with_Sampling_Demons_appendix_supplementary_Section_J.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4618c9345e78af850987f5dfce38ac42ffead2ae
--- /dev/null
+++ b/data/sampled_jsons/Training-Free_Diffusion_Model_Alignment_with_Sampling_Demons_appendix_supplementary_Section_J.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "GitHub - aiiu-lab/DemonSampling: [ICLR'25] Official ...", "date": "", "ddg_snippet": "Official implementation of ICLR 2025 \" Sampling Demon \" (arXiv:2410.05760). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models . By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/aiiu-lab/DemonSampling", "content": "Official implementation of ICLR 2025 \" Sampling Demon \" (arXiv:2410.05760). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models . By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ..."}
+{"idx": 1, "title": "Training - free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "This paper introduces a training - free technique called \" Sampling Demons \" for aligning diffusion models with target objectives. The method works by modifying the diffusion sampling process to match the desired output during inference, without requiring any additional training.", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/training-free-diffusion-model-alignment-sampling-demons", "content": "This paper introduces a training - free technique called \" Sampling Demons \" for aligning diffusion models with target objectives. The method works by modifying the diffusion sampling process to match the desired output during inference, without requiring any additional training."}
+{"idx": 2, "title": "Training-free Diffusion Model Alignment with Sampling Demons Images Training-Free Diffusion Model Alignment with Sampling Demons GitHub - aiiu-lab/DemonSampling: [ICLR'25] Official ... Published as a conference paper at ICLR 2025 - OpenReview Training-free Diffusion Model Alignment with Sampling Demons ... DAS (Diffusion Alignment as Sampling), ICLR'25 Spotlight Training - free Diffusion Model Alignment with Sampling Demons Training - free Diffusion Model Alignment with Sampling Demons GitHub - aiiu-lab/DemonSampling: [ICLR'25] Official implementation of … DAS ( Diffusion Alignment as Sampling), ICLR'25 Spotlight DAS ( Diffusion Alignment as Sampling), ICLR'25 Spotlight Alignment without Over-optimization: Training-Free Solution ...", "date": "", "ddg_snippet": "Oct 8, 2024 · To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models . Our method can be easily integrated with existing diffusion models without further training . View all Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ... Official implementation of ICLR 2025 \" Sampling Demon \" (arXiv:2410.05760). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models . By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ... To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models . Our method can be easily integrated with existing diffusion models without further training . This paper introduces a training-free technique called \" Sampling Demons \" for aligning diffusion models with target objectives. The method works by modifying the diffusion sampling process to match the desired output during inference, without requiring any additional training . Our approach, tailored for diffusion sampling and incorporating tempering techniques, achieves comparable or superior target rewards to fine-tuning methods while preserving diversity and cross-reward generalization. We demonstrate its effectiveness in single-reward optimization, multi-objective scenarios, and online black-box optimization. What is the first inference-time preference alignment method for diffusion models? To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models. Our method can be easily integrated with existing diffusion models without further training . Can diffusion models be aligned with user preferences? Aligning diffusion models with user preferences has been a key challenge . Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. How does sampling demon work? By aligning the denoising process with user preferences via stochastic optimization , Sampling Demon enables the use of non-differentiable reward signals—such as those from Visual-Language Model (VLM) APIs and human judgements—without requiring retraining or fine-tuning of the underlying diffusion model. Is there a test-time alignment of diffusion models without reward over-optimization? This is the official implementation of our paper Test-time Alignment of Diffusion Models without Reward Over-optimization by Sunwoo Kim 1, Minkyu Kim 2, Dongmin Park 2. 1 Seoul National University, 2 KRAFTON AI Who wrote 'test-time alignment of diffusion models without reward over-optimization'? title={Test-time Alignment of Diffusion Models without Reward Over-optimization}, author={ Sunwoo Kim and Minkyu Kim and Dongmin Park }, booktitle={The Thirteenth International Conference on Learning Representations}, year={2025}, Abstract Diffusion models excel in generative tasks, but aligning them with specific objectives while maintaining their versatility remains challenging. Existing fine-tuning methods often suffer from reward over-optimization, while approximate guidance approaches fail to optimize target rewards effectively. Addressing these limitations, we propose a training-free sampling method based on ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.05760", "content": "Oct 8, 2024 · To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models . Our method can be easily integrated with existing diffusion models without further training . View all Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ... Official implementation of ICLR 2025 \" Sampling Demon \" (arXiv:2410.05760). This repository contains the official implementation of Sampling Demon , an inference-time, backpropagation- free preference alignment method for diffusion models . By aligning the denoising process with user preferences via stochastic optimization, Sampling Demon enables the use of non-differentiable reward signals—such ... To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models . Our method can be easily integrated with existing diffusion models without further training . This paper introduces a training-free technique called \" Sampling Demons \" for aligning diffusion models with target objectives. The method works by modifying the diffusion sampling process to match the desired output during inference, without requiring any additional training . Our approach, tailored for diffusion sampling and incorporating tempering techniques, achieves comparable or superior target rewards to fine-tuning methods while preserving diversity and cross-reward generalization. We demonstrate its effectiveness in single-reward optimization, multi-objective scenarios, and online black-box optimization. What is the first inference-time preference alignment method for diffusion models? To the best of our knowledge, the proposed approach is the first inference-time, backpropagation- free preference alignment method for diffusion models. Our method can be easily integrated with existing diffusion models without further training . Can diffusion models be aligned with user preferences? Aligning diffusion models with user preferences has been a key challenge . Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. How does sampling demon work? By aligning the denoising process with user preferences via stochastic optimization , Sampling Demon enables the use of non-differentiable reward signals—such as those from Visual-Language Model (VLM) APIs and human judgements—without requiring retraining or fine-tuning of the underlying diffusion model. Is there a test-time alignment of diffusion models without reward over-optimization? This is the official implementation of our paper Test-time Alignment of Diffusion Models without Reward Over-optimization by Sunwoo Kim 1, Minkyu Kim 2, Dongmin Park 2. 1 Seoul National University, 2 KRAFTON AI Who wrote 'test-time alignment of diffusion models without reward over-optimization'? title={Test-time Alignment of Diffusion Models without Reward Over-optimization}, author={ Sunwoo Kim and Minkyu Kim and Dongmin Park }, booktitle={The Thirteenth International Conference on Learning Representations}, year={2025}, Abstract Diffusion models excel in generative tasks, but aligning them with specific objectives while maintaining their versatility remains challenging. Existing fine-tuning methods often suffer from reward over-optimization, while approximate guidance approaches fail to optimize target rewards effectively. Addressing these limitations, we propose a training-free sampling method based on ..."}
+{"idx": 3, "title": "Training-Free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ...", "subpage_snippet": "", "source": "proceedings.iclr.cc", "link": "https://proceedings.iclr.cc/paper_files/paper/2025/hash/eeab2e00835c71d64458ad1821e05664-Abstract-Conference.html", "content": "Aligning diffusion models with user preferences has been a key challenge.Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions.To address these limitations, we propose a stochastic optimization approach, dubbed Demon , to guide the denoising process at inference time without backpropagation through reward functions or model ..."}
+{"idx": 4, "title": "DAS (Diffusion Alignment as Sampling), ICLR'25 Spotlight", "date": "", "ddg_snippet": "Our approach, tailored for diffusion sampling and incorporating tempering techniques, achieves comparable or superior target rewards to fine-tuning methods while preserving diversity and cross-reward generalization. We demonstrate its effectiveness in single-reward optimization, multi-objective scenarios, and online black-box optimization.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/krafton-ai/DAS", "content": "Our approach, tailored for diffusion sampling and incorporating tempering techniques, achieves comparable or superior target rewards to fine-tuning methods while preserving diversity and cross-reward generalization. We demonstrate its effectiveness in single-reward optimization, multi-objective scenarios, and online black-box optimization."}
+{"idx": 5, "title": "Alignment without Over-optimization: Training-Free Solution ...", "date": "", "ddg_snippet": "Abstract Diffusion models excel in generative tasks, but aligning them with specific objectives while maintaining their versatility remains challenging. Existing fine-tuning methods often suffer from reward over-optimization, while approximate guidance approaches fail to optimize target rewards effectively. Addressing these limitations, we propose a training-free sampling method based on ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2501.05803v1", "content": "Abstract Diffusion models excel in generative tasks, but aligning them with specific objectives while maintaining their versatility remains challenging. Existing fine-tuning methods often suffer from reward over-optimization, while approximate guidance approaches fail to optimize target rewards effectively. Addressing these limitations, we propose a training-free sampling method based on ..."}
+{"idx": 6, "title": "Training - free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "One approach to aligning diffusion models with user preferences is to fine-tune using reinforcement learning (RL) to optimize the models based on rewards signals that reflect the user preferences (Black et al., 2023; Fan et al., 2023) .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.05760v1", "content": "One approach to aligning diffusion models with user preferences is to fine-tune using reinforcement learning (RL) to optimize the models based on rewards signals that reflect the user preferences (Black et al., 2023; Fan et al., 2023) ."}
+{"idx": 7, "title": "Training - free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "alphaXiv. Go Home. Training - free Diffusion Model Alignment with Sampling Demons .Select any part of the paper to ask specific questions about that section . Add Context. Type @ to reference other papers and expand the discussion.", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/abs/2410.05760v2", "content": "alphaXiv. Go Home. Training - free Diffusion Model Alignment with Sampling Demons .Select any part of the paper to ask specific questions about that section . Add Context. Type @ to reference other papers and expand the discussion."}
+{"idx": 8, "title": "aiiu-lab/DemonSampling | DeepWiki", "date": "", "ddg_snippet": "This document introduces the DemonSampling system, a training - free diffusion model alignment technique that enables preference alignment during inference without requiring backpropagation. For installation instructions, see Installation.", "subpage_snippet": "", "source": "deepwiki.com", "link": "https://deepwiki.com/aiiu-lab/DemonSampling", "content": "This document introduces the DemonSampling system, a training - free diffusion model alignment technique that enables preference alignment during inference without requiring backpropagation. For installation instructions, see Installation."}
+{"idx": 9, "title": "Training - free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon ...", "subpage_snippet": "", "source": "synthical.com", "link": "https://synthical.com/article/Training-free-Diffusion-Model-Alignment-with-Sampling-Demons-7d88de32-4c7d-456a-8886-7029b4030b48", "content": "Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a stochastic optimization approach, dubbed Demon ..."}
diff --git a/data/sampled_jsons/UPGNET_HOA_NFR_H-N_N-H_high-order_aggregator_node_feature_regularization_order.jsonl b/data/sampled_jsons/UPGNET_HOA_NFR_H-N_N-H_high-order_aggregator_node_feature_regularization_order.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6cdce94f5d9d7428adc1744f54faf48824634f1d
--- /dev/null
+++ b/data/sampled_jsons/UPGNET_HOA_NFR_H-N_N-H_high-order_aggregator_node_feature_regularization_order.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "On the use of high-order feature propagation in Graph Convolution ...", "date": "", "ddg_snippet": "First, unlike the classical GCN that performs one-hop feature propagation, we use high-order feature propagation so that distant neighbors can improve the data's resulting representation before applying the linear and nonlinear transformations associated with a given GCN block.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0020025521010574", "content": "First, unlike the classical GCN that performs one-hop feature propagation, we use high-order feature propagation so that distant neighbors can improve the data's resulting representation before applying the linear and nonlinear transformations associated with a given GCN block."}
+{"idx": 1, "title": "PDF High-Order Regularization on Graphs - microsoft.com", "date": "", "ddg_snippet": "We address the spam detection issue using the high-order regularization on graph. In most web search en-gines, the more hyperlinks that point to a web page, the more important the web page.", "subpage_snippet": "", "source": "www.microsoft.com", "link": "https://www.microsoft.com/en-us/research/wp-content/uploads/2017/01/high.pdf", "content": "We address the spam detection issue using the high-order regularization on graph. In most web search en-gines, the more hyperlinks that point to a web page, the more important the web page."}
+{"idx": 2, "title": "Higher order heterogeneous graph neural network based on node attribute ...", "date": "", "ddg_snippet": "Specifically, they cannot entirely capture the attributes of higher order neighbors or only use the higher order homogeneous neighbors, thus disregarding the attributes of heterogeneous neighbors. To address these problems, we propose a higher order heterogeneous graph neural network based on heterogeneous node attribute enhancement (HOAE).", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0957417423029068", "content": "Specifically, they cannot entirely capture the attributes of higher order neighbors or only use the higher order homogeneous neighbors, thus disregarding the attributes of heterogeneous neighbors. To address these problems, we propose a higher order heterogeneous graph neural network based on heterogeneous node attribute enhancement (HOAE)."}
+{"idx": 3, "title": "Understanding how regularization penalizes higher order features", "date": "", "ddg_snippet": "I am trying to understand how exactly L2 or ridge regression penalize higher order polynomial features to make fitting curve smoother. I read this article about regularization . I contemplated and I...", "subpage_snippet": "", "source": "stats.stackexchange.com", "link": "https://stats.stackexchange.com/questions/494307/understanding-how-regularization-penalizes-higher-order-features", "content": "I am trying to understand how exactly L2 or ridge regression penalize higher order polynomial features to make fitting curve smoother. I read this article about regularization . I contemplated and I..."}
+{"idx": 4, "title": "High-order Regularization", "date": "", "ddg_snippet": "We reduce computational complexity by building a local first- order approximation of the manifold as a surrogate geometry, and construct our high-order regularizer based on local derivative evaluations therein.", "subpage_snippet": "", "source": "vcai.mpi-inf.mpg.de", "link": "https://vcai.mpi-inf.mpg.de/projects/hreg/index.html", "content": "We reduce computational complexity by building a local first- order approximation of the manifold as a surrogate geometry, and construct our high-order regularizer based on local derivative evaluations therein."}
+{"idx": 5, "title": "GitHub - polarch/Higher-Order-Ambisonics: A set of routines ...", "date": "", "ddg_snippet": "A set of routines implementing higher- order ambisonics for spatial sound encoding, manipulation and decoding. - polarch/Higher- Order -Ambisonics", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/polarch/Higher-Order-Ambisonics", "content": "A set of routines implementing higher- order ambisonics for spatial sound encoding, manipulation and decoding. - polarch/Higher- Order -Ambisonics"}
+{"idx": 6, "title": "Model Search - upgnet.com", "date": "", "ddg_snippet": "Light-Commercial Single-Packages 25-40 Ton Commercial 15-25 Ton Commercial 6.5-12.5 Ton Commercial 3-6 Ton Commercial Enter a Full or Partial Model Number (Ex. DM090E09N4BZZ40001 or BA-08E18DTMZZ30001)", "subpage_snippet": "", "source": "upgnet.com", "link": "https://upgnet.com/Site/ModelSearch/default.aspx?AppID=CommercialProductLookup", "content": "Light-Commercial Single-Packages 25-40 Ton Commercial 15-25 Ton Commercial 6.5-12.5 Ton Commercial 3-6 Ton Commercial Enter a Full or Partial Model Number (Ex. DM090E09N4BZZ40001 or BA-08E18DTMZZ30001)"}
+{"idx": 7, "title": "Going Deeper into Locally Differentially Private Graph Neural Networks", "date": "", "ddg_snippet": "To mitigate these, UPGNET incorporates a Node Feature Regularization ( NFR ) layer using L1- regularization to reduce effective feature dimensions and a High-Order Aggregator ( HOA ) layer to expand effective neighborhood size, thereby minimizing estimation errors .", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=2aKHuXdr7Q", "content": "To mitigate these, UPGNET incorporates a Node Feature Regularization ( NFR ) layer using L1- regularization to reduce effective feature dimensions and a High-Order Aggregator ( HOA ) layer to expand effective neighborhood size, thereby minimizing estimation errors ."}
+{"idx": 8, "title": "Going Deeper into Locally Differentially Private Graph ...", "date": "", "ddg_snippet": "Based on the above analysis, UPGNET enhances utility by introducing two core layers: High-Order Aggregator (HOA) layer and the Node Feature Regularization (NFR) ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46579", "content": "Based on the above analysis, UPGNET enhances utility by introducing two core layers: High-Order Aggregator (HOA) layer and the Node Feature Regularization (NFR) ..."}
+{"idx": 9, "title": "Going Deeper into Locally Differentially Private Graph Neural ...", "date": "", "ddg_snippet": "server side, x′ undergoes processing through an Node Feature Regularization ( NFR ) layer and an High - Order Aggregator ( HOA ) layer to enhance utility before ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/7efbd36f792e0a32a0c290ac11c17eaa63d6680d.pdf", "content": "server side, x′ undergoes processing through an Node Feature Regularization ( NFR ) layer and an High - Order Aggregator ( HOA ) layer to enhance utility before ..."}
diff --git a/data/sampled_jsons/UVGS_Reimagining_Unstructured_3D_Gaussian_Splatting_using_UV_Mapping_spherical_coordinates_UV_mappin.jsonl b/data/sampled_jsons/UVGS_Reimagining_Unstructured_3D_Gaussian_Splatting_using_UV_Mapping_spherical_coordinates_UV_mappin.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..60a964724994d47fcf3f6da7926966263a3203da
--- /dev/null
+++ b/data/sampled_jsons/UVGS_Reimagining_Unstructured_3D_Gaussian_Splatting_using_UV_Mapping_spherical_coordinates_UV_mappin.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "BAGHIRLI, Samid : Taekwondo Data", "date": "", "ddg_snippet": "All about Samid BAGHIRLI, Takewondo full contact fighter at Taekwondo Data.", "subpage_snippet": "", "source": "www.taekwondodata.com", "link": "https://www.taekwondodata.com/samid-baghirli.ak1w.html?lang=en", "content": "All about Samid BAGHIRLI, Takewondo full contact fighter at Taekwondo Data."}
+{"idx": 1, "title": "TAEKWONDO INDIA", "date": "", "ddg_snippet": "All Indian Taekwondo FederationAt Taekwondo India, we believe in empowering individuals through the ancient martial art of Taekwondo . Founded with a mission to inspire and train people to achieve their full potential, we provide top-notch instruction, guidance, and opportunities for learners across all skill levels.", "subpage_snippet": "", "source": "indiataekwondo.in", "link": "https://indiataekwondo.in/", "content": "All Indian Taekwondo FederationAt Taekwondo India, we believe in empowering individuals through the ancient martial art of Taekwondo . Founded with a mission to inspire and train people to achieve their full potential, we provide top-notch instruction, guidance, and opportunities for learners across all skill levels."}
+{"idx": 2, "title": "Taekwondo | Live Taekwondo Scores, Highlights, News ... - NDTV...", "date": "", "ddg_snippet": "Aug 9, 2025 · Get all the latest information on Taekwondo ), live scores, news, results, stats, videos, highlights. Find all the Taekwondo tournament’s schedules at NDTV Sports", "subpage_snippet": "", "source": "sports.ndtv.com", "link": "https://sports.ndtv.com/taekwondo", "content": "Aug 9, 2025 · Get all the latest information on Taekwondo ), live scores, news, results, stats, videos, highlights. Find all the Taekwondo tournament’s schedules at NDTV Sports"}
+{"idx": 3, "title": "HOME | All Indian Taekwondo", "date": "", "ddg_snippet": "All Indian Taekwondo association is a governing body of Taekwondo in India that works in collaboration with Sterling taekwondo , is affiliated with Kukkiwon, and is registered by the Indian government. We encourage self-defense and foster the ideals of Taekwondo in India.", "subpage_snippet": "", "source": "www.nationaltaekwondo.in", "link": "https://www.nationaltaekwondo.in/", "content": "All Indian Taekwondo association is a governing body of Taekwondo in India that works in collaboration with Sterling taekwondo , is affiliated with Kukkiwon, and is registered by the Indian government. We encourage self-defense and foster the ideals of Taekwondo in India."}
+{"idx": 4, "title": "Taekwondo - All the updates, scores, and stories around the...", "date": "", "ddg_snippet": "All the updates, scores, videos, and results around Indian Taekwondo", "subpage_snippet": "", "source": "thebridge.in", "link": "https://thebridge.in/taekwondo/", "content": "All the updates, scores, videos, and results around Indian Taekwondo"}
+{"idx": 5, "title": "Taekwondo : Latest News and Updates, Top Stories, Videos ... -...", "date": "", "ddg_snippet": "Jun 30, 2021 · Taekwondo : Get Taekwondo latest news and headlines, top stories, live updates, special reports, articles, videos, photos and complete coverage at myKhel.com.", "subpage_snippet": "", "source": "www.mykhel.com", "link": "https://www.mykhel.com/topic/taekwondo", "content": "Jun 30, 2021 · Taekwondo : Get Taekwondo latest news and headlines, top stories, live updates, special reports, articles, videos, photos and complete coverage at myKhel.com."}
+{"idx": 6, "title": "TKD Scoring Wi-Fi Server (free) download Windows version", "date": "", "ddg_snippet": "Mar 8, 2025 · TKD Scoring Wi-Fi Server is an innovative electronic scoring program specifically designed for Taekwondo classes and tournaments. The latest version, 5.0.0, comes from the reputable program developer, TKD Scoring Wi-Fi, and is compatible with a range of operating systems including Windows XP, 7, 8, 10, and 11.", "subpage_snippet": "", "source": "en.freedownloadmanager.org", "link": "https://en.freedownloadmanager.org/Windows-PC/TKD-Scoring-Wi-Fi-Server-FREE.html", "content": "Mar 8, 2025 · TKD Scoring Wi-Fi Server is an innovative electronic scoring program specifically designed for Taekwondo classes and tournaments. The latest version, 5.0.0, comes from the reputable program developer, TKD Scoring Wi-Fi, and is compatible with a range of operating systems including Windows XP, 7, 8, 10, and 11."}
+{"idx": 7, "title": "Taekwon-do Association of India (TAI)", "date": "", "ddg_snippet": "TAEKWON-DO was first introduced to INDIA by GRAND MASTER HIL CHOW when he was directed by GEN. CHOI HONG HI the Founder of Taekwon-do & President of ITF (International Taekwon-do Federation) to visit India in the year 1968. But Grand Master Hi Chow could not get support he went back to Korea and later went to USA and developed Taekwon-do there. “ Taekwon-do Association of India” (in short TAI) was born due to the entrance of ITF Taekwon-do to India in the year 1984 under the guidance of ...", "subpage_snippet": "", "source": "www.itfindia.org.in", "link": "https://www.itfindia.org.in/", "content": "TAEKWON-DO was first introduced to INDIA by GRAND MASTER HIL CHOW when he was directed by GEN. CHOI HONG HI the Founder of Taekwon-do & President of ITF (International Taekwon-do Federation) to visit India in the year 1968. But Grand Master Hi Chow could not get support he went back to Korea and later went to USA and developed Taekwon-do there. “ Taekwon-do Association of India” (in short TAI) was born due to the entrance of ITF Taekwon-do to India in the year 1984 under the guidance of ..."}
+{"idx": 8, "title": "Samad Tkd Tkd - Facebook", "date": "", "ddg_snippet": "Samad Tkd Tkd is on Facebook. Join Facebook to connect with Samad Tkd Tkd and others you may know. Facebook gives people the power to share and makes the world more open and connected.", "subpage_snippet": "", "source": "www.facebook.com", "link": "https://www.facebook.com/samadtkd.tkd.3/", "content": "Samad Tkd Tkd is on Facebook. Join Facebook to connect with Samad Tkd Tkd and others you may know. Facebook gives people the power to share and makes the world more open and connected."}
+{"idx": 9, "title": "Taekwondo : Latest News, Photos, Videos on Taekwondo - NDTV.COM", "date": "", "ddg_snippet": "Find Taekwondo Latest News, Videos & Pictures on Taekwondo and see latest updates, news, information from NDTV.COM . Explore more on Taekwondo .", "subpage_snippet": "", "source": "www.ndtv.com", "link": "https://www.ndtv.com/topic/taekwondo", "content": "Find Taekwondo Latest News, Videos & Pictures on Taekwondo and see latest updates, news, information from NDTV.COM . Explore more on Taekwondo ."}
diff --git a/data/sampled_jsons/UVGS_paper_custom_3DGS_dataset_Objaverse_100K_objects_experimental_section_year_2024.jsonl b/data/sampled_jsons/UVGS_paper_custom_3DGS_dataset_Objaverse_100K_objects_experimental_section_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3f6e3502c6461731a48155b786e8212647d28cae
--- /dev/null
+++ b/data/sampled_jsons/UVGS_paper_custom_3DGS_dataset_Objaverse_100K_objects_experimental_section_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "UVGS: Reimagining Unstructured 3D Gaussian Splatting ...", "date": "", "ddg_snippet": "3 Feb 2025 — We performed our experiments on our custom Objaverse [10] 3DGS dataset and evaluate the performance of our model in terms of PSNR, SSIM, and ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.01846v1", "content": "3 Feb 2025 — We performed our experiments on our custom Objaverse [10] 3DGS dataset and evaluate the performance of our model in terms of PSNR, SSIM, and ..."}
+{"idx": 1, "title": "UVGS: Reimagining Unstructured 3D Gaussian Splatting ...", "date": "", "ddg_snippet": "by A Rai · 2025 · Cited by 2 — we create a high-quality and large-scale 3DGS dataset of. ∼400K objects and scenes from Objaverse . We only use static scenes or objects from Objaverse . 11 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Rai_UVGS_Reimagining_Unstructured_3D_Gaussian_Splatting_using_UV_Mapping_CVPR_2025_paper.pdf", "content": "by A Rai · 2025 · Cited by 2 — we create a high-quality and large-scale 3DGS dataset of. ∼400K objects and scenes from Objaverse . We only use static scenes or objects from Objaverse . 11 pages"}
+{"idx": 2, "title": "100 Дней Выживания в Stalker Anomaly Custom — Видео от Leoner", "date": "", "ddg_snippet": "6,5 тыс просмотров. 100 Дней Выживания в Subnautica. 1:03:16.100 Дней Выживания в Mount & Blade II: Bannerlord.", "subpage_snippet": "", "source": "vk.com", "link": "https://vk.com/video-197839979_456239248", "content": "6,5 тыс просмотров. 100 Дней Выживания в Subnautica. 1:03:16.100 Дней Выживания в Mount & Blade II: Bannerlord."}
+{"idx": 3, "title": "CVPR 2025 Friday 06/13", "date": "", "ddg_snippet": "This paper presents a novel framework for modeling and conditional generation of 3D articulated objects . Troubled by flexibility-quality tradeoffs, existing ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/day/6/13", "content": "This paper presents a novel framework for modeling and conditional generation of 3D articulated objects . Troubled by flexibility-quality tradeoffs, existing ..."}
+{"idx": 4, "title": "GitHub - lukasHoel/ 3 DGS -LM: 3 DGS -LM accelerates...", "date": "", "ddg_snippet": "3 DGS -LM accelerates Gaussian-Splatting optimization by replacing the ADAM optimizer with Levenberg-Marquardt. This is the official repository that contains source code for the ICCV 2025 paper 3 DGS -LM.Run bash scripts/fit_all_scenes.sh on an NVIDIA A100 80GB GPU.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/lukasHoel/3DGS-LM", "content": "3 DGS -LM accelerates Gaussian-Splatting optimization by replacing the ADAM optimizer with Levenberg-Marquardt. This is the official repository that contains source code for the ICCV 2025 paper 3 DGS -LM.Run bash scripts/fit_all_scenes.sh on an NVIDIA A100 80GB GPU."}
+{"idx": 5, "title": "stabilityai/TripoSR · Hugging Face", "date": "", "ddg_snippet": "We use renders from the Objaverse dataset , utilizing our enhanced rendering method that more closely replicate the distribution of images found in the real world, significantly improving our model’s ability to generalize.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/stabilityai/TripoSR", "content": "We use renders from the Objaverse dataset , utilizing our enhanced rendering method that more closely replicate the distribution of images found in the real world, significantly improving our model’s ability to generalize."}
+{"idx": 6, "title": "I Love You - 100x to 10,000x Times Copy and Paste", "date": "", "ddg_snippet": "Key Features✔Generate up to 10,000 \"I Love You\" messages with custom counts (100x, 1,000x, 10,000x).✔2 languages supported (English, Español, Français, Hindi, etc.).", "subpage_snippet": "", "source": "www.iloveyoucopyandpaste.com", "link": "https://www.iloveyoucopyandpaste.com/", "content": "Key Features✔Generate up to 10,000 \"I Love You\" messages with custom counts (100x, 1,000x, 10,000x).✔2 languages supported (English, Español, Français, Hindi, etc.)."}
+{"idx": 7, "title": "3D Mesh Editing using Masked LRMs", "date": "", "ddg_snippet": "Objaverse : A universe of annotated 3d objects . In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13142–13153, 2023. Deitke et al.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.08641v2", "content": "Objaverse : A universe of annotated 3d objects . In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13142–13153, 2023. Deitke et al."}
+{"idx": 8, "title": "Polaris Serenade - Official Fisch Wiki", "date": "", "ddg_snippet": "and three Shiny Totems. respectively. The third and final quest will require the player to obtain 50,000,000C$.This section contains subjective information that was last updated for Version 1.39.0.", "subpage_snippet": "", "source": "fischipedia.org", "link": "https://fischipedia.org/wiki/Polaris_Serenade", "content": "and three Shiny Totems. respectively. The third and final quest will require the player to obtain 50,000,000C$.This section contains subjective information that was last updated for Version 1.39.0."}
+{"idx": 9, "title": "Upscale Image Online 100% Free | i2IMG", "date": "", "ddg_snippet": "AI-driven upscaling algorithms, often based on deep learning techniques, are trained on vast datasets of high-resolution images. This training allows them to learn the complex relationships between pixels and to predict the missing details with remarkable accuracy.", "subpage_snippet": "", "source": "www.i2img.com", "link": "https://www.i2img.com/upscale-image", "content": "AI-driven upscaling algorithms, often based on deep learning techniques, are trained on vast datasets of high-resolution images. This training allows them to learn the complex relationships between pixels and to predict the missing details with remarkable accuracy."}
diff --git a/data/sampled_jsons/Universal_convexification_via_risk-aversion_Dvijotham_2014.jsonl b/data/sampled_jsons/Universal_convexification_via_risk-aversion_Dvijotham_2014.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8ef48dd0e5015a2cc6595b9d8320c673a034c580
--- /dev/null
+++ b/data/sampled_jsons/Universal_convexification_via_risk-aversion_Dvijotham_2014.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Universal convexification via risk-aversion | Proceedings of ...", "date": "", "ddg_snippet": "Jul 23, 2014 · We then extend the framework to a general class of discrete-time dynamical systems where our convexification approach falls under the paradigm of risk -sensitive Markov Decision Processes. We derive the first model-based and model-free policy gradient optimization algorithms with guaranteed convergence to the optimal solution.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3020751.3020769", "content": "Jul 23, 2014 · We then extend the framework to a general class of discrete-time dynamical systems where our convexification approach falls under the paradigm of risk -sensitive Markov Decision Processes. We derive the first model-based and model-free policy gradient optimization algorithms with guaranteed convergence to the optimal solution."}
+{"idx": 1, "title": "[1406.0554] Universal Convexification via Risk-Aversion Universal convexification via risk-aversion | Proceedings of ... Universal Convexi cation via Risk-Aversion - UAI (PDF) Universal Convexification via Risk-Aversion - ResearchGate Universal Convexification via Risk-Aversion - paperzz.com \"Universal Convexification via Risk-Aversion.\" - dblp Universal Convexification via Risk Aversion", "date": "", "ddg_snippet": "Jun 3, 2014 · View a PDF of the paper titled Universal Convexification via Risk-Aversion , by Krishnamurthy Dvijotham and 1 other authors Jul 23, 2014 · We then extend the framework to a general class of discrete-time dynamical systems where our convexification approach falls under the paradigm of risk -sensitive Markov Decision Processes. We derive the first model-based and model-free policy gradient optimization algorithms with guaranteed convergence to the optimal solution. Universal Convexi cation via Risk-Aversion ... Abstract We develop a framework for convexifying a general class of optimization problems. We analyze the suboptimality of the solution to the convexi ed problem relative to the original nonconvex problem, and prove ad-ditive approximation guarantees under some assumptions. Jun 2, 2014 · In this context, our convexification approach falls under the well-studied paradigm of risk -sensitive Markov Decision Processes. Universal Convexification via Risk-Aversion Krishnamurthy Dvijotham Dept of Computer Science & Engg University of Washington Seattle, WA 98195 Maryam Fazel Dept of Electrical Engg University of Washington Seattle, WA 98195 Abstract We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution ... Bibliographic details on Universal Convexification via Risk-Aversion . Dvijotham Krishnamurthy, CalTech Abstract: We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution to the convexified problem relative to the original nonconvex problem and prove additive approximation guarantees.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1406.0554", "content": "Jun 3, 2014 · View a PDF of the paper titled Universal Convexification via Risk-Aversion , by Krishnamurthy Dvijotham and 1 other authors Jul 23, 2014 · We then extend the framework to a general class of discrete-time dynamical systems where our convexification approach falls under the paradigm of risk -sensitive Markov Decision Processes. We derive the first model-based and model-free policy gradient optimization algorithms with guaranteed convergence to the optimal solution. Universal Convexi cation via Risk-Aversion ... Abstract We develop a framework for convexifying a general class of optimization problems. We analyze the suboptimality of the solution to the convexi ed problem relative to the original nonconvex problem, and prove ad-ditive approximation guarantees under some assumptions. Jun 2, 2014 · In this context, our convexification approach falls under the well-studied paradigm of risk -sensitive Markov Decision Processes. Universal Convexification via Risk-Aversion Krishnamurthy Dvijotham Dept of Computer Science & Engg University of Washington Seattle, WA 98195 Maryam Fazel Dept of Electrical Engg University of Washington Seattle, WA 98195 Abstract We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution ... Bibliographic details on Universal Convexification via Risk-Aversion . Dvijotham Krishnamurthy, CalTech Abstract: We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution to the convexified problem relative to the original nonconvex problem and prove additive approximation guarantees."}
+{"idx": 2, "title": "Universal Convexi cation via Risk-Aversion - UAI", "date": "", "ddg_snippet": "Universal Convexi cation via Risk-Aversion ... Abstract We develop a framework for convexifying a general class of optimization problems. We analyze the suboptimality of the solution to the convexi ed problem relative to the original nonconvex problem, and prove ad-ditive approximation guarantees under some assumptions.", "subpage_snippet": "", "source": "auai.org", "link": "https://auai.org/uai2014/proceedings/individuals/121.pdf", "content": "Universal Convexi cation via Risk-Aversion ... Abstract We develop a framework for convexifying a general class of optimization problems. We analyze the suboptimality of the solution to the convexi ed problem relative to the original nonconvex problem, and prove ad-ditive approximation guarantees under some assumptions."}
+{"idx": 3, "title": "(PDF) Universal Convexification via Risk-Aversion - ResearchGate", "date": "", "ddg_snippet": "Jun 2, 2014 · In this context, our convexification approach falls under the well-studied paradigm of risk -sensitive Markov Decision Processes.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/262840964_Universal_Convexification_via_Risk-Aversion", "content": "Jun 2, 2014 · In this context, our convexification approach falls under the well-studied paradigm of risk -sensitive Markov Decision Processes."}
+{"idx": 4, "title": "Universal Convexification via Risk-Aversion - paperzz.com", "date": "", "ddg_snippet": "Universal Convexification via Risk-Aversion Krishnamurthy Dvijotham Dept of Computer Science & Engg University of Washington Seattle, WA 98195 Maryam Fazel Dept of Electrical Engg University of Washington Seattle, WA 98195 Abstract We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution ...", "subpage_snippet": "", "source": "paperzz.com", "link": "https://paperzz.com/doc/6870215/universal-convexification-via-risk-aversion", "content": "Universal Convexification via Risk-Aversion Krishnamurthy Dvijotham Dept of Computer Science & Engg University of Washington Seattle, WA 98195 Maryam Fazel Dept of Electrical Engg University of Washington Seattle, WA 98195 Abstract We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution ..."}
+{"idx": 5, "title": "\"Universal Convexification via Risk-Aversion.\" - dblp", "date": "", "ddg_snippet": "Bibliographic details on Universal Convexification via Risk-Aversion .", "subpage_snippet": "", "source": "dblp.org", "link": "https://dblp.org/rec/journals/corr/DvijothamFT14", "content": "Bibliographic details on Universal Convexification via Risk-Aversion ."}
+{"idx": 6, "title": "Universal Convexification via Risk Aversion", "date": "", "ddg_snippet": "Dvijotham Krishnamurthy, CalTech Abstract: We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution to the convexified problem relative to the original nonconvex problem and prove additive approximation guarantees.", "subpage_snippet": "", "source": "lccc.control.lth.se", "link": "http://lccc.control.lth.se/abstractID=680.html", "content": "Dvijotham Krishnamurthy, CalTech Abstract: We develop a framework for convexifying a fairly general class of optimization problems. Under additional assumptions, we analyze the suboptimality of the solution to the convexified problem relative to the original nonconvex problem and prove additive approximation guarantees."}
+{"idx": 7, "title": "Universal Convexification via Risk-Aversion", "date": "", "ddg_snippet": "by K Dvijotham · Cited by 11 — Making the problem convex using risk - aversion and theorem 3.1 leads to the green curve that only preserves the robust minimum as the unique global optimum. −3.", "subpage_snippet": "", "source": "www.roboti.us", "link": "https://www.roboti.us/lab/papers/DvijothamUAI14.pdf", "content": "by K Dvijotham · Cited by 11 — Making the problem convex using risk - aversion and theorem 3.1 leads to the green curve that only preserves the robust minimum as the unique global optimum. −3."}
+{"idx": 8, "title": "Universal Convexification via Risk-Aversion - ADS", "date": "", "ddg_snippet": "by K Dvijotham · 2014 · Cited by 11 — Universal Convexification via Risk - Aversion ... Abstract. We develop a framework for convexifying a fairly general class of optimization problems. Under ...", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "https://ui.adsabs.harvard.edu/abs/2014arXiv1406.0554D/abstract", "content": "by K Dvijotham · 2014 · Cited by 11 — Universal Convexification via Risk - Aversion ... Abstract. We develop a framework for convexifying a fairly general class of optimization problems. Under ..."}
+{"idx": 9, "title": "arXiv:1406.0554v1 [cs.SY] 3 Jun 2014", "date": "", "ddg_snippet": "by K Dvijotham · 2014 · Cited by 11 — Making the problem convex using risk - aversion and theorem 3.1 leads to the green curve that only preserves the robust minimum as the unique.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1406.0554", "content": "by K Dvijotham · 2014 · Cited by 11 — Making the problem convex using risk - aversion and theorem 3.1 leads to the green curve that only preserves the robust minimum as the unique."}
diff --git a/data/sampled_jsons/VRSBench_Visual_Remote_Sensing_Benchmark_Li_et_al..jsonl b/data/sampled_jsons/VRSBench_Visual_Remote_Sensing_Benchmark_Li_et_al..jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cb3897782e41522de9e4b8b6a25d7663d79ae2b9
--- /dev/null
+++ b/data/sampled_jsons/VRSBench_Visual_Remote_Sensing_Benchmark_Li_et_al..jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2406.12384] VRSBench : A Versatile Vision-Language Benchmark ...", "date": "", "ddg_snippet": "View a PDF of the paper titled VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding, by Xiang Li and 2 other authors.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2406.12384", "content": "View a PDF of the paper titled VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding, by Xiang Li and 2 other authors."}
+{"idx": 1, "title": "VRSBench", "date": "", "ddg_snippet": "This benchmark comprises 29,614 images, with 29,614 human-verified detailed captions, 52,472 object references, and 123,221 question-answer pairs. It facilitates the training and evaluation of vision-language models across a broad spectrum of remote sensing image understanding tasks.", "subpage_snippet": "", "source": "vrsbench.github.io", "link": "https://vrsbench.github.io/", "content": "This benchmark comprises 29,614 images, with 29,614 human-verified detailed captions, 52,472 object references, and 123,221 question-answer pairs. It facilitates the training and evaluation of vision-language models across a broad spectrum of remote sensing image understanding tasks."}
+{"idx": 2, "title": "GitHub - lx709/ VRSBench", "date": "", "ddg_snippet": "VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding.Vision-language models in remote sensing : Current progress and future trends.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/lx709/VRSBench", "content": "VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding.Vision-language models in remote sensing : Current progress and future trends."}
+{"idx": 3, "title": "VRSBench : A Versatile Vision-Language Benchmark", "date": "", "ddg_snippet": "Remote Sensing Visual Question Answering Datasets. VRSBench provides a comprehensive benchmark for developing and evaluating generalist vision-language models in both remote sensing and computer vision.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=kwDOxOmGE0", "content": "Remote Sensing Visual Question Answering Datasets. VRSBench provides a comprehensive benchmark for developing and evaluating generalist vision-language models in both remote sensing and computer vision."}
+{"idx": 4, "title": "GitHub - Jack-bo1220/Awesome- Remote - Sensing -Foundation-Models", "date": "", "ddg_snippet": "- Remote Sensing ChatGPT: Solving Remote Sensing Tasks with ChatGPT and Visual Models. VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding.", "subpage_snippet": "", "source": "www.hubp.de", "link": "https://www.hubp.de/Jack-bo1220/Awesome-Remote-Sensing-Foundation-Models", "content": "- Remote Sensing ChatGPT: Solving Remote Sensing Tasks with ChatGPT and Visual Models. VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding."}
+{"idx": 5, "title": "VRSBench : A Versatile Vision-Language Benchmark Dataset for...", "date": "", "ddg_snippet": "• VRSBench contains a large collection of remote sensing images paired with natural language descriptions, providing a versatile benchmark for tasks like image captioning, visual question answering, and multimodal reasoning.", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/vrsbench-versatile-vision-language-benchmark-dataset-remote", "content": "• VRSBench contains a large collection of remote sensing images paired with natural language descriptions, providing a versatile benchmark for tasks like image captioning, visual question answering, and multimodal reasoning."}
+{"idx": 6, "title": "xiang709/ VRSBench · Datasets at Hugging Face", "date": "", "ddg_snippet": "VRSBench is a Versatile Vision-Language Benchmark for Remote Sensing Image Understanding. It consists of 29,614 remote sensing images with detailed captions, 52,472 object refers, and 3123,221 visual question-answer pairs.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/datasets/xiang709/VRSBench", "content": "VRSBench is a Versatile Vision-Language Benchmark for Remote Sensing Image Understanding. It consists of 29,614 remote sensing images with detailed captions, 52,472 object refers, and 3123,221 visual question-answer pairs."}
+{"idx": 7, "title": "Advancements in Vision–Language Models for Remote Sensing ...", "date": "", "ddg_snippet": "Pure visual tasks focus on analyzing remote sensing images to extract meaningful information about the Earth’s surface. Vision–language tasks, which combine natural language processing with visual data analysis, have opened new avenues for remote sensing applications.", "subpage_snippet": "", "source": "www.thefreelibrary.com", "link": "https://www.thefreelibrary.com/Advancements+in+Vision–Language+Models+for+Remote+Sensing:+Datasets,...-a0823355239", "content": "Pure visual tasks focus on analyzing remote sensing images to extract meaningful information about the Earth’s surface. Vision–language tasks, which combine natural language processing with visual data analysis, have opened new avenues for remote sensing applications."}
+{"idx": 8, "title": "Human Benchmark", "date": "", "ddg_snippet": "Human Benchmark . Measure your abilities with brain games and cognitive tests. Get Started.", "subpage_snippet": "", "source": "humanbenchmark.com", "link": "https://humanbenchmark.com/", "content": "Human Benchmark . Measure your abilities with brain games and cognitive tests. Get Started."}
+{"idx": 9, "title": "PowerPoint Presentation", "date": "", "ddg_snippet": "VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image. Understanding. Xiang Li KAUST. Jing Ding KAUST.", "subpage_snippet": "", "source": "nips.cc", "link": "https://nips.cc/media/neurips-2024/Slides/97530.pdf", "content": "VRSBench : A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image. Understanding. Xiang Li KAUST. Jing Ding KAUST."}
diff --git a/data/sampled_jsons/Waymo_Open_Dataset_200000_frames_1000_segments_20_seconds_10Hz_official_year_2020.jsonl b/data/sampled_jsons/Waymo_Open_Dataset_200000_frames_1000_segments_20_seconds_10Hz_official_year_2020.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..165a0620c58cfb7502931fbd033e98da958e8580
--- /dev/null
+++ b/data/sampled_jsons/Waymo_Open_Dataset_200000_frames_1000_segments_20_seconds_10Hz_official_year_2020.jsonl
@@ -0,0 +1,3 @@
+{"idx": 0, "title": "Waymo Shares Autonomous Vehicle Dataset for Machine ...", "date": "", "ddg_snippet": "4 Sept 2019 — The dataset consists of 1,000 segments of 20 seconds each, collected at 10Hz (i.e., 200,000 frames ) which contain: Synchronized data from ...", "subpage_snippet": "", "source": "www.infoq.com", "link": "https://www.infoq.com/news/2019/09/waymo-machine-learning-dataset/", "content": "4 Sept 2019 — The dataset consists of 1,000 segments of 20 seconds each, collected at 10Hz (i.e., 200,000 frames ) which contain: Synchronized data from ..."}
+{"idx": 1, "title": "Dataset list - A list of the biggest machine learning datasets", "date": "", "ddg_snippet": "The Waymo Open Dataset currently contains lidar and camera data from 1,000 segments (20s each): 1,000 segments of 20s each, collected at 10Hz ( 200,000 frames ) ...", "subpage_snippet": "", "source": "www.datasetlist.com", "link": "https://www.datasetlist.com/", "content": "The Waymo Open Dataset currently contains lidar and camera data from 1,000 segments (20s each): 1,000 segments of 20s each, collected at 10Hz ( 200,000 frames ) ..."}
+{"idx": 2, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}
diff --git a/data/sampled_jsons/Winter_et_al_sybilhunter_Tor_anomalous_circuits_year_2020-2023.jsonl b/data/sampled_jsons/Winter_et_al_sybilhunter_Tor_anomalous_circuits_year_2020-2023.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..21766fb4de66fa82e6784d88d7f14b46be9b925b
--- /dev/null
+++ b/data/sampled_jsons/Winter_et_al_sybilhunter_Tor_anomalous_circuits_year_2020-2023.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Tor — Википедия", "date": "", "ddg_snippet": "Tor — свободное и открытое программное обеспечение для реализации второго и третьего поколения так называемой луковой маршрутизации. Это система прокси-серверов, позволяющая устанавливать анонимное сетевое соединение, защищённое от прослушивания.", "subpage_snippet": "", "source": "ru.wikipedia.org", "link": "https://ru.wikipedia.org/wiki/Tor", "content": "Tor — свободное и открытое программное обеспечение для реализации второго и третьего поколения так называемой луковой маршрутизации. Это система прокси-серверов, позволяющая устанавливать анонимное сетевое соединение, защищённое от прослушивания."}
+{"idx": 1, "title": "Anomalous keys in Tor relays | Request PDF", "date": "", "ddg_snippet": "Little is known about real-world Sybils in the Tor network, and we lack practical tools and methods to expose Sybil attacks. In this work, we develop sybilhunter , the first system for detecting Sybil relays based on their appearance, such as configuration; and behavior...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/315782281_Anomalous_keys_in_Tor_relays", "content": "Little is known about real-world Sybils in the Tor network, and we lack practical tools and methods to expose Sybil attacks. In this work, we develop sybilhunter , the first system for detecting Sybil relays based on their appearance, such as configuration; and behavior..."}
+{"idx": 2, "title": "NullHypothesis/ sybilhunter : Hunting for Sybils and anomalies in...", "date": "", "ddg_snippet": "Sybilhunter is a command line tool written in Go to discover and analyse Sybil relays in the Tor network. It does so by implementing a number of analysis techniques that take as input archived Tor network data.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/NullHypothesis/sybilhunter", "content": "Sybilhunter is a command line tool written in Go to discover and analyse Sybil relays in the Tor network. It does so by implementing a number of analysis techniques that take as input archived Tor network data."}
+{"idx": 3, "title": "ESP32 LED Blink Example | Circuits 4you.com", "date": "", "ddg_snippet": "Circuits 4you.com. Arduino interfacing, circuits tutorials with code and ebooks, Step by step guides for all sensor modules used for arduino. Programming tips and tricks.", "subpage_snippet": "", "source": "circuits4you.com", "link": "https://circuits4you.com/2018/02/02/esp32-led-blink-example/", "content": "Circuits 4you.com. Arduino interfacing, circuits tutorials with code and ebooks, Step by step guides for all sensor modules used for arduino. Programming tips and tricks."}
+{"idx": 4, "title": "Solar panel as darkness detector | All About Circuits | Forum", "date": "", "ddg_snippet": "Hello found this this circuit Solar panel darknes detector on the internet i build this circuit on a bread board.", "subpage_snippet": "", "source": "forum.allaboutcircuits.com", "link": "https://forum.allaboutcircuits.com/threads/solar-panel-as-darkness-detector.103120/", "content": "Hello found this this circuit Solar panel darknes detector on the internet i build this circuit on a bread board."}
+{"idx": 5, "title": "Carley E. Winter et al . Comparative Biochemistry and Physiology Part...", "date": "", "ddg_snippet": "Carley E. Winter , Clare L. Kilgour, Colin J. Brauner, Patricia M. Schulte, Chris M. Wood. The effects of pulse exposures to road salt at various stages of early development in rainbow trout (Oncorhynchus mykiss). Comparative Biochemistry and Physiology Part C: Toxicology & Pharmacology.", "subpage_snippet": "", "source": "zoology.ubc.ca", "link": "https://zoology.ubc.ca/article/publication/carley-e-winter-et-al-comparative-biochemistry-and-physiology-part-c", "content": "Carley E. Winter , Clare L. Kilgour, Colin J. Brauner, Patricia M. Schulte, Chris M. Wood. The effects of pulse exposures to road salt at various stages of early development in rainbow trout (Oncorhynchus mykiss). Comparative Biochemistry and Physiology Part C: Toxicology & Pharmacology."}
+{"idx": 6, "title": "Honey Onions: a Framework for Characterizing and", "date": "", "ddg_snippet": "For example, Biryukov et al . [34], document their ndings on probing the network topology and connectivity of Tor relays. The authors demonstrate how the leakage of the Tor network topology can be used in attacks to traceback from an exit node to a small set of possible entry nodes.", "subpage_snippet": "", "source": "www.khoury.northeastern.edu", "link": "https://www.khoury.northeastern.edu/home/noubir/publications-local/SN2016.pdf", "content": "For example, Biryukov et al . [34], document their ndings on probing the network topology and connectivity of Tor relays. The authors demonstrate how the leakage of the Tor network topology can be used in attacks to traceback from an exit node to a small set of possible entry nodes."}
+{"idx": 7, "title": "circuito.io", "date": "", "ddg_snippet": "circuito.io is an automatic generator of circuit schematics and code. You plan your circuit by picking the high level electronic blocks, and circuito.io does all the rest. It calculates all the electrical requirements for each components and produces...", "subpage_snippet": "", "source": "www.circuito.io", "link": "https://www.circuito.io/app", "content": "circuito.io is an automatic generator of circuit schematics and code. You plan your circuit by picking the high level electronic blocks, and circuito.io does all the rest. It calculates all the electrical requirements for each components and produces..."}
+{"idx": 8, "title": "Learn how to use Tinkercad - Tinkercad", "date": "", "ddg_snippet": "Learn how to design and print in 3D, code, and create circuits using our free, easy-to-use app Tinkercad.", "subpage_snippet": "", "source": "www.tinkercad.com", "link": "https://www.tinkercad.com/learn", "content": "Learn how to design and print in 3D, code, and create circuits using our free, easy-to-use app Tinkercad."}
+{"idx": 9, "title": "Почему мы боимся и что с этим делать: тревожный мозг...", "date": "", "ddg_snippet": "Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155-184.Bandelow, B., et al . (2017). Efficacy of treatments for anxiety disorders: a meta-analysis. International Clinical Psychopharmacology, 32(4), 223-233.", "subpage_snippet": "", "source": "PsychoSearch.ru", "link": "https://PsychoSearch.ru/practice/prakticheskaya-psikhologiya/928-why-we-fear-role-of-the-amygdala", "content": "Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155-184.Bandelow, B., et al . (2017). Efficacy of treatments for anxiety disorders: a meta-analysis. International Clinical Psychopharmacology, 32(4), 223-233."}
diff --git a/data/sampled_jsons/abstract_of_Direct_Preference_Optimization_Your_Language_Model_is_Secretly_a_Reward_Model.jsonl b/data/sampled_jsons/abstract_of_Direct_Preference_Optimization_Your_Language_Model_is_Secretly_a_Reward_Model.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..95a0f76cccf99a4c450621d0905e5a43c0737625
--- /dev/null
+++ b/data/sampled_jsons/abstract_of_Direct_Preference_Optimization_Your_Language_Model_is_Secretly_a_Reward_Model.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "While large-scale unsupervised language models (LMs) learn broad world knowledge and some reasoning skills, achieving precise control of their behavior is difficult due to the completely unsupervised nature of their training. Existing methods for gaining such steerability collect human labels of the relative quality of model generations and fine-tune the unsupervised LM to align with these ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2305.18290", "content": "While large-scale unsupervised language models (LMs) learn broad world knowledge and some reasoning skills, achieving precise control of their behavior is difficult due to the completely unsupervised nature of their training. Existing methods for gaining such steerability collect human labels of the relative quality of model generations and fine-tune the unsupervised LM to align with these ..."}
+{"idx": 1, "title": "PDF Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/a85b405ed65c6477a4fe8302b5e06ce7-Paper-Conference.pdf", "content": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning."}
+{"idx": 2, "title": "Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for sampling from the LM during fine-tuning or performing significant hyperparameter tuning.", "subpage_snippet": "", "source": "ui.adsabs.harvard.edu", "link": "https://ui.adsabs.harvard.edu/abs/2023arXiv230518290R/abstract", "content": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for sampling from the LM during fine-tuning or performing significant hyperparameter tuning."}
+{"idx": 3, "title": "Direct preference optimization | Proceedings of the 37th International ...", "date": "", "ddg_snippet": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/abs/10.5555/3666122.3668460", "content": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning."}
+{"idx": 4, "title": "PDF Direct Preference Optimization(DPO)", "date": "", "ddg_snippet": "Direct Preference Optimization(DPO) Rafailov, R., Sharma, A ., Mitchell, E., Manning, C.D., Ermon, S. and Finn, C., 2024. Direct preference optimization : Your language model is secretly a reward model . Advances in Neural Information Processing Systems, 36. Presented by: Di Mu, An Cao", "subpage_snippet": "", "source": "www.cs.toronto.edu", "link": "https://www.cs.toronto.edu/~cmaddis/courses/csc2541_w25/presentations/mu_cao_dpo.pdf", "content": "Direct Preference Optimization(DPO) Rafailov, R., Sharma, A ., Mitchell, E., Manning, C.D., Ermon, S. and Finn, C., 2024. Direct preference optimization : Your language model is secretly a reward model . Advances in Neural Information Processing Systems, 36. Presented by: Di Mu, An Cao"}
+{"idx": 5, "title": "Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "A new parameterization of the reward model in RLHF that enables extraction of the corresponding optimal policy in closed form is introduced, allowing us to solve the standard RLHF problem with only a simple classification loss. While large-scale unsupervised language models (LMs) learn broad world knowledge and some reasoning skills, achieving precise control of their behavior is difficult due ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/Direct-Preference-Optimization:-Your-Language-Model-Rafailov-Sharma/0d1c76d45afa012ded7ab741194baf142117c495", "content": "A new parameterization of the reward model in RLHF that enables extraction of the corresponding optimal policy in closed form is introduced, allowing us to solve the standard RLHF problem with only a simple classification loss. While large-scale unsupervised language models (LMs) learn broad world knowledge and some reasoning skills, achieving precise control of their behavior is difficult due ..."}
+{"idx": 6, "title": "Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant and computationally lightweight, eliminating the need for fitting a reward model , sampling from ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/371136760_Direct_Preference_Optimization_Your_Language_Model_is_Secretly_a_Reward_Model", "content": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant and computationally lightweight, eliminating the need for fitting a reward model , sampling from ..."}
+{"idx": 7, "title": "Abstract Direct Preference Optimization: Your Language Model is ...", "date": "", "ddg_snippet": "In this paper, we show how to directly optimize a language model to adhere to human preferences , without explicit reward modeling or reinforcement learning. We propose Direct Preference Optimiza-tion (DPO), an algorithm that implicitly optimizes the same objective as existing RLHF algorithms ( reward maximization with a KL-divergence constraint) but is simple to implement and straight-forward ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2305.18290v3", "content": "In this paper, we show how to directly optimize a language model to adhere to human preferences , without explicit reward modeling or reinforcement learning. We propose Direct Preference Optimiza-tion (DPO), an algorithm that implicitly optimizes the same objective as existing RLHF algorithms ( reward maximization with a KL-divergence constraint) but is simple to implement and straight-forward ..."}
+{"idx": 8, "title": "NeurIPS 2023 Direct Preference Optimization: Your Language Model is ...", "date": "", "ddg_snippet": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning.", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2023/oral/73865", "content": "The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight, eliminating the need for fitting a reward model , sampling from the LM during fine-tuning, or performing significant hyperparameter tuning."}
+{"idx": 9, "title": "Direct Preference Optimization: Your Language Model is Secretly a ...", "date": "", "ddg_snippet": "View 6 comments: Curious why this phenomenon occurs- intuitively, why is it the case that DPO generalizes better than PPO to a new distribution? My best guess would be that it is simply because the intermediate step o...", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/overview/2305.18290", "content": "View 6 comments: Curious why this phenomenon occurs- intuitively, why is it the case that DPO generalizes better than PPO to a new distribution? My best guess would be that it is simply because the intermediate step o..."}
diff --git a/data/sampled_jsons/accessibility_tree_GUI_automation_text_to_action.jsonl b/data/sampled_jsons/accessibility_tree_GUI_automation_text_to_action.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f4bcbacdb133532b2f49f5bc14e9ab507f37c8cd
--- /dev/null
+++ b/data/sampled_jsons/accessibility_tree_GUI_automation_text_to_action.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ShowUI: One Vision-Language-Action Model for GUI Visual Agent", "date": "", "ddg_snippet": "Early efforts in GUI automation have primarily focused on developing language agents [12, 46, 54] that rely on closed-source, API-based LLMs like GPT-4 [32]. These agents leverage text -rich metadata like HTML inputs or accessibility trees to perform navigation and other tasks.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Lin_ShowUI_One_Vision-Language-Action_Model_for_GUI_Visual_Agent_CVPR_2025_paper.pdf", "content": "Early efforts in GUI automation have primarily focused on developing language agents [12, 46, 54] that rely on closed-source, API-based LLMs like GPT-4 [32]. These agents leverage text -rich metadata like HTML inputs or accessibility trees to perform navigation and other tasks."}
+{"idx": 1, "title": "Accessibility infrastructure for UI toolkits - GitHub ShowUI: Advanced Open-Source Vision-Language-Action Model for GUI Appendix G: Active Accessibility Bridge to UI Automation ShowUI: A Vision-Language-Action Model for GUI Visual Agents ... Accessibility infrastructure for UI toolkits - GitHub Accessibility tree - Glossary | MDN Appendix G: Active Accessibility Bridge to UI Automation ShowUI: Advanced Open-Source Vision-Language- Action Model for … ShowUI : One Vision-Language-Action Model for GUI Visual Agent ShowUI: Advanced Open-Source Vision-Language- Action Model for … Accessibility tree - MDN Web Docs", "date": "", "ddg_snippet": "There are numerous UI toolkits, and new ones continue to proliferate. So far, only the largest UI toolkit projects, with corporate backing, implement accessibility for users who require assistive technologies, such as blind people using screen readers. If the long tail of UI toolkits are ever going to be made fully accessible, then we must pool as ... See full list on github.com Data schema It has often been said that data structures are more important than code. This is certainly the case with AccessKit. The heart of AccessKit is a schema that defines all of the data required to make a UI accessible to screen readers and other assistive technologies. This schema represents a tree structure, in which each node is either a UI element such as a button or text box, or a grouping of elements such as a document, pane, or window. Each node has an integer ID, a role (e.g. button or window), and a variety of optional attributes. The schema also defines actions that can be requested by assistive technologies, such as moving the keyboard focus, invoking a button, or selecting text . The schema is based largely on Chromium's cross-platform accessibility abstraction.The canonical definition of the schema is in the Rust programming language. Rust was chosen for its combination of efficiency and an expressive type system.When both the toolkit and the platform adapter (see below) are written in Rust or another language that can efficiently call functions defined in Rust (such as C or C++), the data defined in the schema can be passed back and forth with no serialization overhead. In other cases, serialization will be required to minimize the overhead of language interoperability. Because the schema supports serde, each language binding of this type can choose its own serialization format.The schema is defined in the accesskit crate, in the common directory. It has not yet been stabilized, but it is reasonably complete. At this point there is just one known issue: the schema doesn't yet define any events. While some events in platform accessibility APIs, such as focus change and property changes, can be implied from tree updates, other events, such as ad-hoc announcements meant for screen reader users, cannot. Platform adapters These are the libraries that implement platform accessibility APIs. The following platform adapters are currently implemented:•The Windows adapter, which implements the UI Automation API, is available in the accesskit_windows crate, in the platforms/windows directory. It doesn't yet support all possible widget types, but it can now be used to make real, non-trivial applications accessible. In particular, it supports both single-line and multi-line text edit controls, but not yet rich text .•The macOS adapter, which implements the Cocoa NSAccessibility protocol, is available in the accesskit_macos crate, in the platforms/macos directory. It is roughly at feature parity with the Windows adapter, including support for text edit controls.•The Unix adapter, which implements the D-Bus-based AT-SPI protocol, is available in the accesskit_unix crate, in the platforms/unix directory. This adapter doesn't yet fully support text edit controls. It is also not yet usable with the Orca screen reader, due to a keyboard input handling issue that we are working with the appropriate GNOME development teams to solve.The following adapters are planned:•Android•iOS/tvOS•Web (creating a hidden HTML DOM)The interaction between the provider (toolkit or application) and the platform adapter is also inspired by Chromium. Because Chromium has a multi-process architecture and does not allow synchronous IPC from the browser process to the sandboxed renderer processes, the browser process cannot pull accessibility information from the renderer on demand. Instead, the renderer process pushes data to the browser process. The renderer process initially pushes a complete accessibility tree , then it pushes incremental updates. The browser process only needs to send a request to the renderer process when an assistive technology requests one of the actions mentioned above. In AccessKit, the platform adapter is like the Chromium browser process, and the UI toolkit is like the Chromium renderer process, except that both components run in the same process and communicate through normal function calls rather than IPC.One notable consequence of this design is that only the platform adapter needs to retain a complete accessibility tree in memory. That means that this design is suitable for immediate-mode GUI toolkits, as long as they can provide a stable ID for each UI element.The platform adapters are written primarily in Rust. We've chosen Rust for its combination of reliability and efficiency, including safe concurrency, which is especially important in modern software. Some future adapters may need to be partially written in another language, such as Java or Kotlin for the Android adapter. Consumer library Some of the code required by the platform adapters is platform-independent. This code is in the accesskit_consumer crate, in the consumer directory. In addition to platform adapters, this library may also be useful for implementing embedded assistive technologies, such as a screen reader running directly inside an application, for platforms that don't yet have an AccessKit platform adapter, or for devices that don't have platform support for accessibility at all, such as game consoles and appliances. See full list on github.com Contributions to AccessKit are welcome. Please see CONTRIBUTING.md. See full list on github.com AccessKit is licensed under the Apache License, Version 2.0 or the MIT license, at your option. Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in AccessKit by you, as defined in the Apache License, shall be dual-licensed as above, without any additional terms or conditions. The list of authors for copyright purposes is in the AUTHORS file. Significant portions of AccessKit are derived from Chromium and are covered by its BSD-style license. See full list on github.com Dec 9, 2024 · The early GUI assistants depended basically on text -based information like HTML or accessibility trees . Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client. Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks. What is accessibility infrastructure for UI toolkits? Accessibility infrastructure for UI toolkits AccessKit makes it easier to implement accessibility, for screen readers and other assistive technologies, in toolkits that render their own user interface elements. What is the accessibility tree? The accessibility tree contains accessibility-related information for most HTML elements . What is Microsoft Active Accessibility and UI Automation? By bridging Microsoft Active Accessibility and UI Automation together, Microsoft Active Accessibility -based clients, such as a screenreader on Windows XP, can programmatically interact with UI Automation -based providers of UI Automation , such as a Windows Presentation Foundation (WPF) application. Why are GUI assistants so hard to understand? The early GUI assistants depended basically on text-based information like HTML or accessibility trees . Because of this, it was hard for them to perceive UI visuals as a human does and interact with the elements without text description. Can showui improve GUI visual agents? From the data as-pect, with a carefully curated, high-quality instruction-following dataset, ShowUI achieves strong performance, with a lightweight model size. These results demonstrate ShowUI’s potential to advance GUI visual agents towards more human-like interaction and perception. Limitation and Future work. What is a graphic user interface (GUI) assistant? Graphical User Interface (GUI) assistants assist users to interact with digital appliances and applications . They can be an ordinary voice-activated help or a complex system with understanding and responding to a natural language command. Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/AccessKit/accesskit", "content": "There are numerous UI toolkits, and new ones continue to proliferate. So far, only the largest UI toolkit projects, with corporate backing, implement accessibility for users who require assistive technologies, such as blind people using screen readers. If the long tail of UI toolkits are ever going to be made fully accessible, then we must pool as ... See full list on github.com Data schema It has often been said that data structures are more important than code. This is certainly the case with AccessKit. The heart of AccessKit is a schema that defines all of the data required to make a UI accessible to screen readers and other assistive technologies. This schema represents a tree structure, in which each node is either a UI element such as a button or text box, or a grouping of elements such as a document, pane, or window. Each node has an integer ID, a role (e.g. button or window), and a variety of optional attributes. The schema also defines actions that can be requested by assistive technologies, such as moving the keyboard focus, invoking a button, or selecting text . The schema is based largely on Chromium's cross-platform accessibility abstraction.The canonical definition of the schema is in the Rust programming language. Rust was chosen for its combination of efficiency and an expressive type system.When both the toolkit and the platform adapter (see below) are written in Rust or another language that can efficiently call functions defined in Rust (such as C or C++), the data defined in the schema can be passed back and forth with no serialization overhead. In other cases, serialization will be required to minimize the overhead of language interoperability. Because the schema supports serde, each language binding of this type can choose its own serialization format.The schema is defined in the accesskit crate, in the common directory. It has not yet been stabilized, but it is reasonably complete. At this point there is just one known issue: the schema doesn't yet define any events. While some events in platform accessibility APIs, such as focus change and property changes, can be implied from tree updates, other events, such as ad-hoc announcements meant for screen reader users, cannot. Platform adapters These are the libraries that implement platform accessibility APIs. The following platform adapters are currently implemented:•The Windows adapter, which implements the UI Automation API, is available in the accesskit_windows crate, in the platforms/windows directory. It doesn't yet support all possible widget types, but it can now be used to make real, non-trivial applications accessible. In particular, it supports both single-line and multi-line text edit controls, but not yet rich text .•The macOS adapter, which implements the Cocoa NSAccessibility protocol, is available in the accesskit_macos crate, in the platforms/macos directory. It is roughly at feature parity with the Windows adapter, including support for text edit controls.•The Unix adapter, which implements the D-Bus-based AT-SPI protocol, is available in the accesskit_unix crate, in the platforms/unix directory. This adapter doesn't yet fully support text edit controls. It is also not yet usable with the Orca screen reader, due to a keyboard input handling issue that we are working with the appropriate GNOME development teams to solve.The following adapters are planned:•Android•iOS/tvOS•Web (creating a hidden HTML DOM)The interaction between the provider (toolkit or application) and the platform adapter is also inspired by Chromium. Because Chromium has a multi-process architecture and does not allow synchronous IPC from the browser process to the sandboxed renderer processes, the browser process cannot pull accessibility information from the renderer on demand. Instead, the renderer process pushes data to the browser process. The renderer process initially pushes a complete accessibility tree , then it pushes incremental updates. The browser process only needs to send a request to the renderer process when an assistive technology requests one of the actions mentioned above. In AccessKit, the platform adapter is like the Chromium browser process, and the UI toolkit is like the Chromium renderer process, except that both components run in the same process and communicate through normal function calls rather than IPC.One notable consequence of this design is that only the platform adapter needs to retain a complete accessibility tree in memory. That means that this design is suitable for immediate-mode GUI toolkits, as long as they can provide a stable ID for each UI element.The platform adapters are written primarily in Rust. We've chosen Rust for its combination of reliability and efficiency, including safe concurrency, which is especially important in modern software. Some future adapters may need to be partially written in another language, such as Java or Kotlin for the Android adapter. Consumer library Some of the code required by the platform adapters is platform-independent. This code is in the accesskit_consumer crate, in the consumer directory. In addition to platform adapters, this library may also be useful for implementing embedded assistive technologies, such as a screen reader running directly inside an application, for platforms that don't yet have an AccessKit platform adapter, or for devices that don't have platform support for accessibility at all, such as game consoles and appliances. See full list on github.com Contributions to AccessKit are welcome. Please see CONTRIBUTING.md. See full list on github.com AccessKit is licensed under the Apache License, Version 2.0 or the MIT license, at your option. Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in AccessKit by you, as defined in the Apache License, shall be dual-licensed as above, without any additional terms or conditions. The list of authors for copyright purposes is in the AUTHORS file. Significant portions of AccessKit are derived from Chromium and are covered by its BSD-style license. See full list on github.com Dec 9, 2024 · The early GUI assistants depended basically on text -based information like HTML or accessibility trees . Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client. Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks. What is accessibility infrastructure for UI toolkits? Accessibility infrastructure for UI toolkits AccessKit makes it easier to implement accessibility, for screen readers and other assistive technologies, in toolkits that render their own user interface elements. What is the accessibility tree? The accessibility tree contains accessibility-related information for most HTML elements . What is Microsoft Active Accessibility and UI Automation? By bridging Microsoft Active Accessibility and UI Automation together, Microsoft Active Accessibility -based clients, such as a screenreader on Windows XP, can programmatically interact with UI Automation -based providers of UI Automation , such as a Windows Presentation Foundation (WPF) application. Why are GUI assistants so hard to understand? The early GUI assistants depended basically on text-based information like HTML or accessibility trees . Because of this, it was hard for them to perceive UI visuals as a human does and interact with the elements without text description. Can showui improve GUI visual agents? From the data as-pect, with a carefully curated, high-quality instruction-following dataset, ShowUI achieves strong performance, with a lightweight model size. These results demonstrate ShowUI’s potential to advance GUI visual agents towards more human-like interaction and perception. Limitation and Future work. What is a graphic user interface (GUI) assistant? Graphical User Interface (GUI) assistants assist users to interact with digital appliances and applications . They can be an ordinary voice-activated help or a complex system with understanding and responding to a natural language command. Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers."}
+{"idx": 2, "title": "ShowUI: Advanced Open-Source Vision-Language-Action Model for GUI Appendix G: Active Accessibility Bridge to UI Automation ShowUI: A Vision-Language-Action Model for GUI Visual Agents ... Accessibility infrastructure for UI toolkits - GitHub Accessibility tree - Glossary | MDN Appendix G: Active Accessibility Bridge to UI Automation ShowUI: Advanced Open-Source Vision-Language- Action Model for … ShowUI : One Vision-Language-Action Model for GUI Visual Agent ShowUI: Advanced Open-Source Vision-Language- Action Model for … Accessibility tree - MDN Web Docs", "date": "", "ddg_snippet": "Dec 9, 2024 · The early GUI assistants depended basically on text -based information like HTML or accessibility trees . Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client. Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks. What is accessibility infrastructure for UI toolkits? Accessibility infrastructure for UI toolkits AccessKit makes it easier to implement accessibility, for screen readers and other assistive technologies, in toolkits that render their own user interface elements. What is the accessibility tree? The accessibility tree contains accessibility-related information for most HTML elements . What is Microsoft Active Accessibility and UI Automation? By bridging Microsoft Active Accessibility and UI Automation together, Microsoft Active Accessibility -based clients, such as a screenreader on Windows XP, can programmatically interact with UI Automation -based providers of UI Automation , such as a Windows Presentation Foundation (WPF) application. Why are GUI assistants so hard to understand? The early GUI assistants depended basically on text-based information like HTML or accessibility trees . Because of this, it was hard for them to perceive UI visuals as a human does and interact with the elements without text description. Can showui improve GUI visual agents? From the data as-pect, with a carefully curated, high-quality instruction-following dataset, ShowUI achieves strong performance, with a lightweight model size. These results demonstrate ShowUI’s potential to advance GUI visual agents towards more human-like interaction and perception. Limitation and Future work. What is a graphic user interface (GUI) assistant? Graphical User Interface (GUI) assistants assist users to interact with digital appliances and applications . They can be an ordinary voice-activated help or a complex system with understanding and responding to a natural language command. Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers.", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/aimonks/showui-advanced-open-source-vision-language-action-model-for-gui-aea8c8f0223a", "content": "Dec 9, 2024 · The early GUI assistants depended basically on text -based information like HTML or accessibility trees . Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client. Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks. What is accessibility infrastructure for UI toolkits? Accessibility infrastructure for UI toolkits AccessKit makes it easier to implement accessibility, for screen readers and other assistive technologies, in toolkits that render their own user interface elements. What is the accessibility tree? The accessibility tree contains accessibility-related information for most HTML elements . What is Microsoft Active Accessibility and UI Automation? By bridging Microsoft Active Accessibility and UI Automation together, Microsoft Active Accessibility -based clients, such as a screenreader on Windows XP, can programmatically interact with UI Automation -based providers of UI Automation , such as a Windows Presentation Foundation (WPF) application. Why are GUI assistants so hard to understand? The early GUI assistants depended basically on text-based information like HTML or accessibility trees . Because of this, it was hard for them to perceive UI visuals as a human does and interact with the elements without text description. Can showui improve GUI visual agents? From the data as-pect, with a carefully curated, high-quality instruction-following dataset, ShowUI achieves strong performance, with a lightweight model size. These results demonstrate ShowUI’s potential to advance GUI visual agents towards more human-like interaction and perception. Limitation and Future work. What is a graphic user interface (GUI) assistant? Graphical User Interface (GUI) assistants assist users to interact with digital appliances and applications . They can be an ordinary voice-activated help or a complex system with understanding and responding to a natural language command. Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers."}
+{"idx": 3, "title": "Appendix G: Active Accessibility Bridge to UI Automation", "date": "", "ddg_snippet": "Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client.", "subpage_snippet": "", "source": "learn.microsoft.com", "link": "https://learn.microsoft.com/en-us/windows/win32/winauto/appendix-g--active-accessibility-bridge-to-ui-automation", "content": "Jul 14, 2025 · The following tables map the Microsoft Active Accessibility IAccessible interface methods and properties to UI Automation . Use these tables to determine appropriate coding practices for developing your Microsoft Active Accessibility -based client."}
+{"idx": 4, "title": "ShowUI: A Vision-Language-Action Model for GUI Visual Agents ...", "date": "", "ddg_snippet": "Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks.", "subpage_snippet": "", "source": "www.aiinteliigence.com", "link": "https://www.aiinteliigence.com/2024/12/02/showui-a-vision-language-action-model-for-gui-visual-agents-that-addresses-key-challenges-in-ui-visual-and-action-modeling/", "content": "Dec 2, 2024 · Early attempts at GUI automation focused on language-based agents that relied on closed-source, API-based Large Language Models like GPT-4. These initial approaches primarily utilized text -rich metadata such as HTML inputs and accessibility trees to perform navigation and related tasks."}
+{"idx": 5, "title": "Accessibility tree - MDN Web Docs", "date": "", "ddg_snippet": "Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers.", "subpage_snippet": "", "source": "developer.mozilla.org", "link": "https://developer.mozilla.org/en-US/docs/Glossary/Accessibility_tree", "content": "Jul 11, 2025 · The DOM tree contains objects representing all the markup's elements, attributes, and text nodes. Browsers then create an accessibility tree based on the DOM tree , which is used by platform-specific Accessibility APIs to provide a representation that can be understood by assistive technologies, such as screen readers."}
+{"idx": 6, "title": "Microsoft Active Accessibility and UI Automation Compared", "date": "", "ddg_snippet": "14 Jul 2025 — UI Automation represents the UI as a hierarchical tree of automation elements, and provides a single interface for navigating the tree .", "subpage_snippet": "", "source": "learn.microsoft.com", "link": "https://learn.microsoft.com/en-us/windows/win32/winauto/microsoft-active-accessibility-and-ui-automation-compared", "content": "14 Jul 2025 — UI Automation represents the UI as a hierarchical tree of automation elements, and provides a single interface for navigating the tree ."}
+{"idx": 7, "title": "Expose basic accessibility information - Windows apps", "date": "", "ddg_snippet": "4 Jan 2023 — The UI Automation framework has a concept of tree views, where UI Automation clients can retrieve the relationships between elements in a UI ...", "subpage_snippet": "", "source": "learn.microsoft.com", "link": "https://learn.microsoft.com/en-us/windows/apps/design/accessibility/basic-accessibility-information", "content": "4 Jan 2023 — The UI Automation framework has a concept of tree views, where UI Automation clients can retrieve the relationships between elements in a UI ..."}
+{"idx": 8, "title": "ui automation - How can Accessibility Insights succeed in ...", "date": "", "ddg_snippet": "I'm using Microsoft's UIAutomation framework to retrieve information from an app that is built on the Chromium Embedded Framework. My code is failing to find ...", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/75871516/how-can-accessibility-insights-succeed-in-walking-uiautomation-tree-of-chromium", "content": "I'm using Microsoft's UIAutomation framework to retrieve information from an app that is built on the Chromium Embedded Framework. My code is failing to find ..."}
+{"idx": 9, "title": "The Accessibility Tree: A Training Guide for Advanced Web ...", "date": "", "ddg_snippet": "The Accessibility Tree is a hierarchical construct of objects that include accessible names and descriptions, plus supporting states and properties.", "subpage_snippet": "", "source": "whatsock.com", "link": "https://whatsock.com/training/", "content": "The Accessibility Tree is a hierarchical construct of objects that include accessible names and descriptions, plus supporting states and properties."}
diff --git a/data/sampled_jsons/advantage_of_constant-Q_transform_over_short-time_fourier_transform_for_music_signals.jsonl b/data/sampled_jsons/advantage_of_constant-Q_transform_over_short-time_fourier_transform_for_music_signals.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7f4de070b27a17c36a2c7945ff5c92749ab4d8ef
--- /dev/null
+++ b/data/sampled_jsons/advantage_of_constant-Q_transform_over_short-time_fourier_transform_for_music_signals.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Constant-Q transform - Wikipedia", "date": "", "ddg_snippet": "In mathematics and signal processing, the constant-Q transform and variable-Q transform , simply known as CQT and VQT, transforms a data series to the frequency domain. It is related to the Fourier transform [1] and very closely related to the complex Morlet wavelet transform . [2] Its design is suited for musical representation.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Constant-Q_transform", "content": "In mathematics and signal processing, the constant-Q transform and variable-Q transform , simply known as CQT and VQT, transforms a data series to the frequency domain. It is related to the Fourier transform [1] and very closely related to the complex Morlet wavelet transform . [2] Its design is suited for musical representation."}
+{"idx": 1, "title": "PDF The Constant-Q Harmonic Coefficients: A timbre feature designed for ...", "date": "", "ddg_snippet": "Unlike the discrete Fourier transform (DFT), which uses the same window length for every frequency bin, leading to a linear frequency resolution, the CQT uses window lengths that decrease with increasing frequency, with a ratio of cen-ter frequency to frequency width—also known as the quality factor (Q)—which is constant , leading to a ...", "subpage_snippet": "", "source": "zafarrafii.com", "link": "https://zafarrafii.com/Documents/Journals/Rafii+-+The+Constant-Q+Harmonic+Coefficients+A+Timbre+Feature+designed+for+Music+Signals+-+2022.pdf", "content": "Unlike the discrete Fourier transform (DFT), which uses the same window length for every frequency bin, leading to a linear frequency resolution, the CQT uses window lengths that decrease with increasing frequency, with a ratio of cen-ter frequency to frequency width—also known as the quality factor (Q)—which is constant , leading to a ..."}
+{"idx": 2, "title": "PDF The Constant Q Transform - Machine Learning Group", "date": "", "ddg_snippet": "1 Introduction The constant Q transform as introduced in [Brown, 1991] is very close related to the Fourier transform . Like the Fourier transform a constant Q transform is a bank of lters, but in contrast to the former it has geometrically spaced center frequencies fk = f0 2k b (k = 0; : : :), where b dictates the number of lters per octave.", "subpage_snippet": "", "source": "doc.ml.tu-berlin.de", "link": "https://doc.ml.tu-berlin.de/bbci/material/publications/Bla_constQ.pdf", "content": "1 Introduction The constant Q transform as introduced in [Brown, 1991] is very close related to the Fourier transform . Like the Fourier transform a constant Q transform is a bank of lters, but in contrast to the former it has geometrically spaced center frequencies fk = f0 2k b (k = 0; : : :), where b dictates the number of lters per octave."}
+{"idx": 3, "title": "PDF Constant-q Transform Toolbox for Music Processing", "date": "", "ddg_snippet": "1. INTRODUCTION Constant-Q transform (CQT) here refers to a technique that transforms a time -domain signal x(n) into the time -frequency domain so that the center frequencies of the fre-quency bins are geometrically spaced and their Q-factors are all equal. In effect, this means that the frequency res-olution is better for low frequencies and the time resolu-tion is better for high frequencies ...", "subpage_snippet": "", "source": "core.ac.uk", "link": "https://core.ac.uk/download/pdf/144846462.pdf", "content": "1. INTRODUCTION Constant-Q transform (CQT) here refers to a technique that transforms a time -domain signal x(n) into the time -frequency domain so that the center frequencies of the fre-quency bins are geometrically spaced and their Q-factors are all equal. In effect, this means that the frequency res-olution is better for low frequencies and the time resolu-tion is better for high frequencies ..."}
+{"idx": 4, "title": "PDF Pitch shifting of audio signals using the constant-Q transform", "date": "", "ddg_snippet": "ABSTRACT Pitch-scale modifications of polyphonic music are usually per-formed by manipulating the time -frequency representation of the input signal . Most approaches proposed in the past are thereby based on the Fourier transform although its linear frequency bin spacing is known to be inadequate to some degree for analysing and processing music signals . Recently invertible constant-Q trans ...", "subpage_snippet": "", "source": "www.dafx12.york.ac.uk", "link": "https://www.dafx12.york.ac.uk/papers/dafx12_submission_81.pdf", "content": "ABSTRACT Pitch-scale modifications of polyphonic music are usually per-formed by manipulating the time -frequency representation of the input signal . Most approaches proposed in the past are thereby based on the Fourier transform although its linear frequency bin spacing is known to be inadequate to some degree for analysing and processing music signals . Recently invertible constant-Q trans ..."}
+{"idx": 5, "title": "Evaluation of spectral transforms for music signal analysis", "date": "", "ddg_snippet": "In this paper we present a study on the spectral analysis of music signals comparing the time domain representation, the short-time Fourier transform (STFT) and the constant-Q transform (CQT) which are additionally combined with different signal -dependent transforms . The comparison is carried out with respect to the spectral compactness, the data compression ability and the temporal continuity ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/abstract/document/6701843", "content": "In this paper we present a study on the spectral analysis of music signals comparing the time domain representation, the short-time Fourier transform (STFT) and the constant-Q transform (CQT) which are additionally combined with different signal -dependent transforms . The comparison is carried out with respect to the spectral compactness, the data compression ability and the temporal continuity ..."}
+{"idx": 6, "title": "(PDF) Constant-Q transform toolbox for music processing", "date": "", "ddg_snippet": "This paper proposes a computationally efficient method for computing the constant-Q transform (CQT) of a time -domain signal . CQT refers to a time -frequency represen-tation where the frequency bins ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/228523955_Constant-Q_transform_toolbox_for_music_processing", "content": "This paper proposes a computationally efficient method for computing the constant-Q transform (CQT) of a time -domain signal . CQT refers to a time -frequency represen-tation where the frequency bins ..."}
+{"idx": 7, "title": "Comparing Short-time Fourier Transform (Stft) and Constant-q Transform ...", "date": "", "ddg_snippet": "This study compares the Short-Time Fourier Transform (STFT) and Constant-Q Transform (CQT) techniques in capturing the intricate spectral characteristics of the rebana, a traditional Malay frame drum known for its rich harmonic content and transient attacks.", "subpage_snippet": "", "source": "ejournals.swu.ac.th", "link": "https://ejournals.swu.ac.th/index.php/vss/article/view/16553", "content": "This study compares the Short-Time Fourier Transform (STFT) and Constant-Q Transform (CQT) techniques in capturing the intricate spectral characteristics of the rebana, a traditional Malay frame drum known for its rich harmonic content and transient attacks."}
+{"idx": 8, "title": "Constant-Q Transform (CQT) Overview - emergentmind.com", "date": "", "ddg_snippet": "Learn how the Constant-Q Transform uses logarithmic scaling and adaptive filter lengths for efficient, real- time audio analysis in music and speech.", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/topics/constant-q-transform-cqt", "content": "Learn how the Constant-Q Transform uses logarithmic scaling and adaptive filter lengths for efficient, real- time audio analysis in music and speech."}
+{"idx": 9, "title": "PDF Music Style Transfer Using Constant-Q Transform Spectrograms", "date": "", "ddg_snippet": "To mitigate this, they recommend mel-frequency and constant-Q transform (CQT) spectrograms over log-magnitude spectrograms. Generation of symbolic music , such as sheet music , MIDI or piano roll, has also been topical [3]. Symbolic representations of audio allow for an unambiguous well-defined representation of music .", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/978-3-031-03789-4_13.pdf", "content": "To mitigate this, they recommend mel-frequency and constant-Q transform (CQT) spectrograms over log-magnitude spectrograms. Generation of symbolic music , such as sheet music , MIDI or piano roll, has also been topical [3]. Symbolic representations of audio allow for an unambiguous well-defined representation of music ."}
diff --git a/data/sampled_jsons/arXiv2401.01192.jsonl b/data/sampled_jsons/arXiv2401.01192.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b9523be63469f266fcadb6a2b0a4f440b00e4d5d
--- /dev/null
+++ b/data/sampled_jsons/arXiv2401.01192.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Deep-ELA: Deep Exploratory Landscape Analysis with Self ...", "date": "", "ddg_snippet": "by MV Seiler · 2024 · Cited by 26 — Abstract page for arXiv paper 2401.01192 : Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained Transformers for ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2401.01192", "content": "by MV Seiler · 2024 · Cited by 26 — Abstract page for arXiv paper 2401.01192 : Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained Transformers for ..."}
+{"idx": 1, "title": "and Multi-Objective Continuous Optimization Problems", "date": "", "ddg_snippet": "by MV Seiler · 2024 · Cited by 26 — arXiv : 2401.01192 v2 [cs.LG] 29 Jul 2024. Page 2. M.V. Seiler et al. 1 Introduction and Related Work. Optimization problems, often found at the ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2401.01192", "content": "by MV Seiler · 2024 · Cited by 26 — arXiv : 2401.01192 v2 [cs.LG] 29 Jul 2024. Page 2. M.V. Seiler et al. 1 Introduction and Related Work. Optimization problems, often found at the ..."}
+{"idx": 2, "title": "and Multi-Objective Continuous Optimization Problems", "date": "", "ddg_snippet": "arXiv : 2401.01192 v1 [cs.LG] 02 Jan 2024. Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained Transformers for Single- and Multi- ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2401.01192v1", "content": "arXiv : 2401.01192 v1 [cs.LG] 02 Jan 2024. Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained Transformers for Single- and Multi- ..."}
+{"idx": 3, "title": "Pascal Kerschke from Technische Universität Dresden", "date": "", "ddg_snippet": "Published by arXiv in arXiv · https://doi.org/10.48550/ arXiv . 2401.01192 · Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained ...", "subpage_snippet": "", "source": "www.scilit.com", "link": "https://www.scilit.com/scholars/9183319", "content": "Published by arXiv in arXiv · https://doi.org/10.48550/ arXiv . 2401.01192 · Deep-ELA: Deep Exploratory Landscape Analysis with Self-Supervised Pretrained ..."}
+{"idx": 4, "title": "and Multi-Objective Continuous Optimization Problems", "date": "", "ddg_snippet": "This paper introduces **Deep-ELA**, a hybrid approach that combines deep learning and Exploratory Landscape Analysis (ELA) features for analyzing continuous ...", "subpage_snippet": "", "source": "www.themoonlight.io", "link": "https://www.themoonlight.io/es/review/deep-ela-deep-exploratory-landscape-analysis-with-self-supervised-pretrained-transformers-for-single-and-multi-objective-continuous-optimization-problems", "content": "This paper introduces **Deep-ELA**, a hybrid approach that combines deep learning and Exploratory Landscape Analysis (ELA) features for analyzing continuous ..."}
+{"idx": 5, "title": "A Comparison of Multi-Objective Optimization Algorithms", "date": "", "ddg_snippet": "5 Feb 2025 — The goal of this paper is to compare the performance of several commonly used meta-heuristic algorithms on a machine learning (ML) surrogate of a BPS model.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3712729.3712795", "content": "5 Feb 2025 — The goal of this paper is to compare the performance of several commonly used meta-heuristic algorithms on a machine learning (ML) surrogate of a BPS model."}
+{"idx": 6, "title": "Towards Robustness and Explainability of Automatic ...", "date": "", "ddg_snippet": "17 Jul 2025 — The goal of automated algorithm selection is to identify the most suitable algorithm for each specific problem instance, thereby enhancing ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/45808", "content": "17 Jul 2025 — The goal of automated algorithm selection is to identify the most suitable algorithm for each specific problem instance, thereby enhancing ..."}
+{"idx": 7, "title": "Scalarization-based Exploratory Landscape Analysis for ...", "date": "", "ddg_snippet": "In this study, we investigated two scalarizing approaches: (1) decomposition and (2) non-dominated sorting. Through experiments on bi-objective continuous ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1145/3712256.3726378", "content": "In this study, we investigated two scalarizing approaches: (1) decomposition and (2) non-dominated sorting. Through experiments on bi-objective continuous ..."}
+{"idx": 8, "title": "Have we hit a wall in algorithm selection generalization?", "date": "", "ddg_snippet": "by G Cenikj · 2025 · Cited by 4 — This study evaluates the generalizability of AS models based on different problem representations in the context of single-objective continuous optimization.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S2210650225000525", "content": "by G Cenikj · 2025 · Cited by 4 — This study evaluates the generalizability of AS models based on different problem representations in the context of single-objective continuous optimization."}
+{"idx": 9, "title": "A TECHNICAL DETAILS Within this paper, the ...", "date": "", "ddg_snippet": "Within this paper, the aim of the landscape analysis is to profile the dynamic optimization status of the current optimization process.", "subpage_snippet": "", "source": "proceedings.iclr.cc", "link": "https://proceedings.iclr.cc/paper_files/paper/2025/file/27b89ae947228a7195861e3eb307d1a8-Supplemental-Conference.pdf", "content": "Within this paper, the aim of the landscape analysis is to profile the dynamic optimization status of the current optimization process."}
diff --git a/data/sampled_jsons/arXiv2401.12160_ScoreDec_abstract_year_2024.jsonl b/data/sampled_jsons/arXiv2401.12160_ScoreDec_abstract_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ec2ff0e8fcb925dd77dda2e29addde77cd88bb51
--- /dev/null
+++ b/data/sampled_jsons/arXiv2401.12160_ScoreDec_abstract_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[ 2401 . 12160 ] ScoreDec : A Phase-preserving High-Fidelity Audio...", "date": "", "ddg_snippet": "arXiv : 2401 . 12160 (eess). [Submitted on 22 Jan 2024].View a PDF of the paper titled ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter, by Yi-Chiao Wu and 4 other authors.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2401.12160", "content": "arXiv : 2401 . 12160 (eess). [Submitted on 22 Jan 2024].View a PDF of the paper titled ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter, by Yi-Chiao Wu and 4 other authors."}
+{"idx": 1, "title": "Scoredec : a phase-preserving high-fidelity audio codec with a", "date": "", "ddg_snippet": "Codec Avatars Lab, Pittsburgh PA, USA. arXiv : 2401 . 12160 v1 [eess.AS] 22 Jan 2024. ABSTRACT .The proposed ScoreDec attains high-fidelity speech reconstruction, preserves the original phase information, and gets rid of the tricky GAN training.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2401.12160", "content": "Codec Avatars Lab, Pittsburgh PA, USA. arXiv : 2401 . 12160 v1 [eess.AS] 22 Jan 2024. ABSTRACT .The proposed ScoreDec attains high-fidelity speech reconstruction, preserves the original phase information, and gets rid of the tricky GAN training."}
+{"idx": 2, "title": "GitHub - hwenjun18/ arxiv -daily: arxiv daily", "date": "", "ddg_snippet": "Contribute to hwenjun18/ arxiv -daily development by creating an account on GitHub.null. 2024-01-22. ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter. Yi-Chiao Wu et.al. 2401 . 12160 .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/hwenjun18/arxiv-daily", "content": "Contribute to hwenjun18/ arxiv -daily development by creating an account on GitHub.null. 2024-01-22. ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter. Yi-Chiao Wu et.al. 2401 . 12160 ."}
+{"idx": 3, "title": "ScoreDec : A Phase-preserving High-Fidelity Audio... | Papers With Code", "date": "", "ddg_snippet": "ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter.", "subpage_snippet": "", "source": "paperswithcode.com", "link": "https://paperswithcode.com/paper/scoredec-a-phase-preserving-high-fidelity", "content": "ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter."}
+{"idx": 4, "title": "ScoreDec : A Phase-preserving High-Fidelity Audio Codec with...", "date": "", "ddg_snippet": "Abstract . Although recent mainstream waveform-domain end-to-end (E2E) neural audio codecs achieve impressive coded audio quality with a very low bitrate, the quality gap between the coded and natural audio is still significant.", "subpage_snippet": "", "source": "bigpon.github.io", "link": "https://bigpon.github.io/ScoreDec_demo/", "content": "Abstract . Although recent mainstream waveform-domain end-to-end (E2E) neural audio codecs achieve impressive coded audio quality with a very low bitrate, the quality gap between the coded and natural audio is still significant."}
+{"idx": 5, "title": "Most discriminative stimuli for functional cell type identification. ( arXiv ...)", "date": "", "ddg_snippet": "In the retina, functional types can be identified by carefully selected stimuli, but this requires expert domain knowledge and biases the procedure towards previously known cell types.", "subpage_snippet": "", "source": "www.machinelearningfreaks.com", "link": "https://www.machinelearningfreaks.com/most-discriminative-stimuli-for-functional-cell-type-identification-arxiv2401-05342v1-q-bio-nc/", "content": "In the retina, functional types can be identified by carefully selected stimuli, but this requires expert domain knowledge and biases the procedure towards previously known cell types."}
+{"idx": 6, "title": "Публикация #1800 — Speech Technology (@speechtech)", "date": "", "ddg_snippet": "ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter.", "subpage_snippet": "", "source": "tgstat.com", "link": "https://tgstat.com/ru/channel/@speechtech/1800", "content": "ScoreDec : A Phase-preserving High-Fidelity Audio Codec with A Generalized Score -based Diffusion Post-filter."}
+{"idx": 7, "title": "A Low-Bitrate Neural Audio Codec Framework with Bandwidth Reduction", "date": "", "ddg_snippet": "Abstract . This paper proposes a novel neural audio codec framework which incorporates bandwidth reduction and recovery, facili-tating its application in scenarios with high sampling rates and low bitrates.", "subpage_snippet": "", "source": "www.isca-archive.org", "link": "https://www.isca-archive.org/interspeech_2024/ai24b_interspeech.pdf", "content": "Abstract . This paper proposes a novel neural audio codec framework which incorporates bandwidth reduction and recovery, facili-tating its application in scenarios with high sampling rates and low bitrates."}
+{"idx": 8, "title": "(PDF) Codec-SUPERB @ SLT 2024: A lightweight benchmark for...", "date": "", "ddg_snippet": "and Alexander Richard, “ Scoredec : A phase-preserving high-. fidelity audio codec with a generalized score -based diffusion. post-filter,” arXiv preprint arXiv : 2401 . 12160 , 2024.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/384267174_Codec-SUPERB_SLT_2024_A_lightweight_benchmark_for_neural_audio_codec_models", "content": "and Alexander Richard, “ Scoredec : A phase-preserving high-. fidelity audio codec with a generalized score -based diffusion. post-filter,” arXiv preprint arXiv : 2401 . 12160 , 2024."}
+{"idx": 9, "title": "Жидкий робот 2025: как создали 5 реальных технологий из...", "date": "", "ddg_snippet": "Lin, Y., et al. (2024). 3D-Printed Hydraulic Fluidic Logic Circuitry for Soft Robots. arXiv preprint arXiv : 2401 .16827.", "subpage_snippet": "", "source": "cifrum.kz", "link": "https://cifrum.kz/zhidkij-robot-2025-kak-sozdali-5-realnyh-tehnologij-iz-terminatora/", "content": "Lin, Y., et al. (2024). 3D-Printed Hydraulic Fluidic Logic Circuitry for Soft Robots. arXiv preprint arXiv : 2401 .16827."}
diff --git a/data/sampled_jsons/arXiv2505.00998_Algorithm_1_D_Drift.jsonl b/data/sampled_jsons/arXiv2505.00998_Algorithm_1_D_Drift.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6e3e6aae896afe581cf88d67adbe354a348f4899
--- /dev/null
+++ b/data/sampled_jsons/arXiv2505.00998_Algorithm_1_D_Drift.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Algorithm - Wikipedia", "date": "", "ddg_snippet": "The first cryptographic algorithm for deciphering encrypted code was developed by Al-Kindi, a 9th-century Arab mathematician, in A Manuscript On Deciphering Cryptographic Messages.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Algorithm", "content": "The first cryptographic algorithm for deciphering encrypted code was developed by Al-Kindi, a 9th-century Arab mathematician, in A Manuscript On Deciphering Cryptographic Messages."}
+{"idx": 1, "title": "[2505.00998] Deterministic-to-Stochastic Diverse Latent ... Deterministic-to-Stochastic Diverse Latent Feature Mapping ... arXiv:2505.00998v1 [cs.CV] 2 May 2025 Detecting Concept Drift in Neural Networks Using Chi-squared ... [2505.04263] Physics-Informed DeepONets for drift-diffusion ... Synthetic Non-stationary Data Streams for Recognition of the ... Regularized Personalization of Text-to-Image Diffusion Models ...", "date": "", "ddg_snippet": "May 2, 2025 · Abstract page for arXiv paper 2505 . 00998 : Deterministic-to-Stochastic Diverse Latent Feature Mapping for Human Motion Synthesis Specifically, we can employ a neural network v θ () with matching samples (𝒛 0, 𝒛 1 ) ∼ 𝝅 to predict the deterministic drift u (x, t) using the drift -estimate loss function: Abstract Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to ... May 7, 2025 · As the adoption of deep learning models has grown beyond human capacity for verification, meta- algorithms are needed to ensure reliable model inference. Concept drift detection is a field dedicated to identifying statistical shifts that is underutilized in monitoring neural networks that may encounter inference data with distributional characteristics diverging from their training data. Given ... May 7, 2025 · Abstract page for arXiv paper 2505 .04263: Physics-Informed DeepONets for drift -diffusion on metric graphs: simulation and parameter identification May 19, 2025 · 2. 1 Concept drift detection The available studies have shown that the evaluation of drift detection using classification quality causes a bias towards overactive methods [4]. For detection quality evaluation, synthetic data streams are often used, where the moments of concept drift occurrence are predetermined and known. In Section 4. 1 , we show that the existing objective fails to constrain distributional drift when personalizing to a small subject-specific distribution. While personalization aims to operate within the inductive biases of a pretrained model, there is no guiding principle to preserve such biases.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.00998", "content": "May 2, 2025 · Abstract page for arXiv paper 2505 . 00998 : Deterministic-to-Stochastic Diverse Latent Feature Mapping for Human Motion Synthesis Specifically, we can employ a neural network v θ () with matching samples (𝒛 0, 𝒛 1 ) ∼ 𝝅 to predict the deterministic drift u (x, t) using the drift -estimate loss function: Abstract Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to ... May 7, 2025 · As the adoption of deep learning models has grown beyond human capacity for verification, meta- algorithms are needed to ensure reliable model inference. Concept drift detection is a field dedicated to identifying statistical shifts that is underutilized in monitoring neural networks that may encounter inference data with distributional characteristics diverging from their training data. Given ... May 7, 2025 · Abstract page for arXiv paper 2505 .04263: Physics-Informed DeepONets for drift -diffusion on metric graphs: simulation and parameter identification May 19, 2025 · 2. 1 Concept drift detection The available studies have shown that the evaluation of drift detection using classification quality causes a bias towards overactive methods [4]. For detection quality evaluation, synthetic data streams are often used, where the moments of concept drift occurrence are predetermined and known. In Section 4. 1 , we show that the existing objective fails to constrain distributional drift when personalizing to a small subject-specific distribution. While personalization aims to operate within the inductive biases of a pretrained model, there is no guiding principle to preserve such biases."}
+{"idx": 2, "title": "[2505.04263] Physics-Informed DeepONets for drift-diffusion ... Synthetic Non-stationary Data Streams for Recognition of the ... Regularized Personalization of Text-to-Image Diffusion Models ...", "date": "", "ddg_snippet": "May 7, 2025 · Abstract page for arXiv paper 2505 .04263: Physics-Informed DeepONets for drift -diffusion on metric graphs: simulation and parameter identification May 19, 2025 · 2. 1 Concept drift detection The available studies have shown that the evaluation of drift detection using classification quality causes a bias towards overactive methods [4]. For detection quality evaluation, synthetic data streams are often used, where the moments of concept drift occurrence are predetermined and known. In Section 4. 1 , we show that the existing objective fails to constrain distributional drift when personalizing to a small subject-specific distribution. While personalization aims to operate within the inductive biases of a pretrained model, there is no guiding principle to preserve such biases.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.04263", "content": "May 7, 2025 · Abstract page for arXiv paper 2505 .04263: Physics-Informed DeepONets for drift -diffusion on metric graphs: simulation and parameter identification May 19, 2025 · 2. 1 Concept drift detection The available studies have shown that the evaluation of drift detection using classification quality causes a bias towards overactive methods [4]. For detection quality evaluation, synthetic data streams are often used, where the moments of concept drift occurrence are predetermined and known. In Section 4. 1 , we show that the existing objective fails to constrain distributional drift when personalizing to a small subject-specific distribution. While personalization aims to operate within the inductive biases of a pretrained model, there is no guiding principle to preserve such biases."}
+{"idx": 3, "title": "Fast Inverse Square Root — A Quake III Algorithm - YouTube", "date": "", "ddg_snippet": "In this video we will take an in depth look at the fast inverse square root and see where the mysterious number 0x5f3759df comes from. This algorithm became ...", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=p8u_k2LIZyo", "content": "In this video we will take an in depth look at the fast inverse square root and see where the mysterious number 0x5f3759df comes from. This algorithm became ..."}
+{"idx": 4, "title": "DSA Tutorial - Learn Data Structures and Algorithms - GeeksforGeeks", "date": "", "ddg_snippet": "Greedy Algorithm builds up the solution one piece at a time and chooses the next piece which gives the most obvious and immediate benefit i .e., which is the most optimal choice at that moment.", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/dsa/dsa-tutorial-learn-data-structures-and-algorithms/", "content": "Greedy Algorithm builds up the solution one piece at a time and chooses the next piece which gives the most obvious and immediate benefit i .e., which is the most optimal choice at that moment."}
+{"idx": 5, "title": "arXiv:2505.00998v1 [cs.CV] 2 May 2025", "date": "", "ddg_snippet": "Abstract Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2505.00998", "content": "Abstract Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to ..."}
+{"idx": 6, "title": "Deterministic-to-Stochastic Diverse Latent Feature ...", "date": "", "ddg_snippet": "The pseudo algorithm of the DSDFM is provided in Algorithm 1 . Report issue for preceding element. Algorithm 1 The process for generating diverse ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.00998v1", "content": "The pseudo algorithm of the DSDFM is provided in Algorithm 1 . Report issue for preceding element. Algorithm 1 The process for generating diverse ..."}
+{"idx": 7, "title": "Detecting Concept Drift in Neural Networks Using Chi-squared ...", "date": "", "ddg_snippet": "May 7, 2025 · As the adoption of deep learning models has grown beyond human capacity for verification, meta- algorithms are needed to ensure reliable model inference. Concept drift detection is a field dedicated to identifying statistical shifts that is underutilized in monitoring neural networks that may encounter inference data with distributional characteristics diverging from their training data. Given ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.04318", "content": "May 7, 2025 · As the adoption of deep learning models has grown beyond human capacity for verification, meta- algorithms are needed to ensure reliable model inference. Concept drift detection is a field dedicated to identifying statistical shifts that is underutilized in monitoring neural networks that may encounter inference data with distributional characteristics diverging from their training data. Given ..."}
+{"idx": 8, "title": "SHA- 1 : Всё об алгоритме хеширования и его... | Belov Solutions", "date": "", "ddg_snippet": "SHA- 1 (Secure Hash Algorithm 1 ) — это криптографический алгоритм хеширования, разработанный Национальным институтом стандартов и технологий США (NIST). Он генерирует хэш длиной 160 бит (40 шестнадцатеричных символов)...", "subpage_snippet": "", "source": "belov.solutions", "link": "https://belov.solutions/blog/hashing-cryptography/sha-1-vsyo-ob-algoritme-heshirovaniya/", "content": "SHA- 1 (Secure Hash Algorithm 1 ) — это криптографический алгоритм хеширования, разработанный Национальным институтом стандартов и технологий США (NIST). Он генерирует хэш длиной 160 бит (40 шестнадцатеричных символов)..."}
+{"idx": 9, "title": "Deterministic-to-Stochastic Diverse Latent Feature Mapping ...", "date": "", "ddg_snippet": "Specifically, we can employ a neural network v θ () with matching samples (𝒛 0, 𝒛 1 ) ∼ 𝝅 to predict the deterministic drift u (x, t) using the drift -estimate loss function:", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.00998", "content": "Specifically, we can employ a neural network v θ () with matching samples (𝒛 0, 𝒛 1 ) ∼ 𝝅 to predict the deterministic drift u (x, t) using the drift -estimate loss function:"}
diff --git a/data/sampled_jsons/arXiv_Taming_Knowledge_Conflicts_in_Language_Models_Gaotang_Li.jsonl b/data/sampled_jsons/arXiv_Taming_Knowledge_Conflicts_in_Language_Models_Gaotang_Li.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1ce240c32a48432e9a9e682d69bbd6eba114d4af
--- /dev/null
+++ b/data/sampled_jsons/arXiv_Taming_Knowledge_Conflicts_in_Language_Models_Gaotang_Li.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "by G Li · 2025 — Abstract: Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.10996", "content": "by G Li · 2025 — Abstract: Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge ."}
+{"idx": 1, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "14 Mar 2025 — Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge. Previous works attribute this ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.10996v1", "content": "14 Mar 2025 — Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge. Previous works attribute this ..."}
+{"idx": 2, "title": "Taming Knowledge Conflicts in Language Models", "date": "", "ddg_snippet": "by G Li · 2025 — Abstract. Language Models (LMs) often encounter knowl- edge conflicts when parametric memory con- tradicts contextual knowledge . Previous works.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.10996", "content": "by G Li · 2025 — Abstract. Language Models (LMs) often encounter knowl- edge conflicts when parametric memory con- tradicts contextual knowledge . Previous works."}
+{"idx": 3, "title": "[Literature Review] Taming Knowledge Conflicts in ...", "date": "", "ddg_snippet": "14 Mar 2025 — The paper titled \" Taming Knowledge Conflicts in Language Models \" by Gaotang Li and colleagues addresses significant issues related to knowledge conflicts in ...", "subpage_snippet": "", "source": "www.themoonlight.io", "link": "https://www.themoonlight.io/en/review/taming-knowledge-conflicts-in-language-models", "content": "14 Mar 2025 — The paper titled \" Taming Knowledge Conflicts in Language Models \" by Gaotang Li and colleagues addresses significant issues related to knowledge conflicts in ..."}
+{"idx": 4, "title": "Taming Knowledge Conflicts in Language Models | Trending arXiv ...", "date": "", "ddg_snippet": "Taming Knowledge Conflicts in Language Models . By Gaotang Li, Hanghang Tong | Learn about this paper—or listen like a podcast—on Ribbit Ribbit.", "subpage_snippet": "", "source": "ribbitribbit.co", "link": "https://ribbitribbit.co/paper/arxiv.2503.10996-Taming-Knowledge-Conflicts-in-Language-Models", "content": "Taming Knowledge Conflicts in Language Models . By Gaotang Li, Hanghang Tong | Learn about this paper—or listen like a podcast—on Ribbit Ribbit."}
+{"idx": 5, "title": "Gaotang Li", "date": "", "ddg_snippet": "arXiv preprint arXiv:2505.02387 , 2025. 35, 2025. Interpretable ... Taming Knowledge Conflicts in Language Models. G Li, Y Chen, H Tong. ICML 2025 ...", "subpage_snippet": "", "source": "scholar.google.com", "link": "https://scholar.google.com/citations?user=0aVJRykAAAAJ&hl=en", "content": "arXiv preprint arXiv:2505.02387 , 2025. 35, 2025. Interpretable ... Taming Knowledge Conflicts in Language Models. G Li, Y Chen, H Tong. ICML 2025 ..."}
+{"idx": 6, "title": "Taming Knowledge Conflicts in Language Models | AI Research ...", "date": "", "ddg_snippet": "Taming Knowledge Conflicts in Language Models . Published 3/16/2025 by GAOTANG LI, Yuzhong Chen, Hanghang Tong ... Original Paper. View on arxiv(opens in a new tab).", "subpage_snippet": "", "source": "www.aimodels.fyi", "link": "https://www.aimodels.fyi/papers/arxiv/taming-knowledge-conflicts-language-models", "content": "Taming Knowledge Conflicts in Language Models . Published 3/16/2025 by GAOTANG LI, Yuzhong Chen, Hanghang Tong ... Original Paper. View on arxiv(opens in a new tab)."}
+{"idx": 7, "title": "Gaotang Li", "date": "", "ddg_snippet": "ICML '25 SPOTLIGHT Taming Knowledge Conflict in Language Models ... More details below ⬇️ Paper: https://arxiv.org/pdf/2503.10996 ...", "subpage_snippet": "", "source": "twitter.com", "link": "https://twitter.com/GaotangLi/status/1918155972050395531", "content": "ICML '25 SPOTLIGHT Taming Knowledge Conflict in Language Models ... More details below ⬇️ Paper: https://arxiv.org/pdf/2503.10996 ..."}
+{"idx": 8, "title": "Gaotang Li", "date": "", "ddg_snippet": "Abstract: Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this ...", "subpage_snippet": "", "source": "www.catalyzex.com", "link": "https://www.catalyzex.com/author/Gaotang+Li", "content": "Abstract: Language Models (LMs) often encounter knowledge conflicts when parametric memory contradicts contextual knowledge . Previous works attribute this ..."}
+{"idx": 9, "title": "Mitigating Knowledge Conflicts in LLM Generation via Identifying ...", "date": "", "ddg_snippet": "This work proposes a novel framework, IRCAN (Identifying and Reweighting Context-Aware Neurons), to capitalize on neurons that are crucial in processing ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/IRCAN:-Mitigating-Knowledge-Conflicts-in-LLM-via-Shi-Jin/f95eb48dd381c04ee797d588c7417d71040c8ac0", "content": "This work proposes a novel framework, IRCAN (Identifying and Reweighting Context-Aware Neurons), to capitalize on neurons that are crucial in processing ..."}
diff --git a/data/sampled_jsons/attention_heads_safety_mechanism_undifferentiated_attention_scaling_contribution_ASR_increase_reason.jsonl b/data/sampled_jsons/attention_heads_safety_mechanism_undifferentiated_attention_scaling_contribution_ASR_increase_reason.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5967e14586ba6788129a7774ecab4215a57e4544
--- /dev/null
+++ b/data/sampled_jsons/attention_heads_safety_mechanism_undifferentiated_attention_scaling_contribution_ASR_increase_reason.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "On the Role of Attention Heads in Large Language Model Safety", "date": "", "ddg_snippet": "To further explore the mechanisms of the safety head , we compare different ablation methods, Undifferentiated Attention (as defined by Eq 7) and Scaling Contribution (Eq 8) on Llama-2-7b-chat (results of Vicuna-7b-v1.5 are deferred to Appendix C.3).", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.13708v1", "content": "To further explore the mechanisms of the safety head , we compare different ablation methods, Undifferentiated Attention (as defined by Eq 7) and Scaling Contribution (Eq 8) on Llama-2-7b-chat (results of Vicuna-7b-v1.5 are deferred to Appendix C.3)."}
+{"idx": 1, "title": "On the Role of Attention Heads in Large Language Model Safety", "date": "", "ddg_snippet": "Two methods, Undifferentiated Attention and Scaling Contribution , are used to ablate the attention head . Safety Head ImPortant Score (Ships) is defined to evaluate the importance of attention heads based on the change in the probability of rejection responses to harmful queries.", "subpage_snippet": "", "source": "www.bohrium.com", "link": "https://www.bohrium.com/paper/arxiv/2410.13708", "content": "Two methods, Undifferentiated Attention and Scaling Contribution , are used to ablate the attention head . Safety Head ImPortant Score (Ships) is defined to evaluate the importance of attention heads based on the change in the probability of rejection responses to harmful queries."}
+{"idx": 2, "title": "Attention heads of large language models - PMC", "date": "", "ddg_snippet": "Attention heads play a pivotal role in reasoning and are thought to share similarities with human brain functions. In this review, we explore the roles and mechanisms of attention heads to help demystify the internal reasoning processes of LLMs. We first introduce a four-stage framework inspired by the human thought process.", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11873009/", "content": "Attention heads play a pivotal role in reasoning and are thought to share similarities with human brain functions. In this review, we explore the roles and mechanisms of attention heads to help demystify the internal reasoning processes of LLMs. We first introduce a four-stage framework inspired by the human thought process."}
+{"idx": 3, "title": "Attention Mechanisms: Theory and Variations | SpringerLink", "date": "", "ddg_snippet": "Attention mechanisms form the core of transformers, enabling them to capture dependencies across input sequences. This chapter defines self- attention and multi- head attention , explores their mathematical properties, and discusses variations like cross- attention and efficient attention mechanisms , emphasizing scalability and interpretability.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-981-96-4706-4_3", "content": "Attention mechanisms form the core of transformers, enabling them to capture dependencies across input sequences. This chapter defines self- attention and multi- head attention , explores their mathematical properties, and discusses variations like cross- attention and efficient attention mechanisms , emphasizing scalability and interpretability."}
+{"idx": 4, "title": "PDF On the Role of Attention Heads in Large Language Model Safety", "date": "", "ddg_snippet": "Ships can be used for 082 attributing safety attention head . Experimental results show that on three harmful query datasets, 083 using Ships to identify safe heads and using undifferentiated attention ablation (only modifying 084 ∼ 0.006% of the parameters) can improve the attack success rate ( ASR ) of Llama-2-7b-chat 085 f", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/notes/edits/attachment?id=ewuTt1DVMb&name=pdf", "content": "Ships can be used for 082 attributing safety attention head . Experimental results show that on three harmful query datasets, 083 using Ships to identify safe heads and using undifferentiated attention ablation (only modifying 084 ∼ 0.006% of the parameters) can improve the attack success rate ( ASR ) of Llama-2-7b-chat 085 f"}
+{"idx": 5, "title": "Attention heads of large language models - ScienceDirect", "date": "", "ddg_snippet": "Attention heads play a pivotal role in reasoning and are thought to share similarities with human brain functions. In this review, we explore the roles and mechanisms of attention heads to help demystify the internal reasoning processes of LLMs. We first introduce a four-stage framework inspired by the human thought process.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S2666389925000248", "content": "Attention heads play a pivotal role in reasoning and are thought to share similarities with human brain functions. In this review, we explore the roles and mechanisms of attention heads to help demystify the internal reasoning processes of LLMs. We first introduce a four-stage framework inspired by the human thought process."}
+{"idx": 6, "title": "On the Role of Attention Heads in Large Language Model Safety", "date": "", "ddg_snippet": "We propose a novel metric which tailored for multi- head attention , the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2410.13708", "content": "We propose a novel metric which tailored for multi- head attention , the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety ."}
+{"idx": 7, "title": "GitHub - IAAR-Shanghai/Awesome-Attention-Heads: An awesome repository ...", "date": "", "ddg_snippet": "The former is crucial for revealing the functional capabilities of the model, leading to a series of studies exploring various functions within attention mechanisms , which we have termed Attention Head Mining.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/IAAR-Shanghai/Awesome-Attention-Heads", "content": "The former is crucial for revealing the functional capabilities of the model, leading to a series of studies exploring various functions within attention mechanisms , which we have termed Attention Head Mining."}
+{"idx": 8, "title": "ICLR 2025 On the Role of Attention Heads in Large Language Model Safety ...", "date": "", "ddg_snippet": "We propose an novel metric which tailored for multi- head attention , the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety .", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2025/oral/31798", "content": "We propose an novel metric which tailored for multi- head attention , the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety ."}
+{"idx": 9, "title": "PDF arXiv:2410.13708v1 [cs.CL] 17 Oct 2024 - ResearchGate", "date": "", "ddg_snippet": "ied by Scaling Contribution abla-tion. Furthermore, we observed that across various datasets, the heads identified by Undifferentiated Attention show greater consistency, whereas the heads ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/385010417_On_the_Role_of_Attention_Heads_in_Large_Language_Model_Safety/fulltext/6711f20e069cb92a811a75e8/On-the-Role-of-Attention-Heads-in-Large-Language-Model-Safety.pdf", "content": "ied by Scaling Contribution abla-tion. Furthermore, we observed that across various datasets, the heads identified by Undifferentiated Attention show greater consistency, whereas the heads ..."}
diff --git a/data/sampled_jsons/curse_of_dimensionality_machine_learning_neural_networks.jsonl b/data/sampled_jsons/curse_of_dimensionality_machine_learning_neural_networks.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..64c608447b3e9f7cb0fd9729d92fb602669cc08f
--- /dev/null
+++ b/data/sampled_jsons/curse_of_dimensionality_machine_learning_neural_networks.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Curse of dimensionality - Wikipedia", "date": "", "ddg_snippet": "5 In machine learning and insofar as predictive performance is concerned, the curse of dimensionality is used interchangeably with the peaking ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Curse_of_dimensionality", "content": "5 In machine learning and insofar as predictive performance is concerned, the curse of dimensionality is used interchangeably with the peaking ..."}
+{"idx": 1, "title": "neural networks - Does Dimensionality curse effect some models", "date": "", "ddg_snippet": "Again this is one interpretation, in reality the curse of dimensionality does in fact impact neural networks , but not at the same level as the models ...", "subpage_snippet": "", "source": "stats.stackexchange.com", "link": "https://stats.stackexchange.com/questions/186184/does-dimensionality-curse-effect-some-models-more-than-others", "content": "Again this is one interpretation, in reality the curse of dimensionality does in fact impact neural networks , but not at the same level as the models ..."}
+{"idx": 2, "title": "machine learning – X-Tough", "date": "", "ddg_snippet": "Today, machine learning techniques can be used to analyze an image and find out which object(s ... Machine Learning is more than neural networks .", "subpage_snippet": "", "source": "texaswinthem.com", "link": "https://texaswinthem.com/blog/tag/machine-learning/", "content": "Today, machine learning techniques can be used to analyze an image and find out which object(s ... Machine Learning is more than neural networks ."}
+{"idx": 3, "title": "Understanding the Curse of Dimensionality in Machine Learning", "date": "", "ddg_snippet": "In machine learning , the ‘ curse of dimensionality ’ is an important challenge that practitioners work on. ... curse of dimensionality presents ...", "subpage_snippet": "", "source": "www.hitechnectar.com", "link": "https://www.hitechnectar.com/blogs/effective-strategies-to-combat-the-curse-of-dimensional-data-analysis/", "content": "In machine learning , the ‘ curse of dimensionality ’ is an important challenge that practitioners work on. ... curse of dimensionality presents ..."}
+{"idx": 4, "title": "The Curse Of Dimensionality And How To Overcome It", "date": "", "ddg_snippet": "... abnormality detection, and more general machine learning problems are the three main areas that are affected by the curse of dimensionality ...", "subpage_snippet": "", "source": "spotintelligence.com", "link": "https://spotintelligence.com/2022/11/29/curse-of-dimensionality/", "content": "... abnormality detection, and more general machine learning problems are the three main areas that are affected by the curse of dimensionality ..."}
+{"idx": 5, "title": "The Curse of Dimensionality in Machine Learning: Challenges,", "date": "", "ddg_snippet": "Explore The Curse of Dimensionality in data analysis and machine learning , including its challenges, effects on algorithms, and techniques like PCA ...", "subpage_snippet": "", "source": "www.datacamp.com", "link": "https://www.datacamp.com/blog/curse-of-dimensionality-machine-learning", "content": "Explore The Curse of Dimensionality in data analysis and machine learning , including its challenges, effects on algorithms, and techniques like PCA ..."}
+{"idx": 6, "title": "curse of dimensionality | mt. solitary", "date": "", "ddg_snippet": "... of this great survey of “ the mathematics of modern deep learning ” , which includes a really nice explanation of why neural networks don ...", "subpage_snippet": "", "source": "www.mtsolitary.com", "link": "https://www.mtsolitary.com/20210703183435-curse-of-dimensionality/", "content": "... of this great survey of “ the mathematics of modern deep learning ” , which includes a really nice explanation of why neural networks don ..."}
+{"idx": 7, "title": "Curse of dimensionality - HandWiki", "date": "", "ddg_snippet": "5 In machine learning and insofar as predictive performance is concerned, the curse of dimensionality is used interchangeably with the peaking ...", "subpage_snippet": "", "source": "handwiki.org", "link": "https://handwiki.org/wiki/Curse_of_dimensionality", "content": "5 In machine learning and insofar as predictive performance is concerned, the curse of dimensionality is used interchangeably with the peaking ..."}
+{"idx": 8, "title": "Curse of Dimensionality in Machine Learning - Analytics Vidhya", "date": "", "ddg_snippet": "In this article, we tackle the Curse of Dimensionality in machine learning , examining its origins and impact on algorithm performance.", "subpage_snippet": "", "source": "www.analyticsvidhya.com", "link": "https://www.analyticsvidhya.com/blog/2021/04/the-curse-of-dimensionality-in-machine-learning/", "content": "In this article, we tackle the Curse of Dimensionality in machine learning , examining its origins and impact on algorithm performance."}
+{"idx": 9, "title": "Neural Networks", "date": "", "ddg_snippet": "The curse of dimensionality is the phenomena ... Neural networks are one of the most popular and powerful classes of machine learning algorithms.", "subpage_snippet": "", "source": "www.turingfinance.com", "link": "https://www.turingfinance.com/category/neural-networks/", "content": "The curse of dimensionality is the phenomena ... Neural networks are one of the most popular and powerful classes of machine learning algorithms."}
diff --git a/data/sampled_jsons/flow_matching_machine_learning_high_dimensional_generative_modeling_year_2023.jsonl b/data/sampled_jsons/flow_matching_machine_learning_high_dimensional_generative_modeling_year_2023.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e805937efc75d7231e61369cbf80929198e3df4d
--- /dev/null
+++ b/data/sampled_jsons/flow_matching_machine_learning_high_dimensional_generative_modeling_year_2023.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Published in Transactions on Machine Learning Research (03/2024)", "date": "", "ddg_snippet": "Flow matching for generative modeling .International Conference on Machine Learning (ICML), 2015. Noam Rozen, Aditya Grover, Maximilian Nickel, and Yaron Lipman. Moser flow : Divergence-based generative modeling on manifolds.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2302.00482", "content": "Flow matching for generative modeling .International Conference on Machine Learning (ICML), 2015. Noam Rozen, Aditya Grover, Maximilian Nickel, and Yaron Lipman. Moser flow : Divergence-based generative modeling on manifolds."}
+{"idx": 1, "title": "Flow Matching for Generative Modeling -Bohrium", "date": "", "ddg_snippet": "Flow Matching for Generative Modeling . 2022-10-06. 110. Cite. Share. publication cover Machine Learning .Conclusion. Flow Matching , a new simulation - free framework for training CNF models , was introduced. It uses conditional constructions for high - dimensional scalability.", "subpage_snippet": "", "source": "www.bohrium.com", "link": "https://www.bohrium.com/paper-details/flow-matching-for-generative-modeling/867749377035206795-108614", "content": "Flow Matching for Generative Modeling . 2022-10-06. 110. Cite. Share. publication cover Machine Learning .Conclusion. Flow Matching , a new simulation - free framework for training CNF models , was introduced. It uses conditional constructions for high - dimensional scalability."}
+{"idx": 2, "title": "Free Video: Discrete Flow Matching for High - Dimensional Discrete...", "date": "", "ddg_snippet": "Explore Discrete Flow Matching for generating high - dimensional discrete data. Learn about probability paths, sampling techniques, and improved generative perplexity in AI applications.", "subpage_snippet": "", "source": "www.classcentral.com", "link": "https://www.classcentral.com/course/youtube-discrete-flow-matching-andrew-campbell-345368", "content": "Explore Discrete Flow Matching for generating high - dimensional discrete data. Learn about probability paths, sampling techniques, and improved generative perplexity in AI applications."}
+{"idx": 3, "title": "Wasserstein Flow Matching : Transforming Data Creation", "date": "", "ddg_snippet": "# Computer Science # Machine Learning . Wasserstein Flow Matching : Transforming Data...What’s the Deal with Generative Modeling ?So, What is Wasserstein Flow Matching ?", "subpage_snippet": "", "source": "scisimple.com", "link": "https://scisimple.com/en/articles/2025-06-01-wasserstein-flow-matching-transforming-data-creation--a9p0d18", "content": "# Computer Science # Machine Learning . Wasserstein Flow Matching : Transforming Data...What’s the Deal with Generative Modeling ?So, What is Wasserstein Flow Matching ?"}
+{"idx": 4, "title": "An Introduction to Flow Matching and Conditional Flow Matching ...", "date": "", "ddg_snippet": "Sampling: Because Flow Matching learns a deterministic flow via an ODE, sampling can be significantly faster, requiring fewer function evaluations (NFE) from the model to generate a high -quality sample.", "subpage_snippet": "", "source": "ksml4.com", "link": "https://ksml4.com/an-introduction-to-flow-matching-and-conditional-flow-matching/", "content": "Sampling: Because Flow Matching learns a deterministic flow via an ODE, sampling can be significantly faster, requiring fewer function evaluations (NFE) from the model to generate a high -quality sample."}
+{"idx": 5, "title": "Decomposable Flow Matching (DFM)", "date": "", "ddg_snippet": "Decomposable Flow Matching (DFM) breaks down complex generative models and data analysis flows into simpler, independent parts for scalable and interpretable generation .", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/topics/decomposable-flow-matching-dfm", "content": "Decomposable Flow Matching (DFM) breaks down complex generative models and data analysis flows into simpler, independent parts for scalable and interpretable generation ."}
+{"idx": 6, "title": "(PDF) Probabilistic Forecasting via Autoregressive Flow Matching", "date": "", "ddg_snippet": "FLOW MATCHING . Ahmed El-Gazzar & Marcel van Gerven. Department of Machine Learning and Neural Computing.forecasting. 2 backgrou nd: flow M atch in g. Flow matching (FM) (Lipman et al., 2022) is a paradigm for generative modeling that enables train", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/389821674_Probabilistic_Forecasting_via_Autoregressive_Flow_Matching", "content": "FLOW MATCHING . Ahmed El-Gazzar & Marcel van Gerven. Department of Machine Learning and Neural Computing.forecasting. 2 backgrou nd: flow M atch in g. Flow matching (FM) (Lipman et al., 2022) is a paradigm for generative modeling that enables train"}
+{"idx": 7, "title": "Topic 20: What is Flow Matching ? | Turing Post", "date": "", "ddg_snippet": "Why is Flow Matching worth discussing now? It’s gaining attention for its role in top generative models like Flux (text-to-image), F5-TTS and E2-TTS (text-to-speech), and Meta’s MovieGen (text-to-video).", "subpage_snippet": "", "source": "www.turingpost.com", "link": "https://www.turingpost.com/p/flowmatching", "content": "Why is Flow Matching worth discussing now? It’s gaining attention for its role in top generative models like Flux (text-to-image), F5-TTS and E2-TTS (text-to-speech), and Meta’s MovieGen (text-to-video)."}
+{"idx": 8, "title": "Pullback Flow Matching on Data Manifolds", "date": "", "ddg_snippet": "The Pullback Flow Matching (PFM) framework introduces a novel method for simulation-free training of generative models on data manifolds, utilizing pullback geometry and isometric learning . PFM enables closed-form mappings, precise interpolation, and efficient generation .", "subpage_snippet": "", "source": "powerdrill.ai", "link": "https://powerdrill.ai/discover/discover-Pullback-Flow-Matching-cm20xfwq81rj3019gt6c7z811", "content": "The Pullback Flow Matching (PFM) framework introduces a novel method for simulation-free training of generative models on data manifolds, utilizing pullback geometry and isometric learning . PFM enables closed-form mappings, precise interpolation, and efficient generation ."}
+{"idx": 9, "title": "Energy Matching : Unifying Flow Matching and Energy-Based Models ...", "date": "", "ddg_snippet": "Improving and generalizing flow -based generative models with minibatch optimal transport.Transactions on Machine Learning Research, 2023.", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/overview/2504.10612v1", "content": "Improving and generalizing flow -based generative models with minibatch optimal transport.Transactions on Machine Learning Research, 2023."}
diff --git a/data/sampled_jsons/httpsgithub.comFLAIROxah2ac2blobproductionREADME.md.jsonl b/data/sampled_jsons/httpsgithub.comFLAIROxah2ac2blobproductionREADME.md.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..07375a21d911f272b9e5cbc7d4fbdce16dfe6571
--- /dev/null
+++ b/data/sampled_jsons/httpsgithub.comFLAIROxah2ac2blobproductionREADME.md.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Github 上月趋势 | GitHub 中文社区", "date": "", "ddg_snippet": "回车: Github 搜索. Shift+回车: Google搜索...", "subpage_snippet": "", "source": "www.github-zh.com", "link": "https://www.github-zh.com/trends/monthly", "content": "回车: Github 搜索. Shift+回车: Google搜索..."}
+{"idx": 1, "title": "AI Product Photo Generator & Editor | Create E-Commerce... | Flair.ai", "date": "", "ddg_snippet": "Create studio-quality e-commerce photoshoots in seconds with our drag-and-drop AI editor. Try it free today.", "subpage_snippet": "", "source": "flair.ai", "link": "https://flair.ai/", "content": "Create studio-quality e-commerce photoshoots in seconds with our drag-and-drop AI editor. Try it free today."}
+{"idx": 2, "title": "GitHub - ChrisTitusTech/winutil: Chris Titus Tech's Windows Utility...", "date": "", "ddg_snippet": "Usage. Winutil must be run in Admin mode because it performs system-wide tweaks. To achieve this, run PowerShell as an administrator. Here are a few ways to do it...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/ChrisTitusTech/winutil", "content": "Usage. Winutil must be run in Admin mode because it performs system-wide tweaks. To achieve this, run PowerShell as an administrator. Here are a few ways to do it..."}
+{"idx": 3, "title": "Пользователь GitHub разобрал (https...)", "date": "", "ddg_snippet": "Ну, тут Макс не оригинален. Данные тырят все. С того дня как ты впервые зарегистрировался хоть где-то в интернете.", "subpage_snippet": "", "source": "vk.com", "link": "https://vk.com/wall-10354961_219743", "content": "Ну, тут Макс не оригинален. Данные тырят все. С того дня как ты впервые зарегистрировался хоть где-то в интернете."}
+{"idx": 4, "title": "buzz Re-merged https:// github . com /jaysalvat/buzz/pull/31 by hand", "date": "", "ddg_snippet": "Buzz is a small but powerful Javascript library that allows you to easily take advantage of the new HTML5 audio element. It tries to degrade silently on non-modern browsers.", "subpage_snippet": "", "source": "gitmemories.com", "link": "https://gitmemories.com/jaysalvat/buzz/issues/49", "content": "Buzz is a small but powerful Javascript library that allows you to easily take advantage of the new HTML5 audio element. It tries to degrade silently on non-modern browsers."}
+{"idx": 5, "title": "amnezia-wg-easy/ README . md at master - forks/amnezia-wg-easy...", "date": "", "ddg_snippet": "amnezia-wg-easy - The easiest way to run WireGuard VPN + Web-based Admin UI.", "subpage_snippet": "", "source": "git.zhirov.kz", "link": "https://git.zhirov.kz/forks/amnezia-wg-easy/src/branch/master/README.md", "content": "amnezia-wg-easy - The easiest way to run WireGuard VPN + Web-based Admin UI."}
+{"idx": 6, "title": "91视频最新网址 · GitLab", "date": "", "ddg_snippet": "dizhi8. README . md .", "subpage_snippet": "", "source": "gitlab.com", "link": "https://gitlab.com/dizhi8", "content": "dizhi8. README . md ."}
+{"idx": 7, "title": "Как записывать и читать файлы, изображения в SQLite на Python...", "date": "", "ddg_snippet": "В этом материале вы узнаете, как вставлять и получать BLOB -данные из таблицы SQLite с помощью Python и модуля sqlite3.", "subpage_snippet": "", "source": "PythonRu.com", "link": "https://PythonRu.com/biblioteki/rabota-s-izobrazhenijami-i-fajlami-v-sqlite", "content": "В этом материале вы узнаете, как вставлять и получать BLOB -данные из таблицы SQLite с помощью Python и модуля sqlite3."}
+{"idx": 8, "title": "Почему не запускается fail2ban? — Хабр Q&A", "date": "", "ddg_snippet": "Ответили на вопрос 2 человека.", "subpage_snippet": "", "source": "qna.habr.com", "link": "https://qna.habr.com/q/1328574", "content": "Ответили на вопрос 2 человека."}
+{"idx": 9, "title": "Прямой эфир ТВ онлайн смотреть бесплатно каналы и трансляции...", "date": "", "ddg_snippet": "Телевидение в прямом эфире: фильмы, сериалы, шоу и передачи в онлайн-кинотеатре KION! Доступно на всех устройствах. Хорошее качество видео и звука, актуальная...", "subpage_snippet": "", "source": "kion.ru", "link": "https://kion.ru/tv", "content": "Телевидение в прямом эфире: фильмы, сериалы, шоу и передачи в онлайн-кинотеатре KION! Доступно на всех устройствах. Хорошее качество видео и звука, актуальная..."}
diff --git a/data/sampled_jsons/httpsraw.githubusercontent.comfiveaiunderstanding_safety_finetuningmainpretrain.py_year_2024.jsonl b/data/sampled_jsons/httpsraw.githubusercontent.comfiveaiunderstanding_safety_finetuningmainpretrain.py_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..abf040fe4b06f3983b6049332f795df3d7ea7296
--- /dev/null
+++ b/data/sampled_jsons/httpsraw.githubusercontent.comfiveaiunderstanding_safety_finetuningmainpretrain.py_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Sign in to Gmail - Computer - Gmail Help - Google Help", "date": "", "ddg_snippet": "Sign in to Gmail Tip: If you sign in to a public computer, make sure to sign out before you leave the computer. Learn how to sign in on a device that's not yours.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/mail/answer/8494?hl=en&co=GENIE.Platform=Desktop", "content": "Sign in to Gmail Tip: If you sign in to a public computer, make sure to sign out before you leave the computer. Learn how to sign in on a device that's not yours."}
+{"idx": 1, "title": "security - Are HTTPS headers encrypted? - Stack Overflow", "date": "", "ddg_snippet": "Oct 9, 2008 · When sending data over HTTPS , I know the content is encrypted, however I hear mixed answers about whether the headers are encrypted, or how much of the header is encrypted. How much of HTTPS heade...", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/187655/are-https-headers-encrypted", "content": "Oct 9, 2008 · When sending data over HTTPS , I know the content is encrypted, however I hear mixed answers about whether the headers are encrypted, or how much of the header is encrypted. How much of HTTPS heade..."}
+{"idx": 2, "title": "ssl - Are HTTPS URLs encrypted? - Stack Overflow", "date": "", "ddg_snippet": "Are all URLs encrypted when using TLS/SSL ( HTTPS ) encryption? I would like to know because I want all URL data to be hidden when using TLS/SSL ( HTTPS ). If TLS/SSL gives you total URL encryption th...", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/499591/are-https-urls-encrypted", "content": "Are all URLs encrypted when using TLS/SSL ( HTTPS ) encryption? I would like to know because I want all URL data to be hidden when using TLS/SSL ( HTTPS ). If TLS/SSL gives you total URL encryption th..."}
+{"idx": 3, "title": "Track: Poster Session 2 East", "date": "", "ddg_snippet": "Code is available at https :// github.com/fiveai / understandingsafetyfinetuning . Poster. #3307. A Concept-Based Explainability Framework for Large Multimodal ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2024/session/108364", "content": "Code is available at https :// github.com/fiveai / understandingsafetyfinetuning . Poster. #3307. A Concept-Based Explainability Framework for Large Multimodal ..."}
+{"idx": 4, "title": "Download and install Google Chrome", "date": "", "ddg_snippet": "How to install Chrome Important: Before you download, you can check if Chrome supports your operating system and other system requirements.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/chrome/answer/95346?hl=en&co=GENIE.Platform=Desktop", "content": "How to install Chrome Important: Before you download, you can check if Chrome supports your operating system and other system requirements."}
+{"idx": 5, "title": "How to fix \"running scripts is disabled on this system\"?", "date": "", "ddg_snippet": "Nov 1, 2020 · The \"Unrestricted\" execution policy is generally considered risky. A better choice would be \"Remote-Signed\", which doesn't block scripts created and stored locally, but does prevent scripts downloaded from the internet from running unless you specifically check and unblock them.", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/64633727/how-to-fix-running-scripts-is-disabled-on-this-system", "content": "Nov 1, 2020 · The \"Unrestricted\" execution policy is generally considered risky. A better choice would be \"Remote-Signed\", which doesn't block scripts created and stored locally, but does prevent scripts downloaded from the internet from running unless you specifically check and unblock them."}
+{"idx": 6, "title": "Google Account Help", "date": "", "ddg_snippet": "Official Google Account Help Center where you can find tips and tutorials on using Google Account and other answers to frequently asked questions.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/accounts/?hl=en", "content": "Official Google Account Help Center where you can find tips and tutorials on using Google Account and other answers to frequently asked questions."}
+{"idx": 7, "title": "Newest Questions - Stack Overflow", "date": "", "ddg_snippet": "Ask questions, find answers and collaborate at work with Stack Overflow for Teams. Try Teams for free Explore Teams", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/", "content": "Ask questions, find answers and collaborate at work with Stack Overflow for Teams. Try Teams for free Explore Teams"}
+{"idx": 8, "title": "How to recover your Google Account or Gmail", "date": "", "ddg_snippet": "If you forgot your password or username, or you can’t get verification codes, follow these steps to recover your Google Account. That way, you can use services like Gmail, Pho", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/accounts/answer/7682439?hl=en", "content": "If you forgot your password or username, or you can’t get verification codes, follow these steps to recover your Google Account. That way, you can use services like Gmail, Pho"}
+{"idx": 9, "title": "Access your Google Analytics account", "date": "", "ddg_snippet": "If you are unable to sign in to your account, try these steps.", "subpage_snippet": "", "source": "support.google.com", "link": "https://support.google.com/analytics/answer/1009692?hl=en", "content": "If you are unable to sign in to your account, try these steps."}
diff --git a/data/sampled_jsons/icml.cc_virtual2025poster46029_MultiPDENet.jsonl b/data/sampled_jsons/icml.cc_virtual2025poster46029_MultiPDENet.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c5c19da5fc9552f185b4bba00c2dc7e9c44099db
--- /dev/null
+++ b/data/sampled_jsons/icml.cc_virtual2025poster46029_MultiPDENet.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ICML Poster MultiPDENet: PDE-embedded Learning with Multi-time-stepping ...", "date": "", "ddg_snippet": "MultiPDENet : PDE-embedded Learning with Multi-time-stepping for Accelerated Flow Simulation Qi Wang · Yuan Mi · Wang Haoyun · Yi Zhang · Ruizhi Chengze · Hongsheng Liu · Ji-Rong Wen · Hao Sun", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46029", "content": "MultiPDENet : PDE-embedded Learning with Multi-time-stepping for Accelerated Flow Simulation Qi Wang · Yuan Mi · Wang Haoyun · Yi Zhang · Ruizhi Chengze · Hongsheng Liu · Ji-Rong Wen · Hao Sun"}
+{"idx": 1, "title": "ICML 2025 Posters", "date": "", "ddg_snippet": "Zoom in/out with mouse wheel; double click to reset zoom Shift-drag to select a cluster Click on dot to open ICML 2025 Posters Alt+Click opens the openreview page for the paper.", "subpage_snippet": "", "source": "icml2025.vizhub.ai", "link": "https://icml2025.vizhub.ai/", "content": "Zoom in/out with mouse wheel; double click to reset zoom Shift-drag to select a cluster Click on dot to open ICML 2025 Posters Alt+Click opens the openreview page for the paper."}
+{"idx": 2, "title": "Poster Presenter Instructions (Virtual) | Wiki.EventHosts NeurIPS/ICML ...", "date": "", "ddg_snippet": "This section applies if the conference is producing a virtual poster session. Check with the conference or workshop organizers to see if a hybrid conference is including a virtual poster session.", "subpage_snippet": "", "source": "wiki.eventhosts.cc", "link": "https://wiki.eventhosts.cc/en/reference/poster-presenter-instructions", "content": "This section applies if the conference is producing a virtual poster session. Check with the conference or workshop organizers to see if a hybrid conference is including a virtual poster session."}
+{"idx": 3, "title": "ICML 2025 Poster Instructions", "date": "", "ddg_snippet": "ICML 2025 Poster Instructions Poster Visit your My Stuff page to see your papers. Changes you make to your paper on OpenReview will not appear on icml.cc until after we manually pull the changes Missing Papers see this FAQ if you do not see all your papers Author Names On icml.cc we use the names that authors have set in their profile. Authors can change the name that is displayed by editing ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Conferences/2025/PosterInstructions", "content": "ICML 2025 Poster Instructions Poster Visit your My Stuff page to see your papers. Changes you make to your paper on OpenReview will not appear on icml.cc until after we manually pull the changes Missing Papers see this FAQ if you do not see all your papers Author Names On icml.cc we use the names that authors have set in their profile. Authors can change the name that is displayed by editing ..."}
+{"idx": 4, "title": "GitHub - smallfryy/icml2025-robotics-papers: 21 robotics ...", "date": "", "ddg_snippet": "ICML 2025 Robotics Papers A curated list of 21 robotics and robotics-adjacent papers presented at ICML 2025 . This year’s highlights include robot learning, sim2real transfer, VLAs, skill abstraction, motion planning, and scalable policy architectures. Explore interactive papers on Bytez or browse the virtual posters on ICML .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/smallfryy/icml2025-robotics-papers", "content": "ICML 2025 Robotics Papers A curated list of 21 robotics and robotics-adjacent papers presented at ICML 2025 . This year’s highlights include robot learning, sim2real transfer, VLAs, skill abstraction, motion planning, and scalable policy architectures. Explore interactive papers on Bytez or browse the virtual posters on ICML ."}
+{"idx": 5, "title": "2025 Spotlight Posters - icml.cc", "date": "", "ddg_snippet": "To the best of our knowledge, this is the first polynomial-time algorithm to achieve exact (as opposed to approximate) invariances in this setting, partially addressing a question posed by Diaz ( 2025 ) regarding the avoidance of prohibitively large and computationally intensive group averaging methods in kernel regression with exact invariances.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/events/2025SpotlightPosters", "content": "To the best of our knowledge, this is the first polynomial-time algorithm to achieve exact (as opposed to approximate) invariances in this setting, partially addressing a question posed by Diaz ( 2025 ) regarding the avoidance of prohibitively large and computationally intensive group averaging methods in kernel regression with exact invariances."}
+{"idx": 6, "title": "2025 Conference - icml.cc", "date": "", "ddg_snippet": "Announcements ICML 2025 virtual content is now free. An account is required. Beware of emails that claim they can provide a list of ICML attendees. We do not sell attendee information. Beware of Unofficial Events: Unless an event is listed on icml.cc , it is not authorized or endorsed by ICML . The ICML Career site is open. The organizing committee for 2025 has been announced! Poster Order pick ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Conferences/2025/index.html", "content": "Announcements ICML 2025 virtual content is now free. An account is required. Beware of emails that claim they can provide a list of ICML attendees. We do not sell attendee information. Beware of Unofficial Events: Unless an event is listed on icml.cc , it is not authorized or endorsed by ICML . The ICML Career site is open. The organizing committee for 2025 has been announced! Poster Order pick ..."}
+{"idx": 7, "title": "2025 Presenter Instructions - icml.cc", "date": "", "ddg_snippet": "ICML 2025 Oral Presenter Instructions Poster Oral presenters will also present papers during the poster sessions. Information on poster presentations in available here. Oral Presenters Please arrive at least 5 minutes before your session starts and check in with the session chair. Information for your paper presentation is now available on your My Stuff page under Presentation History All oral ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Conferences/2025/PresenterInstructions", "content": "ICML 2025 Oral Presenter Instructions Poster Oral presenters will also present papers during the poster sessions. Information on poster presentations in available here. Oral Presenters Please arrive at least 5 minutes before your session starts and check in with the session chair. Information for your paper presentation is now available on your My Stuff page under Presentation History All oral ..."}
+{"idx": 8, "title": "ICML 2025 Getting Started", "date": "", "ddg_snippet": "ICML 2025 Getting Started Badges/Checkin There are two registration desks (West Level 1) and (East ground floor), here is the schedule Attendees with minors need to use the West registration desk. Treat your badge like cash there is a $60 lost badge fee, Details Do not anyone to use your badge. Details Whova our Event App We will be using Whova as our official event app! Get Whova to quickly ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025", "content": "ICML 2025 Getting Started Badges/Checkin There are two registration desks (West Level 1) and (East ground floor), here is the schedule Attendees with minors need to use the West registration desk. Treat your badge like cash there is a $60 lost badge fee, Details Do not anyone to use your badge. Details Whova our Event App We will be using Whova as our official event app! Get Whova to quickly ..."}
+{"idx": 9, "title": "ICML 2025 Papers", "date": "", "ddg_snippet": "Enable Javascript in your browser to see the papers page.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/papers.html?search=agent", "content": "Enable Javascript in your browser to see the papers page."}
diff --git a/data/sampled_jsons/knowledge_distillation_teacher_student_one-way_transfer_mutual_exchange_Hinton_2015_mechanism_year_2015.jsonl b/data/sampled_jsons/knowledge_distillation_teacher_student_one-way_transfer_mutual_exchange_Hinton_2015_mechanism_year_2015.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2fe6561552fdd872398f4b253b8797d2490dbf50
--- /dev/null
+++ b/data/sampled_jsons/knowledge_distillation_teacher_student_one-way_transfer_mutual_exchange_Hinton_2015_mechanism_year_2015.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Knowledge Distillation in Federated Edge Learning: A Survey", "date": "", "ddg_snippet": "... knowledge transfer and model collaborative training, Knowledge Distillation (KD) [ 4 , 5 ] transfers knowledge from one ML model to another, allowing ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2301.05849v3", "content": "... knowledge transfer and model collaborative training, Knowledge Distillation (KD) [ 4 , 5 ] transfers knowledge from one ML model to another, allowing ..."}
+{"idx": 1, "title": "(PDF) Two-stage optimization based on heterogeneous branch", "date": "", "ddg_snippet": "... optimization based on heterogeneous branch fusion for knowledge distillation (THFKD) ( 1 ) It consists of pre-trained teacher and multi-branch student ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/393328890_Two-stage_optimization_based_on_heterogeneous_branch_fusion_for_knowledge_distillation", "content": "... optimization based on heterogeneous branch fusion for knowledge distillation (THFKD) ( 1 ) It consists of pre-trained teacher and multi-branch student ..."}
+{"idx": 2, "title": "Lightweight Contenders: Navigating Semi-Supervised Text Mining", "date": "", "ddg_snippet": "Student cohorts engage in mutual learning for collaborative optimization, where multiple students are trained together using complementary knowledge ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.00883v1", "content": "Student cohorts engage in mutual learning for collaborative optimization, where multiple students are trained together using complementary knowledge ..."}
+{"idx": 3, "title": "A lightweight speech recognition method with target-swap", "date": "", "ddg_snippet": "As a teacher - student training strategy, knowledge distillation (KD) aims to transfer knowledge from the heavy pre-trained teacher model into the ...", "subpage_snippet": "", "source": "peerj.com", "link": "https://peerj.com/articles/cs-1650/", "content": "As a teacher - student training strategy, knowledge distillation (KD) aims to transfer knowledge from the heavy pre-trained teacher model into the ..."}
+{"idx": 4, "title": "ICML 2021 Papers", "date": "", "ddg_snippet": "LogME: Practical Assessment of Pre-trained Models for Transfer Learning ... REPAINT: Knowledge Transfer in Deep Reinforcement Learning", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2021/papers.html", "content": "LogME: Practical Assessment of Pre-trained Models for Transfer Learning ... REPAINT: Knowledge Transfer in Deep Reinforcement Learning"}
+{"idx": 5, "title": "GitHub - zh320/realtime-semantic-segmentation-pytorch: PyTorch", "date": "", "ddg_snippet": "... of realtime semantic segmentation models, support multi-gpu training and validating, automatic mixed precision training, knowledge distillation ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/zh320/realtime-semantic-segmentation-pytorch", "content": "... of realtime semantic segmentation models, support multi-gpu training and validating, automatic mixed precision training, knowledge distillation ..."}
+{"idx": 6, "title": "WO2021042828A1 - Neural network model compression method and", "date": "", "ddg_snippet": "... computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge ...", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/WO2021042828A1/en", "content": "... computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge ..."}
+{"idx": 7, "title": "The Future of AI is Generative not Discriminative 5/26/2021 |", "date": "", "ddg_snippet": "People, in contrast, can learn from just a few examples, generalize far beyond their experience, and can easily transfer and reuse knowledge .", "subpage_snippet": "", "source": "www.slideshare.net", "link": "https://www.slideshare.net/slideshow/the-future-of-ai-is-generative-not-discriminative-5262021/249884337", "content": "People, in contrast, can learn from just a few examples, generalize far beyond their experience, and can easily transfer and reuse knowledge ."}
+{"idx": 8, "title": "A Computational Model of Inclusive Pedagogy: From Understanding", "date": "", "ddg_snippet": "Human education transcends mere knowledge transfer , it relies on co-adaptation dynamics — the mutual adjustment of teaching and learning strategies ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.02853v1", "content": "Human education transcends mere knowledge transfer , it relies on co-adaptation dynamics — the mutual adjustment of teaching and learning strategies ..."}
+{"idx": 9, "title": "Downloads", "date": "", "ddg_snippet": "1 -bit Adam: Communication Efficient ... Accelerated Algorithms for Smooth Convex-Concave Minimax Problems with O( 1 /k^2) Rate on Squared Gradient Norm", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Downloads/2021", "content": "1 -bit Adam: Communication Efficient ... Accelerated Algorithms for Smooth Convex-Concave Minimax Problems with O( 1 /k^2) Rate on Squared Gradient Norm"}
diff --git a/data/sampled_jsons/nuScenes_autonomous_driving_vehicle_trajectories_pedestrian_interactions_abstract.jsonl b/data/sampled_jsons/nuScenes_autonomous_driving_vehicle_trajectories_pedestrian_interactions_abstract.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3727d3e9cca0644ef2502917566190aba8886bf9
--- /dev/null
+++ b/data/sampled_jsons/nuScenes_autonomous_driving_vehicle_trajectories_pedestrian_interactions_abstract.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "(PDF) Evaluating the safety of autonomous vehicle - pedestrian ...", "date": "", "ddg_snippet": "This study aims to conduct a thorough assessment of pedestrian and cyclist safety in autonomous vehicle (AV) environments. To that end, the study utilized AV sensor data of over 1,500 driving hours from five sources in Canada, the United States, and Singapore.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/360770861_Evaluating_the_safety_of_autonomous_vehicle-pedestrian_interactions_An_extreme_value_theory_approach", "content": "This study aims to conduct a thorough assessment of pedestrian and cyclist safety in autonomous vehicle (AV) environments. To that end, the study utilized AV sensor data of over 1,500 driving hours from five sources in Canada, the United States, and Singapore."}
+{"idx": 1, "title": "Pedestrian Models for Autonomous Driving Part II: High-Level...", "date": "", "ddg_snippet": "[32] presented nuScenes a dataset for autonomous driving composed of multiple sensor data (RGB, LIDAR, RADAR) from two cities and containing 1k scenes. A summary of vehicle datasets is given in the supplementary material Sec.", "subpage_snippet": "", "source": "eprints.whiterose.ac.uk", "link": "https://eprints.whiterose.ac.uk/id/eprint/162669/1/Camara2020pedestriana_IEEE_T-ITS_Part_II.pdf", "content": "[32] presented nuScenes a dataset for autonomous driving composed of multiple sensor data (RGB, LIDAR, RADAR) from two cities and containing 1k scenes. A summary of vehicle datasets is given in the supplementary material Sec."}
+{"idx": 2, "title": "Pedestrian -Aware Motion Planning for Autonomous", "date": "", "ddg_snippet": "Index Terms—Autonomous systems, Autonomous driving , Mo-tion planning, Motion planning under uncertainty, Pedestrians.Open datasets such as nuScenes [48] provide extensive real-world driving logs but lack the dynamic feedback needed.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2504.01409", "content": "Index Terms—Autonomous systems, Autonomous driving , Mo-tion planning, Motion planning under uncertainty, Pedestrians.Open datasets such as nuScenes [48] provide extensive real-world driving logs but lack the dynamic feedback needed."}
+{"idx": 3, "title": "Image Analysis in Autonomous Vehicles: A Review", "date": "", "ddg_snippet": "Datasets such as Cityscapes, NuScenes , and CARLA form the benchmarks for training and evaluating different AI models, with unique characteristics catering to various aspects of autonomous driving .", "subpage_snippet": "", "source": "www.preprints.org", "link": "https://www.preprints.org/manuscript/202407.1857/v1/download", "content": "Datasets such as Cityscapes, NuScenes , and CARLA form the benchmarks for training and evaluating different AI models, with unique characteristics catering to various aspects of autonomous driving ."}
+{"idx": 4, "title": "Autonomous vehicle -road user dynamics for safer roads", "date": "", "ddg_snippet": "Evaluating the Safety of Autonomous Vehicle - Pedestrian Interactions : An Extreme Value Theory Approach.Processing, assessing, and enhancing the Waymo autonomous vehicle open dataset for driving behavior research.", "subpage_snippet": "", "source": "macsphere.mcmaster.ca", "link": "https://macsphere.mcmaster.ca/bitstream/11375/29321/2/alozi_abdul_razak_2023Dec_phd_civil_engineering.pdf", "content": "Evaluating the Safety of Autonomous Vehicle - Pedestrian Interactions : An Extreme Value Theory Approach.Processing, assessing, and enhancing the Waymo autonomous vehicle open dataset for driving behavior research."}
+{"idx": 5, "title": "D-STGCN: Dynamic Pedestrian Trajectory Prediction Using...", "date": "", "ddg_snippet": "Abstract . Predicting pedestrian trajectories in urban scenarios is a challenging task that has a wide range of applications, from video surveillance to autonomous driving . The task is difficult since pedestrian behavior is affected by both their individual path’s history, their interactions ...", "subpage_snippet": "", "source": "www.mdpi.com", "link": "https://www.mdpi.com/2079-9292/12/3/611", "content": "Abstract . Predicting pedestrian trajectories in urban scenarios is a challenging task that has a wide range of applications, from video surveillance to autonomous driving . The task is difficult since pedestrian behavior is affected by both their individual path’s history, their interactions ..."}
+{"idx": 6, "title": "Evaluating the safety of autonomous vehicle–pedestrian ...", "date": "", "ddg_snippet": "by AR Alozi · 2022 · Cited by 55 — Still, the impact of AVs on pedestrian safety needs to be investigated and quantified, considering the different characteristics of AV- pedestrian interactions ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S2213665722000197", "content": "by AR Alozi · 2022 · Cited by 55 — Still, the impact of AVs on pedestrian safety needs to be investigated and quantified, considering the different characteristics of AV- pedestrian interactions ..."}
+{"idx": 7, "title": "Pedestrian Models for Autonomous Driving Part II", "date": "", "ddg_snippet": "by F Camara · 2020 · Cited by 181 — Caesar et al. [32] presented nuScenes a dataset for autonomous driving composed of multiple sensor data (RGB,. LIDAR, RADAR) from two cities and ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2003.11959", "content": "by F Camara · 2020 · Cited by 181 — Caesar et al. [32] presented nuScenes a dataset for autonomous driving composed of multiple sensor data (RGB,. LIDAR, RADAR) from two cities and ..."}
+{"idx": 8, "title": "PEDESTRIAN MOTION RECONSTRUCTION:ALARGE", "date": "", "ddg_snippet": "2021) focus on vehicle trajectories with sparse pedestrian interactions . PIE ... nuscenes : A multimodal dataset for autonomous driving . In Proceedings ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/789739cfd83f5876545163c9f6ee9f74cfe121fa.pdf", "content": "2021) focus on vehicle trajectories with sparse pedestrian interactions . PIE ... nuscenes : A multimodal dataset for autonomous driving . In Proceedings ..."}
+{"idx": 9, "title": "DrivingDojo Dataset: Advancing Interactive and ...", "date": "", "ddg_snippet": "14 Oct 2024 — Autonomous driving datasets commonly used in current world model literature like nuScenes ... pedestrian interactions , 5) manually sorted ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.10738v1", "content": "14 Oct 2024 — Autonomous driving datasets commonly used in current world model literature like nuScenes ... pedestrian interactions , 5) manually sorted ..."}
diff --git a/data/sampled_jsons/observed_agreement_c_obs_probabilistic_formula_two_models_output_probabilities_year_2024.jsonl b/data/sampled_jsons/observed_agreement_c_obs_probabilistic_formula_two_models_output_probabilities_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ba292ffb2911361da627e8645ca265f47ac7f626
--- /dev/null
+++ b/data/sampled_jsons/observed_agreement_c_obs_probabilistic_formula_two_models_output_probabilities_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Great Models Think Alike and this Undermines AI Oversight", "date": "", "ddg_snippet": "We study how model similarity affects both aspects of AI oversight by proposing Chance Adjusted Probabilistic Agreement (CAPA): a metric for LM ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.04313v2", "content": "We study how model similarity affects both aspects of AI oversight by proposing Chance Adjusted Probabilistic Agreement (CAPA): a metric for LM ..."}
+{"idx": 1, "title": "Diffusion models for multivariate subsurface generation and", "date": "", "ddg_snippet": "Diffusion models (DM; Ho et al., 2020 ; Sohl-Dickstein et al., 2015 ) represent the current state-of-the-art in deep generative modeling ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.15809v1", "content": "Diffusion models (DM; Ho et al., 2020 ; Sohl-Dickstein et al., 2015 ) represent the current state-of-the-art in deep generative modeling ..."}
+{"idx": 2, "title": "34 questions with answers in PROBABILISTIC RISK ANALYSIS |", "date": "", "ddg_snippet": "Moreover, If there is any formula or methodology in which soil type from the epicenter to the target point is taken to account for obtaining an ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/topic/Probabilistic-Risk-Analysis", "content": "Moreover, If there is any formula or methodology in which soil type from the epicenter to the target point is taken to account for obtaining an ..."}
+{"idx": 3, "title": "PyMC: a modern, and comprehensive probabilistic programming", "date": "", "ddg_snippet": "Additionally, the handling of generalized linear models ( GLMs) has found a new home in the Bambi library ( Capretto et al., 2022 ), which has been ...", "subpage_snippet": "", "source": "peerj.com", "link": "https://peerj.com/articles/cs-1516/", "content": "Additionally, the handling of generalized linear models ( GLMs) has found a new home in the Bambi library ( Capretto et al., 2022 ), which has been ..."}
+{"idx": 4, "title": "Chapter 5 Bayesian hierarchical models | Introduction to", "date": "", "ddg_snippet": "5. 2 .5 By-subjects and by-items correlated varying intercept varying slopes model ( \\(M_{sih}\\) ) ... 2 .1 Accounting for measurement error in ...", "subpage_snippet": "", "source": "bruno.nicenboim.me", "link": "https://bruno.nicenboim.me/bayescogsci/ch-hierarchical.html", "content": "5. 2 .5 By-subjects and by-items correlated varying intercept varying slopes model ( \\(M_{sih}\\) ) ... 2 .1 Accounting for measurement error in ..."}
+{"idx": 5, "title": "Credible Uncertainty Quantification under Noise and System", "date": "", "ddg_snippet": "... probabilistic -forecasting and ML/DL communities conduct multi-criteria credibility evaluation [ 12 , 13 ] : they combine different proper scoring ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.03311v1", "content": "... probabilistic -forecasting and ML/DL communities conduct multi-criteria credibility evaluation [ 12 , 13 ] : they combine different proper scoring ..."}
+{"idx": 6, "title": "On the limit to the accuracy of regional-scale air quality", "date": "", "ddg_snippet": "Conse- tions and indicated that errors in the specification of the quently, comparisons of modeled and observed concentra- physical atmosphere such ...", "subpage_snippet": "", "source": "www.readkong.com", "link": "https://www.readkong.com/page/on-the-limit-to-the-accuracy-of-regional-scale-air-quality-2092367", "content": "Conse- tions and indicated that errors in the specification of the quently, comparisons of modeled and observed concentra- physical atmosphere such ..."}
+{"idx": 7, "title": "A Case Study on Quantifying Reliability under Extreme Risk", "date": "", "ddg_snippet": "... combination of three factors critical in extreme reliability assessment: (1) sparse or limited empirical data on input distributions; ( 2 ) surrogate ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2408.10083v2", "content": "... combination of three factors critical in extreme reliability assessment: (1) sparse or limited empirical data on input distributions; ( 2 ) surrogate ..."}
+{"idx": 8, "title": "Global Estimation of Subsurface Eddy Kinetic Energy of", "date": "", "ddg_snippet": "The subsurface EKE can be inferred both theoretically and numerically from sea surface observations but is limited by the issue of decreasing ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.10656v1", "content": "The subsurface EKE can be inferred both theoretically and numerically from sea surface observations but is limited by the issue of decreasing ..."}
+{"idx": 9, "title": "Chapter 14 Cross-validation | Introduction to Bayesian Data", "date": "", "ddg_snippet": "5. 2 .5 By-subjects and by-items correlated varying intercept varying slopes model ( \\(M_{sih}\\) ) ... 2 .1 Accounting for measurement error in ...", "subpage_snippet": "", "source": "bruno.nicenboim.me", "link": "https://bruno.nicenboim.me/bayescogsci/ch-cv.html", "content": "5. 2 .5 By-subjects and by-items correlated varying intercept varying slopes model ( \\(M_{sih}\\) ) ... 2 .1 Accounting for measurement error in ..."}
diff --git a/data/sampled_jsons/peeling_device_probability_concentration_inequalities_martingales.jsonl b/data/sampled_jsons/peeling_device_probability_concentration_inequalities_martingales.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8a6bd10cb4b2192d28f8c468618f8de15ec98cc5
--- /dev/null
+++ b/data/sampled_jsons/peeling_device_probability_concentration_inequalities_martingales.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Concentration inequalities for the maximum of the rescaled ...", "date": "", "ddg_snippet": "30 Oct 2013 — In this case, we can easily derive concentration inequalities on Mn using the fact that Sn is a martingale and applying Azuma-Hoeffding's ...", "subpage_snippet": "", "source": "mathoverflow.net", "link": "https://mathoverflow.net/questions/146397/concentration-inequalities-for-the-maximum-of-the-rescaled-normalized-sum-of-iid", "content": "30 Oct 2013 — In this case, we can easily derive concentration inequalities on Mn using the fact that Sn is a martingale and applying Azuma-Hoeffding's ..."}
+{"idx": 1, "title": "Mixture Martingales Revisited with Applications to ...", "date": "", "ddg_snippet": "by E Kaufmann · 2021 · Cited by 169 — Abstract. This paper presents new deviation inequalities that are valid uniformly in time under adaptive sampling in a multi-armed bandit model. 44 pages", "subpage_snippet": "", "source": "www.jmlr.org", "link": "https://www.jmlr.org/papers/volume22/18-798/18-798.pdf", "content": "by E Kaufmann · 2021 · Cited by 169 — Abstract. This paper presents new deviation inequalities that are valid uniformly in time under adaptive sampling in a multi-armed bandit model. 44 pages"}
+{"idx": 2, "title": "Martingales and potentials", "date": "", "ddg_snippet": "1 Jan 2016 — In Section 3.2, we show how martingales can be used to derive a sub- stantial generalization of our previous concentration inequalities —from the ...", "subpage_snippet": "", "source": "people.math.wisc.edu", "link": "https://people.math.wisc.edu/~roch/mdp/roch-mdp-chap3.pdf", "content": "1 Jan 2016 — In Section 3.2, we show how martingales can be used to derive a sub- stantial generalization of our previous concentration inequalities —from the ..."}
+{"idx": 3, "title": "Matrix concentration inequalities via the method of ...", "date": "", "ddg_snippet": "by L Mackey · 2012 · Cited by 190 — This paper derives exponential concentration inequalities and polynomial moment inequalities for the spectral norm of a random matrix.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1201.6002", "content": "by L Mackey · 2012 · Cited by 190 — This paper derives exponential concentration inequalities and polynomial moment inequalities for the spectral norm of a random matrix."}
+{"idx": 4, "title": "Mixture martingales revisited with applications to ...", "date": "", "ddg_snippet": "by E Kaufmann · 2021 · Cited by 169 — Abstract. This paper presents new deviation inequalities that are valid uniformly in time under adaptive sampling in a multi-armed bandit model.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/pdf/10.5555/3546258.3546504", "content": "by E Kaufmann · 2021 · Cited by 169 — Abstract. This paper presents new deviation inequalities that are valid uniformly in time under adaptive sampling in a multi-armed bandit model."}
+{"idx": 5, "title": "Probability in High Dimension - Princeton Math", "date": "", "ddg_snippet": "by R van Handel · Cited by 505 — These notes were written for the course APC 550: Probability in High Dimen- sion that I taught at Princeton in the Spring 2014 and Fall 2016 semesters. 326 pages", "subpage_snippet": "", "source": "web.math.princeton.edu", "link": "https://web.math.princeton.edu/~rvan/APC550.pdf", "content": "by R van Handel · Cited by 505 — These notes were written for the course APC 550: Probability in High Dimen- sion that I taught at Princeton in the Spring 2014 and Fall 2016 semesters. 326 pages"}
+{"idx": 6, "title": "Sequential and Adaptive Inference Based on Martingale ...", "date": "", "ddg_snippet": "by SR Howard · 2019 — Concentration inequalities play an important role in probability and statistics, giving non-asymptotic tail probability bounds for random variables or ...", "subpage_snippet": "", "source": "escholarship.org", "link": "https://escholarship.org/content/qt63m9j4hw/qt63m9j4hw.pdf", "content": "by SR Howard · 2019 — Concentration inequalities play an important role in probability and statistics, giving non-asymptotic tail probability bounds for random variables or ..."}
+{"idx": 7, "title": "Martingale methods for sequential estimation of convex ...", "date": "", "ddg_snippet": "by T Manole · 2021 · Cited by 41 — We construct an offline-to-sequential device that converts a wide array of existing offline concentration inequalities into time-uniform ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2103.09267", "content": "by T Manole · 2021 · Cited by 41 — We construct an offline-to-sequential device that converts a wide array of existing offline concentration inequalities into time-uniform ..."}
+{"idx": 8, "title": "Applications in Covariance Estimation and Linear Regression", "date": "", "ddg_snippet": "by AK Kuchibhotla · Cited by 179 — Concentration inequalities and probabilistic tail bounds form a major part of the toolkit required for such analyses. Some of the most commonly ... 71 pages", "subpage_snippet": "", "source": "faculty.wharton.upenn.edu", "link": "https://faculty.wharton.upenn.edu/wp-content/uploads/2018/04/MBS-Arxiv-V1.pdf", "content": "by AK Kuchibhotla · Cited by 179 — Concentration inequalities and probabilistic tail bounds form a major part of the toolkit required for such analyses. Some of the most commonly ... 71 pages"}
+{"idx": 9, "title": "Dealing with Unknown Variances in Best-Arm Identification", "date": "", "ddg_snippet": "by M Jourdan · 2023 · Cited by 24 — In this paper we introduce and analyze two approaches to deal with unknown variances, either by plugging in the empirical variance or by adapting the ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v201/jourdan23a/jourdan23a.pdf", "content": "by M Jourdan · 2023 · Cited by 24 — In this paper we introduce and analyze two approaches to deal with unknown variances, either by plugging in the empirical variance or by adapting the ..."}
diff --git a/data/sampled_jsons/probability_of_necessity_causal_bayesian_networks.jsonl b/data/sampled_jsons/probability_of_necessity_causal_bayesian_networks.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..dff31889c4fdfdf14ff71d6ac98247d53350803f
--- /dev/null
+++ b/data/sampled_jsons/probability_of_necessity_causal_bayesian_networks.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Intervention and Conditioning in Causal Bayesian Networks", "date": "", "ddg_snippet": "by S Galhotra · 2024 · Cited by 3 — We show that by making simple yet often realistic independence assumptions, it is possible to uniquely estimate the probability of an interventional formula.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2405.14728", "content": "by S Galhotra · 2024 · Cited by 3 — We show that by making simple yet often realistic independence assumptions, it is possible to uniquely estimate the probability of an interventional formula."}
+{"idx": 1, "title": "Bayesian networks for causal analysis in socioecological ...", "date": "", "ddg_snippet": "by R Cabañas · 2025 · Cited by 2 — This paper proposes the application of counterfactual reasoning to enhance the analysis by quantifying the degrees of necessity and sufficiency of various ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S1574954125001827", "content": "by R Cabañas · 2025 · Cited by 2 — This paper proposes the application of counterfactual reasoning to enhance the analysis by quantifying the degrees of necessity and sufficiency of various ..."}
+{"idx": 2, "title": "Intervention and Conditioning in Causal Bayesian Networks", "date": "", "ddg_snippet": "by J Halpern — Pearl (2000) gives examples showing that neither the probability of necessity nor the probability of sufficiency in a CBN can be identified; we can just ...", "subpage_snippet": "", "source": "papers.nips.cc", "link": "https://papers.nips.cc/paper_files/paper/2024/file/a2118322165fffb648d1e341ff5a5b05-Paper-Conference.pdf", "content": "by J Halpern — Pearl (2000) gives examples showing that neither the probability of necessity nor the probability of sufficiency in a CBN can be identified; we can just ..."}
+{"idx": 3, "title": "Introduction to Probabilities, Graphs, and Causal Models", "date": "", "ddg_snippet": "1.1.1 Why Probabilities? Causality connotes lawlike necessity , whereas probabilities connote exceptionality, doubt, and lack of regularity.", "subpage_snippet": "", "source": "web.cs.ucla.edu", "link": "http://web.cs.ucla.edu/~kaoru/ch1-final.pdf", "content": "1.1.1 Why Probabilities? Causality connotes lawlike necessity , whereas probabilities connote exceptionality, doubt, and lack of regularity."}
+{"idx": 4, "title": "Learning causal Bayesian networks based on ...", "date": "", "ddg_snippet": "by L Wang · 2022 · Cited by 16 — In this paper, heuristic search strategy is applied to explore the causal interpretation in the form of directed acyclic graph (DAG) for classification.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0952197622002962", "content": "by L Wang · 2022 · Cited by 16 — In this paper, heuristic search strategy is applied to explore the causal interpretation in the form of directed acyclic graph (DAG) for classification."}
+{"idx": 5, "title": "Functional Bayesian Networks for Discovering Causality ...", "date": "", "ddg_snippet": "by F Zhou · 2023 · Cited by 6 — This article develops a novel functional Bayesian network for modeling directed conditional independence and causal relationships of multivariate functional ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10840881/", "content": "by F Zhou · 2023 · Cited by 6 — This article develops a novel functional Bayesian network for modeling directed conditional independence and causal relationships of multivariate functional ..."}
+{"idx": 6, "title": "Can causal Bayes Nets compute counterfactuals? If so, are ...", "date": "", "ddg_snippet": "3 Nov 2018 — Under some conditions such as monotonicity, same thing for Probability of Necessity . There are similar possibilities for Natural Indirect ...", "subpage_snippet": "", "source": "stats.stackexchange.com", "link": "https://stats.stackexchange.com/questions/375090/can-causal-bayes-nets-compute-counterfactuals-if-so-are-they-worse-than-stru", "content": "3 Nov 2018 — Under some conditions such as monotonicity, same thing for Probability of Necessity . There are similar possibilities for Natural Indirect ..."}
+{"idx": 7, "title": "Modeling Causal Learning Using Bayesian Generic Priors ...", "date": "", "ddg_snippet": "by H Lu · Cited by 23 — Bayesian inference involves two basic components, likelihood probabilities and prior probabilities. Likelihoods assess the probability that particular observed.", "subpage_snippet": "", "source": "www.cs.jhu.edu", "link": "https://www.cs.jhu.edu/~ayuille/pubs/ucla/A205_hjlu_COGSCI2006.pdf", "content": "by H Lu · Cited by 23 — Bayesian inference involves two basic components, likelihood probabilities and prior probabilities. Likelihoods assess the probability that particular observed."}
+{"idx": 8, "title": "Intervention and Conditioning in Causal Bayesian Networks", "date": "", "ddg_snippet": "by S Galhotra · 2024 · Cited by 3 — We show that by making simple yetoften realistic independence assumptions, it is possible to uniquely estimate the probability of an interventional formula.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2024/hash/a2118322165fffb648d1e341ff5a5b05-Abstract-Conference.html", "content": "by S Galhotra · 2024 · Cited by 3 — We show that by making simple yetoften realistic independence assumptions, it is possible to uniquely estimate the probability of an interventional formula."}
+{"idx": 9, "title": "Probabilistic Modelling is Sufficient for Causal Inference", "date": "", "ddg_snippet": "by BK Mlodozeniec — In this paper, we make it clear that you can answer any causal inference question within the realm of probabilistic modelling and inference, without causal - ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=V1FP9WDKa7", "content": "by BK Mlodozeniec — In this paper, we make it clear that you can answer any causal inference question within the realm of probabilistic modelling and inference, without causal - ..."}
diff --git a/data/sampled_jsons/qcnePVejeV_Do_Not_Trust_What_They_Tell_Section_4.3_clustering_features_Intrinsic_Attribute_Similarit.jsonl b/data/sampled_jsons/qcnePVejeV_Do_Not_Trust_What_They_Tell_Section_4.3_clustering_features_Intrinsic_Attribute_Similarit.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9cf0c84dc74387c9411fdadc78cb32bd86680891
--- /dev/null
+++ b/data/sampled_jsons/qcnePVejeV_Do_Not_Trust_What_They_Tell_Section_4.3_clustering_features_Intrinsic_Attribute_Similarit.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Cluster analysis - Wikipedia", "date": "", "ddg_snippet": "The result of a cluster analysis shown as the coloring of the squares into three clusters . Cluster analysis, or clustering , is a data analysis technique aimed at partitioning a set of objects into groups such that objects within the same group exhibi...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Cluster_analysis", "content": "The result of a cluster analysis shown as the coloring of the squares into three clusters . Cluster analysis, or clustering , is a data analysis technique aimed at partitioning a set of objects into groups such that objects within the same group exhibi..."}
+{"idx": 1, "title": "Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor...", "date": "", "ddg_snippet": "Intrinsic Attribute Similarity . This feature measures the sim - ilarity of some intrinsic attributes of Tor nodes.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qcnePVejeV", "content": "Intrinsic Attribute Similarity . This feature measures the sim - ilarity of some intrinsic attributes of Tor nodes."}
+{"idx": 2, "title": "Do Not Trust What They Tell : Exposing Malicious... | OpenReview", "date": "", "ddg_snippet": "In Section 5.4, clustering analysis reveals potential Family relationships among anomalous nodes, but the discussion of these results is limited. Could the authors expand on how these relationships could be validated or associated with specific entities or organizations...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=qcnePVejeV", "content": "In Section 5.4, clustering analysis reveals potential Family relationships among anomalous nodes, but the discussion of these results is limited. Could the authors expand on how these relationships could be validated or associated with specific entities or organizations..."}
+{"idx": 3, "title": "Clustering Visualizer is a Web Application for visualizing of Machine...", "date": "", "ddg_snippet": "Clustering Visualizer is a Web Application for visualizing of Machine Learning Clustering Algorithms.", "subpage_snippet": "", "source": "clustering-visualizer.web.app", "link": "https://clustering-visualizer.web.app/", "content": "Clustering Visualizer is a Web Application for visualizing of Machine Learning Clustering Algorithms."}
+{"idx": 4, "title": "Turbo AI - AI Note Taker", "date": "", "ddg_snippet": "Join 5 million students using Turbo AI for beautiful notes, flashcards, quizzes, and more.", "subpage_snippet": "", "source": "www.turbo.ai", "link": "https://www.turbo.ai/", "content": "Join 5 million students using Turbo AI for beautiful notes, flashcards, quizzes, and more."}
+{"idx": 5, "title": "AI Detector - Trusted AI Checker (Free & No SignUp)", "date": "", "ddg_snippet": "Free AI Checker not only supports the generation of detection reports in multiple formats to meet your needs in different scenarios, but also can give you the most authoritative content analysis results. AI Detector that highlight text allows you to tell at a glance which content is AI or not.", "subpage_snippet": "", "source": "decopy.ai", "link": "https://decopy.ai/ai-detector/", "content": "Free AI Checker not only supports the generation of detection reports in multiple formats to meet your needs in different scenarios, but also can give you the most authoritative content analysis results. AI Detector that highlight text allows you to tell at a glance which content is AI or not."}
+{"idx": 6, "title": "Uncovering multi-technology convergence patterns with hypergraphs...", "date": "", "ddg_snippet": "4 . 3 .2. Intrinsic features Intrinsic features capture properties of the individual technologies that may correlate with convergence propensity, and are aggregated to the hyperedge level.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.15521", "content": "4 . 3 .2. Intrinsic features Intrinsic features capture properties of the individual technologies that may correlate with convergence propensity, and are aggregated to the hyperedge level."}
+{"idx": 7, "title": "Silent Hill f All Puzzle Solutions Guide", "date": "", "ddg_snippet": "Stairwell Puzzle. Story Difficulty – In order to open the locked drawer in the teachers’ storage room, you must first go into the one open classroom and pick up the scrunched-up paper. This will tell you that you can open it with a hairpin.", "subpage_snippet": "", "source": "www.powerpyx.com", "link": "https://www.powerpyx.com/silent-hill-f-all-puzzle-solutions-guide/", "content": "Stairwell Puzzle. Story Difficulty – In order to open the locked drawer in the teachers’ storage room, you must first go into the one open classroom and pick up the scrunched-up paper. This will tell you that you can open it with a hairpin."}
+{"idx": 8, "title": "Present Simple — правила и примеры употребления.", "date": "", "ddg_snippet": "формула отрицательного предложения. Примеры отрицательных предложений в Present Simple. I do not play football. — Я не играю в футбол. Mary does not play piano. — Мэри не играет на пианино. Students don’t sit in the classroom.", "subpage_snippet": "", "source": "skysmart.ru", "link": "https://skysmart.ru/articles/english/present-simple-tense", "content": "формула отрицательного предложения. Примеры отрицательных предложений в Present Simple. I do not play football. — Я не играю в футбол. Mary does not play piano. — Мэри не играет на пианино. Students don’t sit in the classroom."}
+{"idx": 9, "title": "Полный гайд по автотестам для лидов и разработчиков. / Хабр", "date": "", "ddg_snippet": "This pattern is designed to make it easier for large teams to work simultaneously on features while keeping mainline in a releasable state. They can also be used in other patterns that do not allow for continuous integration, but may still be effective patterns for delivering software.", "subpage_snippet": "", "source": "habr.com", "link": "https://habr.com/ru/articles/949148/", "content": "This pattern is designed to make it easier for large teams to work simultaneously on features while keeping mainline in a releasable state. They can also be used in other patterns that do not allow for continuous integration, but may still be effective patterns for delivering software."}
diff --git a/data/sampled_jsons/reinforcement_learning_trajectory_replanning_robotics_2024_year_2024.jsonl b/data/sampled_jsons/reinforcement_learning_trajectory_replanning_robotics_2024_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..627dc7b2470d8670d7d49bd2f5e7186187e12f8f
--- /dev/null
+++ b/data/sampled_jsons/reinforcement_learning_trajectory_replanning_robotics_2024_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "inverse- reinforcement - learning · GitHub Topics · GitHub", "date": "", "ddg_snippet": "machine- learning reinforcement - learning tensorflow ml rl lfd imitation- learning inverse- reinforcement - learning learning -from-demonstration imitation irl. Updated May 10, 2024 . Python.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/topics/inverse-reinforcement-learning", "content": "machine- learning reinforcement - learning tensorflow ml rl lfd imitation- learning inverse- reinforcement - learning learning -from-demonstration imitation irl. Updated May 10, 2024 . Python."}
+{"idx": 1, "title": "(PDF) A Hybrid Human-in-the-Loop Deep Reinforcement Learning ...", "date": "", "ddg_snippet": "TL;DR: Deep reinforcement learning (DRL) is poised to revolutionize the field of artificial intelligence (AI) and represents a step toward building autonomous systems with a higher-level understanding of the visual world as discussed by the authors.", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/papers/a-hybrid-human-in-the-loop-deep-reinforcement-learning-j1em2nr7", "content": "TL;DR: Deep reinforcement learning (DRL) is poised to revolutionize the field of artificial intelligence (AI) and represents a step toward building autonomous systems with a higher-level understanding of the visual world as discussed by the authors."}
+{"idx": 2, "title": "Reinforcement Learning Allows Underwater Robots To... - Astrobiology", "date": "", "ddg_snippet": "Specifically, this new work reveals that reinforcement learning , widely used in the field of control and robotics , as well as in the development of tools related to natural language processing such as ChatGPT, allows underwater robots to learn what actions to perform at any given time to...", "subpage_snippet": "", "source": "astrobiology.com", "link": "https://astrobiology.com/2023/07/reinforcement-learning-allows-underwater-robots-to-locate-and-track-objects-underwater.html", "content": "Specifically, this new work reveals that reinforcement learning , widely used in the field of control and robotics , as well as in the development of tools related to natural language processing such as ChatGPT, allows underwater robots to learn what actions to perform at any given time to..."}
+{"idx": 3, "title": "Beyond Human Preferences: Exploring Reinforcement Learning ...", "date": "", "ddg_snippet": "Jul 1, 2024 . Can LLMs Help AI Learn Faster? A New Approach to Reinforcement Learning . Beyond Human Preferences: Exploring Reinforcement Learning Trajectory Evaluation and Improvement through LLMs.", "subpage_snippet": "", "source": "www.promptlayer.com", "link": "https://www.promptlayer.com/research-papers/beyond-human-preferences-exploring-reinforcement-learning-trajectory-evaluation-and-improvement-through-llms", "content": "Jul 1, 2024 . Can LLMs Help AI Learn Faster? A New Approach to Reinforcement Learning . Beyond Human Preferences: Exploring Reinforcement Learning Trajectory Evaluation and Improvement through LLMs."}
+{"idx": 4, "title": "Multi Agent Reinforcement Learning Trajectory ... | IEEE DataPort", "date": "", "ddg_snippet": "Mohammad Reza Maleki, Mohammad Robat Mili, Mohammad Reza Javan, Nader Mokari, Eduard A. Jorswieck. (2021). Multi Agent Reinforcement Learning Trajectory Design and Two-Stage Resource Management in CoMP UAV VLC Networks.", "subpage_snippet": "", "source": "ieee-dataport.org", "link": "https://ieee-dataport.org/documents/multi-agent-reinforcement-learning-trajectory-design-and-two-stage-resource-management", "content": "Mohammad Reza Maleki, Mohammad Robat Mili, Mohammad Reza Javan, Nader Mokari, Eduard A. Jorswieck. (2021). Multi Agent Reinforcement Learning Trajectory Design and Two-Stage Resource Management in CoMP UAV VLC Networks."}
+{"idx": 5, "title": "RoboNav: Robot Path Planning Using Reinforcement Learning and...", "date": "", "ddg_snippet": "Rl4co: an extensive reinforcement learning for combinatorial optimization benchmark, 2024 . [3] Pengzhan Chen, Jiean Pei, Weiqing Lu, and Mingzhen Li. A deep reinforcement learning based method for real-time path planning and dynamic obstacle avoidance.", "subpage_snippet": "", "source": "publications.waset.org", "link": "https://publications.waset.org/10014231/robonav-robot-path-planning-using-reinforcement-learning-and-attention-models", "content": "Rl4co: an extensive reinforcement learning for combinatorial optimization benchmark, 2024 . [3] Pengzhan Chen, Jiean Pei, Weiqing Lu, and Mingzhen Li. A deep reinforcement learning based method for real-time path planning and dynamic obstacle avoidance."}
+{"idx": 6, "title": "ARtonomous: Introducing Middle School Students to Reinforcement ...", "date": "", "ddg_snippet": "With ARtonomous, students employ reinforcement learning (RL) alongside code to train and customize virtual autonomous robotic vehicles. Through a study evaluating ARtonomous, we found that middle-school students developed an understanding of RL...", "subpage_snippet": "", "source": "pr-mlr-shield-prod.apple.com", "link": "https://pr-mlr-shield-prod.apple.com/research/reinforcement-learning-robotics", "content": "With ARtonomous, students employ reinforcement learning (RL) alongside code to train and customize virtual autonomous robotic vehicles. Through a study evaluating ARtonomous, we found that middle-school students developed an understanding of RL..."}
+{"idx": 7, "title": "Robotic Adaptability through Deep Reinforcement Learning (DRL)", "date": "", "ddg_snippet": "Through the power of deep reinforcement learning (DRL), the robot seamlessly transitions between walking, trotting, and pronking, echoing the fluidity and grace of natural animal movements.", "subpage_snippet": "", "source": "onpassive.financialfreedom-llc.com", "link": "https://onpassive.financialfreedom-llc.com/2024/robotic-adaptability-through-deep-reinforcement-learning-drl/", "content": "Through the power of deep reinforcement learning (DRL), the robot seamlessly transitions between walking, trotting, and pronking, echoing the fluidity and grace of natural animal movements."}
+{"idx": 8, "title": "Computer Vision Group - Robot Vision - Trajectory replanning", "date": "", "ddg_snippet": "Trajectory replanning Contact: Vladyslav Usenko. In this paper, we present a real-time approach to local trajectory replanning for microaerial vehicles (MAVs).", "subpage_snippet": "", "source": "cvg.cit.tum.de", "link": "https://cvg.cit.tum.de/research/robotvision/replanning", "content": "Trajectory replanning Contact: Vladyslav Usenko. In this paper, we present a real-time approach to local trajectory replanning for microaerial vehicles (MAVs)."}
+{"idx": 9, "title": "Deep Local Trajectory Replanning and Control for Robot Navigation", "date": "", "ddg_snippet": "Deep Reinforcement Learning based Robot Navigation in Dynamic Environments using Occupancy Values of Motion Primitives.", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/103435825/Deep_Local_Trajectory_Replanning_and_Control_for_Robot_Navigation", "content": "Deep Reinforcement Learning based Robot Navigation in Dynamic Environments using Occupancy Values of Motion Primitives."}
diff --git a/data/sampled_jsons/reward_estimation_diffusion_models_practical_challenges_computational_cost_non-differentiable.jsonl b/data/sampled_jsons/reward_estimation_diffusion_models_practical_challenges_computational_cost_non-differentiable.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c1075a2d20cbd9079738fa681768d056b4ab7c17
--- /dev/null
+++ b/data/sampled_jsons/reward_estimation_diffusion_models_practical_challenges_computational_cost_non-differentiable.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Stable Diffusion - Wikipedia", "date": "", "ddg_snippet": "Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Stable_Diffusion", "content": "Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial..."}
+{"idx": 1, "title": "Inference-Time Text-to-Video Alignment with Diffusion Latent Beam...", "date": "", "ddg_snippet": "Directly fine-tuning diffusion models on differentiable rewards .This section describes the reward calibration procedure and results for two challenging prompt sets, DEVIL-very-high and MovieGen, which were introduced to evaluate our method with larger T2V models .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2501.19252v2", "content": "Directly fine-tuning diffusion models on differentiable rewards .This section describes the reward calibration procedure and results for two challenging prompt sets, DEVIL-very-high and MovieGen, which were introduced to evaluate our method with larger T2V models ."}
+{"idx": 2, "title": "Simulation of conditioned diffusion and application to parameter...", "date": "", "ddg_snippet": "The implied computational cost impacts applications in both geometric statistics and in geometric deep learning. The weighted diffusion mean offers an alternative to the weighted Fréchet mean.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/227423566_Simulation_of_conditioned_diffusion_and_application_to_parameter_estimation", "content": "The implied computational cost impacts applications in both geometric statistics and in geometric deep learning. The weighted diffusion mean offers an alternative to the weighted Fréchet mean."}
+{"idx": 3, "title": "Positivity and global existence for nonlocal advection- diffusion models ...", "date": "", "ddg_snippet": "We prove the global existence of both non -negative weak solutions in any spatial dimension and positive classical solutions in one spatial dimension. These results generalise and unify various existing results regarding existence of nonlocal advection- diffusion equations.", "subpage_snippet": "", "source": "www.aimspress.com", "link": "https://www.aimspress.com/article/doi/10.3934/math.2025949?viewType=HTML", "content": "We prove the global existence of both non -negative weak solutions in any spatial dimension and positive classical solutions in one spatial dimension. These results generalise and unify various existing results regarding existence of nonlocal advection- diffusion equations."}
+{"idx": 4, "title": "Automated Efficient Estimation using Monte Carlo Efficient Influence...", "date": "", "ddg_snippet": "Extending MC-EIF to handle non - differentiable functionals is crucial, as many real-world problems involve such scenarios. This might involve exploring techniques like smoothing or using alternative approximation methods.", "subpage_snippet": "", "source": "deep-diver.github.io", "link": "https://deep-diver.github.io/neurips2024/spotlight-ai-theory/2wfd3pti8v/", "content": "Extending MC-EIF to handle non - differentiable functionals is crucial, as many real-world problems involve such scenarios. This might involve exploring techniques like smoothing or using alternative approximation methods."}
+{"idx": 5, "title": "Directly Aligning the Full Diffusion Trajectory with Fine-Grained...", "date": "", "ddg_snippet": "Recent studies have demonstrated the effectiveness of directly aligningdiffusion models with human preferences using differentiable reward .", "subpage_snippet": "", "source": "deeplearn.org", "link": "https://deeplearn.org/arxiv/635326/directly-aligning-the-full-diffusion-trajectory-with-fine-grained-human-preference", "content": "Recent studies have demonstrated the effectiveness of directly aligningdiffusion models with human preferences using differentiable reward ."}
+{"idx": 6, "title": "Free Acceleration for Diffusion ! TGATE: Cross-Attention... - StatedAI", "date": "", "ddg_snippet": "2. Contributions. The research team designed a simple, effective, and training-free method called Time-Gated Cross-Attention (TGATE) to improve model inference efficiency while maintaining the generation quality of existing diffusion models .", "subpage_snippet": "", "source": "statedai.com", "link": "https://statedai.com/free-acceleration-for-diffusion-tgate-cross-attention-not-always-effective-in-denoising.html", "content": "2. Contributions. The research team designed a simple, effective, and training-free method called Time-Gated Cross-Attention (TGATE) to improve model inference efficiency while maintaining the generation quality of existing diffusion models ."}
+{"idx": 7, "title": "Examples Of Fine-Tuning Vision Models | Restackio", "date": "", "ddg_snippet": "Computational Cost : Implementing attention can significantly increase the computational load, requiring more resources and time for training. Data Requirements: High-quality attention models often necessitate large datasets to learn effectively, which may not always be available.", "subpage_snippet": "", "source": "d2wozrt205r2fu.cloudfront.net", "link": "https://d2wozrt205r2fu.cloudfront.net/p/vision-fine-tuning-answer-examples-of-fine-tuning-vision-models-cat-ai", "content": "Computational Cost : Implementing attention can significantly increase the computational load, requiring more resources and time for training. Data Requirements: High-quality attention models often necessitate large datasets to learn effectively, which may not always be available."}
+{"idx": 8, "title": "[Revisión de artículo] Dynamic Search for Inference-Time Alignment in...", "date": "", "ddg_snippet": "...Inference-Time Alignment in Diffusion Models \" presents an innovative methodology aimed at improving the alignment of outputs generated by diffusion models with desired reward functions, particularly in cases where these rewards are non - differentiable .", "subpage_snippet": "", "source": "www.themoonlight.io", "link": "https://www.themoonlight.io/es/review/dynamic-search-for-inference-time-alignment-in-diffusion-models", "content": "...Inference-Time Alignment in Diffusion Models \" presents an innovative methodology aimed at improving the alignment of outputs generated by diffusion models with desired reward functions, particularly in cases where these rewards are non - differentiable ."}
+{"idx": 9, "title": "Track: Poster Session 1", "date": "", "ddg_snippet": "Chemistry-Inspired Diffusion with Non - Differentiable Guidance.RAG-DDR: Optimizing Retrieval-Augmented Generation Using Differentiable Data Rewards .", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2025/session/31971", "content": "Chemistry-Inspired Diffusion with Non - Differentiable Guidance.RAG-DDR: Optimizing Retrieval-Augmented Generation Using Differentiable Data Rewards ."}
diff --git a/data/sampled_jsons/risk_aversion_parameter_alpha_quantile_regret_optimization_reinforcement_learning.jsonl b/data/sampled_jsons/risk_aversion_parameter_alpha_quantile_regret_optimization_reinforcement_learning.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..30a315ba230ab7d4fcd148cdd5c5254d246e17cc
--- /dev/null
+++ b/data/sampled_jsons/risk_aversion_parameter_alpha_quantile_regret_optimization_reinforcement_learning.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Risk-Averse Total-Reward Reinforcement Learning", "date": "", "ddg_snippet": "by X Su · 2025 — Risk-averse total-reward Markov Decision Processes (MDPs ) offer a promising framework for modeling and solving undiscounted infinite-horizon ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2506.21683?", "content": "by X Su · 2025 — Risk-averse total-reward Markov Decision Processes (MDPs ) offer a promising framework for modeling and solving undiscounted infinite-horizon ..."}
+{"idx": 1, "title": "Efficient Risk-Averse Reinforcement Learning", "date": "", "ddg_snippet": "by I Greenberg · Cited by 68 — In risk-averse reinforcement learning (RL), the goal is to optimize some risk mea- sure of the returns . A risk measure often focuses on the worst returns ... 33 pages", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2022/file/d2511dfb731fa336739782ba825cd98c-Supplemental-Conference.pdf", "content": "by I Greenberg · Cited by 68 — In risk-averse reinforcement learning (RL), the goal is to optimize some risk mea- sure of the returns . A risk measure often focuses on the worst returns ... 33 pages"}
+{"idx": 2, "title": "Efficient Risk-Averse Reinforcement Learning", "date": "", "ddg_snippet": "by I Greenberg · 2022 · Cited by 67 — In risk-averse reinforcement learning (RL), the goal is to optimize some risk mea- sure of the returns . A risk measure often focuses on the worst returns ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=LdAxczs3m0", "content": "by I Greenberg · 2022 · Cited by 67 — In risk-averse reinforcement learning (RL), the goal is to optimize some risk mea- sure of the returns . A risk measure often focuses on the worst returns ..."}
+{"idx": 3, "title": "Online Bayesian Risk-Averse Reinforcement Learning", "date": "", "ddg_snippet": "5 days ago — In this paper, we study the Bayesian risk-averse formulation in reinforcement learning (RL). To address the epistemic uncertainty due to a lack ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.14077v1", "content": "5 days ago — In this paper, we study the Bayesian risk-averse formulation in reinforcement learning (RL). To address the epistemic uncertainty due to a lack ..."}
+{"idx": 4, "title": "Burning RED: Unlocking Subtask-Driven Reinforcement ...", "date": "", "ddg_snippet": "by JS Rojas · Cited by 1 — This paper introduces a Reward Extended Differential (RED) approach for risk - averse AMDP that aims to handle multiple subtasks concurrently, by ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=5y3QbuK6HD", "content": "by JS Rojas · Cited by 1 — This paper introduces a Reward Extended Differential (RED) approach for risk - averse AMDP that aims to handle multiple subtasks concurrently, by ..."}
+{"idx": 5, "title": "RISK-AVERSE OFFLINE REINFORCEMENT LEARNING", "date": "", "ddg_snippet": "Training Reinforcement Learning (RL) agents online in high-stakes applications is often prohibitive due to the risk associated with exploration.", "subpage_snippet": "", "source": "offline-rl-neurips.github.io", "link": "https://offline-rl-neurips.github.io/supplement/35supp.pdf", "content": "Training Reinforcement Learning (RL) agents online in high-stakes applications is often prohibitive due to the risk associated with exploration."}
+{"idx": 6, "title": "Mean-Semivariance Policy Optimization via Risk-Averse ...", "date": "", "ddg_snippet": "by X Ma · 2022 · Cited by 7 — This paper aims at optimizing the mean-semivariance (MSV) criterion in reinforcement learning w.r.t. steady reward distribution. Since semivariance is time- ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/pdf/10.1613/jair.1.13833", "content": "by X Ma · 2022 · Cited by 7 — This paper aims at optimizing the mean-semivariance (MSV) criterion in reinforcement learning w.r.t. steady reward distribution. Since semivariance is time- ..."}
+{"idx": 7, "title": "RA-PbRL: Provably Efficient Risk-Aware Preference-Based ...", "date": "", "ddg_snippet": "9 Dec 2024 — On the other hand, the static risk objective regret is mainly determined by the quantile function . We have also identified the following four ...", "subpage_snippet": "", "source": "nips.cc", "link": "https://nips.cc/virtual/2024/poster/95716", "content": "9 Dec 2024 — On the other hand, the static risk objective regret is mainly determined by the quantile function . We have also identified the following four ..."}
+{"idx": 8, "title": "RISK-AWARE REINFORCEMENT LEARNING WITH ...", "date": "", "ddg_snippet": "by T Lam · Cited by 21 — Our work provides a unified framework to analyze the regret of risk -aware RL policy with coherent risk measures in conjunction with non-linear function ...", "subpage_snippet": "", "source": "web.mit.edu", "link": "https://web.mit.edu/jaillet/www/general/iclr23b.pdf", "content": "by T Lam · Cited by 21 — Our work provides a unified framework to analyze the regret of risk -aware RL policy with coherent risk measures in conjunction with non-linear function ..."}
+{"idx": 9, "title": "Regret Bounds for Risk-Sensitive Reinforcement Learning", "date": "", "ddg_snippet": "by O Bastani · 2022 · Cited by 26 — For CVaR objectives, we have LG = 1/ α , where α is the size of the tail considered—e.g., when α is small, it averages over outliers with particularly small ... 11 pages", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2022/file/eb4898d622e9a48b5f9713ea1fcff2bf-Paper-Conference.pdf", "content": "by O Bastani · 2022 · Cited by 26 — For CVaR objectives, we have LG = 1/ α , where α is the size of the tail considered—e.g., when α is small, it averages over outliers with particularly small ... 11 pages"}
diff --git a/data/sampled_jsons/scaling_factor_min_max_epoch_threshold_linear_decay_OR_exponential_decay.jsonl b/data/sampled_jsons/scaling_factor_min_max_epoch_threshold_linear_decay_OR_exponential_decay.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..003e786157fbb4b4b4ecc9373fb39eeef395e07c
--- /dev/null
+++ b/data/sampled_jsons/scaling_factor_min_max_epoch_threshold_linear_decay_OR_exponential_decay.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Exponential smoothing - Wikipedia", "date": "", "ddg_snippet": "Exponential smoothing or exponential moving average (EMA) is a rule of thumb technique for smoothing time series data using the exponential window function. Whereas in the simple moving average the past observations are weighted equally, exponential functions are used to assign exponentially decreasing weights over time. It is an easily learned and easily applied procedure for making some ...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Exponential_smoothing", "content": "Exponential smoothing or exponential moving average (EMA) is a rule of thumb technique for smoothing time series data using the exponential window function. Whereas in the simple moving average the past observations are weighted equally, exponential functions are used to assign exponentially decreasing weights over time. It is an easily learned and easily applied procedure for making some ..."}
+{"idx": 1, "title": "tf.keras.optimizers.schedules.ExponentialDecay | TensorFlow ...", "date": "", "ddg_snippet": "def decayed_learning_rate(step): return initial_learning_rate * decay_rate ^ (step / decay_steps) If the argument staircase is True, then step / decay_steps is an integer division and the decayed learning rate follows a staircase function. You can pass this schedule directly into a keras.optimizers.Optimizer as the learning rate. Example: When fitting a Keras model, decay every 100000 steps ...", "subpage_snippet": "", "source": "www.tensorflow.org", "link": "https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/schedules/ExponentialDecay", "content": "def decayed_learning_rate(step): return initial_learning_rate * decay_rate ^ (step / decay_steps) If the argument staircase is True, then step / decay_steps is an integer division and the decayed learning rate follows a staircase function. You can pass this schedule directly into a keras.optimizers.Optimizer as the learning rate. Example: When fitting a Keras model, decay every 100000 steps ..."}
+{"idx": 2, "title": "Pytorch Change the learning rate based on number of epochs Code sample", "date": "", "ddg_snippet": "Feb 4, 2020 · Epoch -20 lr: 0.00010000000000000003 More on How to adjust Learning Rate - torch.optim.lr_scheduler provides several methods to adjust the learning rate based on the number of epochs. See more on stackoverflow", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/60050586/pytorch-change-the-learning-rate-based-on-number-of-epochs", "content": "Feb 4, 2020 · Epoch -20 lr: 0.00010000000000000003 More on How to adjust Learning Rate - torch.optim.lr_scheduler provides several methods to adjust the learning rate based on the number of epochs. See more on stackoverflow"}
+{"idx": 3, "title": "How is the decay rate in exponential smoothing optimized?", "date": "", "ddg_snippet": "Jan 19, 2022 · For the sake of simplicity, I just want to focus on single/level exponential smoothing. When alpha, the decay rate, is near 1, the most recent observation has the highest weight and influence of re...", "subpage_snippet": "", "source": "stats.stackexchange.com", "link": "https://stats.stackexchange.com/questions/561124/how-is-the-decay-rate-in-exponential-smoothing-optimized", "content": "Jan 19, 2022 · For the sake of simplicity, I just want to focus on single/level exponential smoothing. When alpha, the decay rate, is near 1, the most recent observation has the highest weight and influence of re..."}
+{"idx": 4, "title": "GraphPad Prism 10 Curve Fitting Guide - Equation: Two phase decay Exponential Decay & Other LR Schedules - apxml.com Exponential smoothing - Wikipedia", "date": "", "ddg_snippet": "An exponential decay equation models many chemical and biological processes. It is used whenever the rate at which something happens is proportional to the amount which is left. A two-phase model is used when the outcome you measure is the result of the sum of a fast and slow exponential decay . This is also called a double exponential decay . See full list on graphpad.com Create an XY data table. Enter time into X, and response (binding, concentration ..) into Y. If you have several experimental conditions, place the first into column A, the second into column B, etc. After entering data, click Analyze, choose nonlinear regression, choose the panel of exponential equations, and choose Two phasedecay. See full list on graphpad.com If you have subtracted off any background signal, then you know the curve has to plateau at Y=0. In this case, you should constrain the parameter Plateau to be a constant value equal to zero. To do this, go to the Constrain tab of the nonlinear regression dialog, set the drop down next to Plateau to \"Constant equal to\" and enter the value 0.0. In s... See full list on graphpad.com SpanFast=(Y0-Plateau)*PercentFast*.01 SpanSlow=(Y0-Plateau)*(100-PercentFast)*.01 Y=Plateau + SpanFast*exp(-KFast*X) + SpanSlow*exp(-KSlow*X) Y0is the Y value when X (time) is zero. It is expressed in the same units as Y, Plateauis the Y value at infinite times, expressed in the same units as Y. Kfast andKsloware the two rate constants, expressed i... See full list on graphpad.com Comparison of Step Decay (blue), Exponential Decay (orange), and one cycle of Cosine Annealing (green) over 50 epochs, starting from an initial learning rate of 0.1. Note the smooth decrease of Exponential and Cosine compared to the sharp drops of Step Decay. The Cosine curve returns to the max value if restarts are used (shown partially after epoch 25 for illustration). Implementing Learning ... Exponential smoothing or exponential moving average (EMA) is a rule of thumb technique for smoothing time series data using the exponential window function. Whereas in the simple moving average the past observations are weighted equally, exponential functions are used to assign exponentially decreasing weights over time. It is an easily learned and easily applied procedure for making some ...", "subpage_snippet": "", "source": "www.graphpad.com", "link": "https://www.graphpad.com/guides/prism/latest/curve-fitting/reg_exponential_decay_2phase.htm", "content": "An exponential decay equation models many chemical and biological processes. It is used whenever the rate at which something happens is proportional to the amount which is left. A two-phase model is used when the outcome you measure is the result of the sum of a fast and slow exponential decay . This is also called a double exponential decay . See full list on graphpad.com Create an XY data table. Enter time into X, and response (binding, concentration ..) into Y. If you have several experimental conditions, place the first into column A, the second into column B, etc. After entering data, click Analyze, choose nonlinear regression, choose the panel of exponential equations, and choose Two phasedecay. See full list on graphpad.com If you have subtracted off any background signal, then you know the curve has to plateau at Y=0. In this case, you should constrain the parameter Plateau to be a constant value equal to zero. To do this, go to the Constrain tab of the nonlinear regression dialog, set the drop down next to Plateau to \"Constant equal to\" and enter the value 0.0. In s... See full list on graphpad.com SpanFast=(Y0-Plateau)*PercentFast*.01 SpanSlow=(Y0-Plateau)*(100-PercentFast)*.01 Y=Plateau + SpanFast*exp(-KFast*X) + SpanSlow*exp(-KSlow*X) Y0is the Y value when X (time) is zero. It is expressed in the same units as Y, Plateauis the Y value at infinite times, expressed in the same units as Y. Kfast andKsloware the two rate constants, expressed i... See full list on graphpad.com Comparison of Step Decay (blue), Exponential Decay (orange), and one cycle of Cosine Annealing (green) over 50 epochs, starting from an initial learning rate of 0.1. Note the smooth decrease of Exponential and Cosine compared to the sharp drops of Step Decay. The Cosine curve returns to the max value if restarts are used (shown partially after epoch 25 for illustration). Implementing Learning ... Exponential smoothing or exponential moving average (EMA) is a rule of thumb technique for smoothing time series data using the exponential window function. Whereas in the simple moving average the past observations are weighted equally, exponential functions are used to assign exponentially decreasing weights over time. It is an easily learned and easily applied procedure for making some ..."}
+{"idx": 5, "title": "Exponential Decay & Other LR Schedules - apxml.com", "date": "", "ddg_snippet": "Comparison of Step Decay (blue), Exponential Decay (orange), and one cycle of Cosine Annealing (green) over 50 epochs, starting from an initial learning rate of 0.1. Note the smooth decrease of Exponential and Cosine compared to the sharp drops of Step Decay. The Cosine curve returns to the max value if restarts are used (shown partially after epoch 25 for illustration). Implementing Learning ...", "subpage_snippet": "", "source": "apxml.com", "link": "https://apxml.com/courses/deep-learning-regularization-optimization/chapter-7-optimization-refinements-tuning/other-lr-schedules", "content": "Comparison of Step Decay (blue), Exponential Decay (orange), and one cycle of Cosine Annealing (green) over 50 epochs, starting from an initial learning rate of 0.1. Note the smooth decrease of Exponential and Cosine compared to the sharp drops of Step Decay. The Cosine curve returns to the max value if restarts are used (shown partially after epoch 25 for illustration). Implementing Learning ..."}
+{"idx": 6, "title": "STAT 224 Lecture 13 Chapter 6 Transformation of Variables", "date": "", "ddg_snippet": "Example: Bacteria Deaths Due to X-Ray Radiation According to theory, we expect a exponential decay in the count of bacteria in time: nt n0eβ1t = , n0 initial population size where = β1 decay rate =", "subpage_snippet": "", "source": "www.stat.uchicago.edu", "link": "https://www.stat.uchicago.edu/~yibi/teaching/stat224/L13.pdf", "content": "Example: Bacteria Deaths Due to X-Ray Radiation According to theory, we expect a exponential decay in the count of bacteria in time: nt n0eβ1t = , n0 initial population size where = β1 decay rate ="}
+{"idx": 7, "title": "Dynamic Rank Adjustment in Diffusion Policies for Efficient ...", "date": "", "ddg_snippet": "26 Apr 2025 — ... epoch if the number of trainable ranks has changed since the previous epoch . ... r cosine = ⌊ r min + 0.5 × ( r max − r min ) × ( 1 + cos ( π × ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.03822v3", "content": "26 Apr 2025 — ... epoch if the number of trainable ranks has changed since the previous epoch . ... r cosine = ⌊ r min + 0.5 × ( r max − r min ) × ( 1 + cos ( π × ..."}
+{"idx": 8, "title": "Improving the Accuracy of Progress Indication for ... - cs.wisc.edu", "date": "", "ddg_snippet": "exponential decay works better than linear decay . ... min ( 4[c0(1 + P1) – cv] / C , g – 1) if c0(1 + ... epoch . In the test, early stopping happened on ...", "subpage_snippet": "", "source": "pages.cs.wisc.edu", "link": "https://pages.cs.wisc.edu/~gangluo/deep_learning_PI2_full_version.pdf", "content": "exponential decay works better than linear decay . ... min ( 4[c0(1 + P1) – cv] / C , g – 1) if c0(1 + ... epoch . In the test, early stopping happened on ..."}
+{"idx": 9, "title": "District heating optimization in residential buildings using ...", "date": "", "ddg_snippet": "by SSV Kalidindi · 2025 — ... min measurement intervals. The buildings from both Sweden and ... max ( T F i ) denote the minimum and maximum values observed for that ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S2666546825001351", "content": "by SSV Kalidindi · 2025 — ... min measurement intervals. The buildings from both Sweden and ... max ( T F i ) denote the minimum and maximum values observed for that ..."}
diff --git a/data/sampled_jsons/sitearxiv.org_2405.14728_Example_3.2_abd(1-e)_+_(1-a)cd(1-e).jsonl b/data/sampled_jsons/sitearxiv.org_2405.14728_Example_3.2_abd(1-e)_+_(1-a)cd(1-e).jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bcf346d4b8eccd23c7c8657e9f84b270c00f62f3
--- /dev/null
+++ b/data/sampled_jsons/sitearxiv.org_2405.14728_Example_3.2_abd(1-e)_+_(1-a)cd(1-e).jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "The Space-Time Complexity of Sum-Product Queries - arXiv.org", "date": "", "ddg_snippet": "For example , the cache ME for variable E has key con( E ) = ABD , and its space exponent is ρ∗( ABD ) = 1 .5. When solve( E , abcd) is called, Algorithm 3 stores the result in ME( abd ); in later calls, if c changed while abd are the same, the algorithm immediately returns the cached value.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.11920", "content": "For example , the cache ME for variable E has key con( E ) = ABD , and its space exponent is ρ∗( ABD ) = 1 .5. When solve( E , abcd) is called, Algorithm 3 stores the result in ME( abd ); in later calls, if c changed while abd are the same, the algorithm immediately returns the cached value."}
+{"idx": 1, "title": "HOW TO BUILD A HOPF ALGEBRA - arXiv.org", "date": "", "ddg_snippet": "The unital algebra and counital coalgebra structures on the composite ( 1 ) come from recognizing it as either a horizontal composition of an adjoint pair of 2-morphisms in C or as a vertical composition of a different adjoint pair of 2-morphisms, see Remark 4.9. Example 1 . 1 .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2508.16787v2", "content": "The unital algebra and counital coalgebra structures on the composite ( 1 ) come from recognizing it as either a horizontal composition of an adjoint pair of 2-morphisms in C or as a vertical composition of a different adjoint pair of 2-morphisms, see Remark 4.9. Example 1 . 1 ."}
+{"idx": 2, "title": "[ 2405 . 14728 ] Intervention and Conditioning in Causal Bayesian Networks", "date": "", "ddg_snippet": "In case of Causal Bayesian Networks (CBNs), Pearl assumes autonomy of mechanisms that determine interventions to calculate a range of probabilities.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2405.14728", "content": "In case of Causal Bayesian Networks (CBNs), Pearl assumes autonomy of mechanisms that determine interventions to calculate a range of probabilities."}
+{"idx": 3, "title": "Monitoring Machine Learning Systems: A Multivocal Literature ...", "date": "", "ddg_snippet": "4 days ago · These guidelines can be organized under two broad categories: ( 1 ) general monitoring recommendations that apply to all ML systems, e .g., track data quality, integrate monitoring in CI/ CD pipelines, and ensure operational reliability; and (2) context dependent monitoring recommendations that can vary based on various factors of the ML system, e ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.14294v1", "content": "4 days ago · These guidelines can be organized under two broad categories: ( 1 ) general monitoring recommendations that apply to all ML systems, e .g., track data quality, integrate monitoring in CI/ CD pipelines, and ensure operational reliability; and (2) context dependent monitoring recommendations that can vary based on various factors of the ML system, e ..."}
+{"idx": 4, "title": "Intervention and Conditioning in Causal Bayesian Networks", "date": "", "ddg_snippet": "These examples underscore the versatility and utility of causal models for providing a formal representation of system variables.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2405.14728v1", "content": "These examples underscore the versatility and utility of causal models for providing a formal representation of system variables."}
+{"idx": 5, "title": "Intervention and Conditioning in Causal Bayesian", "date": "", "ddg_snippet": "variable X, the value of U is the same as its value in the conjunct determining the value of U (the examples should make clear what this means). Example 3 . 1 . Consider the CBN M ∗ with the following dag: U.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2405.14728", "content": "variable X, the value of U is the same as its value in the conjunct determining the value of U (the examples should make clear what this means). Example 3 . 1 . Consider the CBN M ∗ with the following dag: U."}
+{"idx": 6, "title": "AgentMaster: A Multi-Agent Conversational Framework Using A2A and...", "date": "", "ddg_snippet": "Figure 3: AgentMaster demonstration example and verification.2025. Agentdistill: Training-free agent distillation with generalizable mcp boxes. Preprint, arXiv:2506.14728. Sarkar and Sarkar (2025). ↑.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.21105v2", "content": "Figure 3: AgentMaster demonstration example and verification.2025. Agentdistill: Training-free agent distillation with generalizable mcp boxes. Preprint, arXiv:2506.14728. Sarkar and Sarkar (2025). ↑."}
+{"idx": 7, "title": "LightRetriever: A LLM-based Text Retrieval Architecture with Extremely...", "date": "", "ddg_snippet": "For example , Full-Llama8b and Full-Qwen7b require over 100 seconds solely for query encoding (109.5s and 100.7s, respectively), making them a bottleneck for large-scale, real-time retrieval.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.12260v4", "content": "For example , Full-Llama8b and Full-Qwen7b require over 100 seconds solely for query encoding (109.5s and 100.7s, respectively), making them a bottleneck for large-scale, real-time retrieval."}
+{"idx": 8, "title": "Entanglement Asymmetry for Higher and Noninvertible Symmetries", "date": "", "ddg_snippet": "Abstract Entanglement asymmetry is an observable in quantum systems, constructed using quantum- information methods, suited to detecting symmetry breaking in states — possibly out of equi- librium — relative to a subsystem. In this paper we define the asymmetry for generalized finite symmetries, including higher-form and noninvertible ones. To this end, we introduce a “sym- metrizer ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.16311v1", "content": "Abstract Entanglement asymmetry is an observable in quantum systems, constructed using quantum- information methods, suited to detecting symmetry breaking in states — possibly out of equi- librium — relative to a subsystem. In this paper we define the asymmetry for generalized finite symmetries, including higher-form and noninvertible ones. To this end, we introduce a “sym- metrizer ..."}
+{"idx": 9, "title": "Computing the soft anomalous dimension with massless particles using...", "date": "", "ddg_snippet": "(3.18). This forms a ( E + 1 ) dimensional space, where the first E entries correspond to the powers of xj with 1 ≤ j ≤ E , while the ( E + 1 )-th component corresponds to the scaling of each monomial with λ in the limit considered.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.18017", "content": "(3.18). This forms a ( E + 1 ) dimensional space, where the first E entries correspond to the powers of xj with 1 ≤ j ≤ E , while the ( E + 1 )-th component corresponds to the scaling of each monomial with λ in the limit considered."}
diff --git a/data/sampled_jsons/sitearxiv.org_2503.17332_Table_6_CVE_CVSS_3.x_rating_highest.jsonl b/data/sampled_jsons/sitearxiv.org_2503.17332_Table_6_CVE_CVSS_3.x_rating_highest.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1a6d93ec8993a4db89a6e704266804681e9ffebd
--- /dev/null
+++ b/data/sampled_jsons/sitearxiv.org_2503.17332_Table_6_CVE_CVSS_3.x_rating_highest.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "CVE -Bench: A Benchmark for AI Agents' Ability to Exploit Real-World...", "date": "", "ddg_snippet": "shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3.1 base score. Table 6 . Details of reproduced CVEs . CVSS 3 . x Rating Affected Web Application.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.17332", "content": "shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3.1 base score. Table 6 . Details of reproduced CVEs . CVSS 3 . x Rating Affected Web Application."}
+{"idx": 1, "title": "[2503.17332] CVE-Bench: A Benchmark for AI Agents' Ability to ... CVE-Bench: A Benchmark for AI Agents' Ability to Exploit Real ... CVE-Bench: A Benchmark for AI Agents’ Ability to Exploit Real ... Common Vulnerability Scoring System Prediction based on Open ... Shedding Light on CVSS Scoring Inconsistencies: A User ... Dynamic Vulnerability Criticality Calculator for Industrial ... A Model for Android and iOS Applications Risk Calculation ...", "date": "", "ddg_snippet": "Mar 21, 2025 · Building a benchmark for real-world vulnerabilities involves both specialized expertise to reproduce exploits and a systematic approach to evaluating unpredictable threats. To address this challenge, we introduce CVE -Bench, a real-world cybersecurity benchmark based on critical-severity Common Vulnerabilities and Exposures. We show the details of each CVE in Table 6 , including its identifier, publication date, CVSS 3.x score, affected web application, and our reproduced attack type. Matching the scale of Cybench (Zhang et al., 2024), CVE -Bench includes 40 CVEs. As shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3 .1 base score. We further categorize CVEs based on type of the vulnerable web applications in Table 3 . ABSTRACT The number of newly published vulnerabilities is constantly increas-ing. Until now, the information available when a new vulnerability is published is manually assessed by experts using a Common Vul-nerability Scoring System ( CVSS ) vector and score. This assessment is time consuming and requires expertise. Various works already try to predict CVSS vectors or scores using machine ... Which factors influence CVSS assessments? We systematically investigate these questions in an online survey with 196 CVSS users. We show that specific CVSS metrics are inconsistently evaluated for widespread vulnerability types, including Top 3 vulnerabilities from the “2022 CWE Top 25 Most Dangerous Software Weaknesses” list. Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment. Before vulnerabilities are registered in a database they are assigned a CVE number and a CVSS score which can be used for software product compatibility testing, within threat intelligence community to build security alerts and within public watch lists such as Open Web Application Security Project (OWASP) to rank and manage risks [9].", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.17332", "content": "Mar 21, 2025 · Building a benchmark for real-world vulnerabilities involves both specialized expertise to reproduce exploits and a systematic approach to evaluating unpredictable threats. To address this challenge, we introduce CVE -Bench, a real-world cybersecurity benchmark based on critical-severity Common Vulnerabilities and Exposures. We show the details of each CVE in Table 6 , including its identifier, publication date, CVSS 3.x score, affected web application, and our reproduced attack type. Matching the scale of Cybench (Zhang et al., 2024), CVE -Bench includes 40 CVEs. As shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3 .1 base score. We further categorize CVEs based on type of the vulnerable web applications in Table 3 . ABSTRACT The number of newly published vulnerabilities is constantly increas-ing. Until now, the information available when a new vulnerability is published is manually assessed by experts using a Common Vul-nerability Scoring System ( CVSS ) vector and score. This assessment is time consuming and requires expertise. Various works already try to predict CVSS vectors or scores using machine ... Which factors influence CVSS assessments? We systematically investigate these questions in an online survey with 196 CVSS users. We show that specific CVSS metrics are inconsistently evaluated for widespread vulnerability types, including Top 3 vulnerabilities from the “2022 CWE Top 25 Most Dangerous Software Weaknesses” list. Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment. Before vulnerabilities are registered in a database they are assigned a CVE number and a CVSS score which can be used for software product compatibility testing, within threat intelligence community to build security alerts and within public watch lists such as Open Web Application Security Project (OWASP) to rank and manage risks [9]."}
+{"idx": 2, "title": "CVE -Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World...", "date": "", "ddg_snippet": "Measuring the common vulnerability scoring system base score equation. US Department of Commerce, National Institute of Standards and Technology, 2022. Table 6 : Details of reproduced CVEs .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332", "content": "Measuring the common vulnerability scoring system base score equation. US Department of Commerce, National Institute of Standards and Technology, 2022. Table 6 : Details of reproduced CVEs ."}
+{"idx": 3, "title": "CVE -Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World...", "date": "", "ddg_snippet": "Measuring the common vulnerability scoring system base score equation. CVE ID. Publication Date. CVSS 3 . x Rating . Web Application. Reproduced Attack Type.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332v3", "content": "Measuring the common vulnerability scoring system base score equation. CVE ID. Publication Date. CVSS 3 . x Rating . Web Application. Reproduced Attack Type."}
+{"idx": 4, "title": "CVE -Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World...", "date": "", "ddg_snippet": "arXiv: 2503 . 17332 v1 [cs.CR] 21 Mar 2025. Table 2: Distribution of based severity scores of CVEs in CVE -Bench. The severity score is calculated according to the base score of the Common Vulnerability Scoring System ( CVSS ) version 3 .1. Report issue for preceding element.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332v1", "content": "arXiv: 2503 . 17332 v1 [cs.CR] 21 Mar 2025. Table 2: Distribution of based severity scores of CVEs in CVE -Bench. The severity score is calculated according to the base score of the Common Vulnerability Scoring System ( CVSS ) version 3 .1. Report issue for preceding element."}
+{"idx": 5, "title": "CVE-Bench: A Benchmark for AI Agents’ Ability to Exploit Real ...", "date": "", "ddg_snippet": "Matching the scale of Cybench (Zhang et al., 2024), CVE -Bench includes 40 CVEs. As shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3 .1 base score. We further categorize CVEs based on type of the vulnerable web applications in Table 3 .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332v2", "content": "Matching the scale of Cybench (Zhang et al., 2024), CVE -Bench includes 40 CVEs. As shown in Table 2, CVEs in CVE -Bench have a minimum of 9.0 severity score measured by the Common Vulnerability Scoring System ( CVSS ) version 3 .1 base score. We further categorize CVEs based on type of the vulnerable web applications in Table 3 ."}
+{"idx": 6, "title": "Shedding Light on CVSS Scoring Inconsistencies: A User ... Dynamic Vulnerability Criticality Calculator for Industrial ... A Model for Android and iOS Applications Risk Calculation ...", "date": "", "ddg_snippet": "Which factors influence CVSS assessments? We systematically investigate these questions in an online survey with 196 CVSS users. We show that specific CVSS metrics are inconsistently evaluated for widespread vulnerability types, including Top 3 vulnerabilities from the “2022 CWE Top 25 Most Dangerous Software Weaknesses” list. Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment. Before vulnerabilities are registered in a database they are assigned a CVE number and a CVSS score which can be used for software product compatibility testing, within threat intelligence community to build security alerts and within public watch lists such as Open Web Application Security Project (OWASP) to rank and manage risks [9].", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2308.15259v2", "content": "Which factors influence CVSS assessments? We systematically investigate these questions in an online survey with 196 CVSS users. We show that specific CVSS metrics are inconsistently evaluated for widespread vulnerability types, including Top 3 vulnerabilities from the “2022 CWE Top 25 Most Dangerous Software Weaknesses” list. Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment. Before vulnerabilities are registered in a database they are assigned a CVE number and a CVSS score which can be used for software product compatibility testing, within threat intelligence community to build security alerts and within public watch lists such as Open Web Application Security Project (OWASP) to rank and manage risks [9]."}
+{"idx": 7, "title": "CVE -Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World...", "date": "", "ddg_snippet": "Measuring the common vulnerability scoring system base score equation.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332v4", "content": "Measuring the common vulnerability scoring system base score equation."}
+{"idx": 8, "title": "Common Vulnerability Scoring System Prediction based on Open ...", "date": "", "ddg_snippet": "ABSTRACT The number of newly published vulnerabilities is constantly increas-ing. Until now, the information available when a new vulnerability is published is manually assessed by experts using a Common Vul-nerability Scoring System ( CVSS ) vector and score. This assessment is time consuming and requires expertise. Various works already try to predict CVSS vectors or scores using machine ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2210.02143", "content": "ABSTRACT The number of newly published vulnerabilities is constantly increas-ing. Until now, the information available when a new vulnerability is published is manually assessed by experts using a Common Vul-nerability Scoring System ( CVSS ) vector and score. This assessment is time consuming and requires expertise. Various works already try to predict CVSS vectors or scores using machine ..."}
+{"idx": 9, "title": "Dynamic Vulnerability Criticality Calculator for Industrial ...", "date": "", "ddg_snippet": "Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2404.16854v1", "content": "Mar 20, 2024 · These adjustments allow our model to generate a dynamic vulnerability-only value. Our model leverages the Common Vulnerabilities and Exposures ( CVE ) along with the Common Vulnerability Scoring System ( CVSS ) records to identify vulnerabilities within the assets of the target environment."}
diff --git a/data/sampled_jsons/sitegithub.com_FLAIROxah2ac2_README.md.jsonl b/data/sampled_jsons/sitegithub.com_FLAIROxah2ac2_README.md.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..80ac071511f321c7aac4f7fd9fc2cd2635bf92dc
--- /dev/null
+++ b/data/sampled_jsons/sitegithub.com_FLAIROxah2ac2_README.md.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "GitHub - FLAIROx/JaxMARL: Multi-Agent Reinforcement Learning ... JaxMARL/README.md at main · FLAIROx/JaxMARL · GitHub behaviour-distillation/README.md at main · FLAIROx ... - GitHub", "date": "", "ddg_snippet": "JaxMARL combines ease-of-use with GPU-enabled efficiency, and supports a wide range of commonly used MARL environments as well as popular baseline algorithms. Our aim is for one library that enables thorough evaluation of MARL methods across a wide range of tasks and against relevant baselines. We also introduce SMAX, a vectorised, simplified version of the popular StarCraft Multi-Agent Challenge, which removes the need to run the StarCraft II game engine. For more details, take a look at our blog post or our Colab notebook, which walks through the basic usage. See full list on github . com We follow CleanRL's philosophy of providing single file implementations which can be found within the baselines directory. We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config. See full list on github . com Environments - Before installing, ensure you have the correct JAX version for your hardware accelerator. The JaxMARL environments can be installed directly from PyPi: Algorithms - If you would like to also run the algorithms, install the source code as follows: 1.Clone the repository: 2.The requirements for IPPO & MAPPO can be installed with: 3.If you would also like to run the Q-learning algorithms, Python 3.9 is required along with additional dependencies: Test Scripts - To run our test scripts, some additional dependencies are required (for comparisons against existing implementations), these can be installed with: See full list on github . com Basic JaxMARL API Usage 🖥️ Actions, observations, rewards and done values are passed as dictionaries keyed by agent name, allowing for differing action and observation spaces. The done dictionary contains an additional \"__all__\" key, specifying whether the episode has ended. We follow a parallel structure, with each agent passing an action at each timestep. For asynchronous games, such as Hanabi, a dummy action is passed for agents not acting at a given timestep. Dockerfile 🐋 To help get experiments up and running we include a Dockerfile and its corresponding Makefile. With Docker and the Nvidia Container Toolkit installed, the container can be built with:The built container can then be run: See full list on github . com Please contribute! Please take a look at our contributing guide for how to add an environment/algorithm or submit a bug report. Our roadmap also lives there. See full list on github . com If you use JaxMARL in your work, please cite us as follows: See full list on github . com There are a number of other libraries which inspired this work, we encourage you to take a look! JAX-native algorithms: •Mava: JAX implementations of IPPO and MAPPO, two popular MARL algorithms. •PureJaxRL: JAX implementation of PPO, and demonstration of end-to-end JAX-based RL training. •Minimax: JAX implementations of autocurricula baselines for RL. JAX-native environments: See full list on github . com We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config. Code for Behaviour Distillation (ICML 2024). Contribute to FLAIROx /behaviour-distillation development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/JaxMARL", "content": "JaxMARL combines ease-of-use with GPU-enabled efficiency, and supports a wide range of commonly used MARL environments as well as popular baseline algorithms. Our aim is for one library that enables thorough evaluation of MARL methods across a wide range of tasks and against relevant baselines. We also introduce SMAX, a vectorised, simplified version of the popular StarCraft Multi-Agent Challenge, which removes the need to run the StarCraft II game engine. For more details, take a look at our blog post or our Colab notebook, which walks through the basic usage. See full list on github . com We follow CleanRL's philosophy of providing single file implementations which can be found within the baselines directory. We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config. See full list on github . com Environments - Before installing, ensure you have the correct JAX version for your hardware accelerator. The JaxMARL environments can be installed directly from PyPi: Algorithms - If you would like to also run the algorithms, install the source code as follows: 1.Clone the repository: 2.The requirements for IPPO & MAPPO can be installed with: 3.If you would also like to run the Q-learning algorithms, Python 3.9 is required along with additional dependencies: Test Scripts - To run our test scripts, some additional dependencies are required (for comparisons against existing implementations), these can be installed with: See full list on github . com Basic JaxMARL API Usage 🖥️ Actions, observations, rewards and done values are passed as dictionaries keyed by agent name, allowing for differing action and observation spaces. The done dictionary contains an additional \"__all__\" key, specifying whether the episode has ended. We follow a parallel structure, with each agent passing an action at each timestep. For asynchronous games, such as Hanabi, a dummy action is passed for agents not acting at a given timestep. Dockerfile 🐋 To help get experiments up and running we include a Dockerfile and its corresponding Makefile. With Docker and the Nvidia Container Toolkit installed, the container can be built with:The built container can then be run: See full list on github . com Please contribute! Please take a look at our contributing guide for how to add an environment/algorithm or submit a bug report. Our roadmap also lives there. See full list on github . com If you use JaxMARL in your work, please cite us as follows: See full list on github . com There are a number of other libraries which inspired this work, we encourage you to take a look! JAX-native algorithms: •Mava: JAX implementations of IPPO and MAPPO, two popular MARL algorithms. •PureJaxRL: JAX implementation of PPO, and demonstration of end-to-end JAX-based RL training. •Minimax: JAX implementations of autocurricula baselines for RL. JAX-native environments: See full list on github . com We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config. Code for Behaviour Distillation (ICML 2024). Contribute to FLAIROx /behaviour-distillation development by creating an account on GitHub ."}
+{"idx": 1, "title": "GitHub - FLAIROx/ah2ac2: Ad-Hoc Human-AI Coordination ...", "date": "", "ddg_snippet": "Welcome to the Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 )! The objective of AH2AC2 is to facilitate the development of AI agents capable of effective collaboration with human-like partners, especially in scenarios with limited prior interaction data.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/ah2ac2", "content": "Welcome to the Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 )! The objective of AH2AC2 is to facilitate the development of AI agents capable of effective collaboration with human-like partners, especially in scenarios with limited prior interaction data."}
+{"idx": 2, "title": "ah2ac2/README.md at production · FLAIROx/ah2ac2 · GitHub", "date": "", "ddg_snippet": "Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 ). Contribute to FLAIROx / ah2ac2 development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/ah2ac2/blob/production/README.md", "content": "Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 ). Contribute to FLAIROx / ah2ac2 development by creating an account on GitHub ."}
+{"idx": 3, "title": "GitHub · Where software is built", "date": "", "ddg_snippet": "Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 ). Contribute to FLAIROx / ah2ac2 development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/ah2ac2/labels", "content": "Ad-Hoc Human-AI Coordination Challenge ( AH2AC2 ). Contribute to FLAIROx / ah2ac2 development by creating an account on GitHub ."}
+{"idx": 4, "title": "JaxMARL/README.md at main · FLAIROx/JaxMARL · GitHub", "date": "", "ddg_snippet": "We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/JaxMARL/blob/main/README.md", "content": "We use Hydra to manage our config files, with specifics explained in each algorithm's README . Most files include wandb logging code, this is disabled by default but can be enabled within the file's config."}
+{"idx": 5, "title": "behaviour-distillation/README.md at main · FLAIROx ... - GitHub", "date": "", "ddg_snippet": "Code for Behaviour Distillation (ICML 2024). Contribute to FLAIROx /behaviour-distillation development by creating an account on GitHub .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/behaviour-distillation/blob/main/README.md", "content": "Code for Behaviour Distillation (ICML 2024). Contribute to FLAIROx /behaviour-distillation development by creating an account on GitHub ."}
+{"idx": 6, "title": "multiagent-reinforcement-learning · GitHub Topics · GitHub", "date": "", "ddg_snippet": "FLAIROx / ah 2 ac 2 . Star 4. Code. Issues. Pull requests.This is the second paper of my PhD career. Focusing on the general stochastic approximation framework, it proposes a federated algorithm that finds the optimum of an average of contractive operators.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/topics/multiagent-reinforcement-learning?o=desc&s=updated", "content": "FLAIROx / ah 2 ac 2 . Star 4. Code. Issues. Pull requests.This is the second paper of my PhD career. Focusing on the general stochastic approximation framework, it proposes a federated algorithm that finds the optimum of an average of contractive operators."}
+{"idx": 7, "title": "Foerster Lab for AI Research · GitHub", "date": "", "ddg_snippet": "FLAIROx / ah 2 ac 2 ’s past year of commit activity.Hugo theme for the FLAIR website.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx", "content": "FLAIROx / ah 2 ac 2 ’s past year of commit activity.Hugo theme for the FLAIR website."}
+{"idx": 8, "title": "GitHub - FLAIROx /Kinetix: Reinforcement learning on general...", "date": "", "ddg_snippet": "README . md . pyproject.toml.Repository files navigation. README . MIT license.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/Kinetix", "content": "README . md . pyproject.toml.Repository files navigation. README . MIT license."}
+{"idx": 9, "title": "GitHub - FLAIROx /jaxirl: Contains JAX implementation of algorithms...", "date": "", "ddg_snippet": "README . md . requirements.txt.Repository files navigation. README . Apache-2.0 license.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/FLAIROx/jaxirl", "content": "README . md . requirements.txt.Repository files navigation. README . Apache-2.0 license."}
diff --git "a/data/sampled_jsons/sitegithub.com_fiveai_understanding_safety_finetuning_learning_rates_\316\267M_\316\267S.jsonl" "b/data/sampled_jsons/sitegithub.com_fiveai_understanding_safety_finetuning_learning_rates_\316\267M_\316\267S.jsonl"
new file mode 100644
index 0000000000000000000000000000000000000000..48e8421ab66e8dcc0903b2d423b9662e26a79732
--- /dev/null
+++ "b/data/sampled_jsons/sitegithub.com_fiveai_understanding_safety_finetuning_learning_rates_\316\267M_\316\267S.jsonl"
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "understanding_safety_finetuning/ssft at main · fiveai ... - GitHub", "date": "", "ddg_snippet": "Official Code for What Makes and Breaks Safety Fine-tuning? A Mechanistic Study (NeurIPS 2024) - fiveai/understanding_safety_finetuning", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/tree/main/ssft", "content": "Official Code for What Makes and Breaks Safety Fine-tuning? A Mechanistic Study (NeurIPS 2024) - fiveai/understanding_safety_finetuning"}
+{"idx": 1, "title": "FiveAI - GitHub", "date": "", "ddg_snippet": "We're developing the software and tools to make self-driving vehicle services a reality. - FiveAI", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai", "content": "We're developing the software and tools to make self-driving vehicle services a reality. - FiveAI"}
+{"idx": 2, "title": "understanding_safety_finetuning/pretrain.py at main · fiveai ...", "date": "", "ddg_snippet": "fiveai / understanding_safety_finetuning Public Notifications Fork 0 Star 0 Code Issues Actions Projects Security Insights", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/blob/main/pretrain.py", "content": "fiveai / understanding_safety_finetuning Public Notifications Fork 0 Star 0 Code Issues Actions Projects Security Insights"}
+{"idx": 3, "title": "Actions · fiveai/understanding_safety_finetuning · GitHub", "date": "", "ddg_snippet": "GitHub Actions makes it easy to automate all your software workflows, now with world-class CI/CD. Build, test, and deploy your code right from GitHub.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/actions", "content": "GitHub Actions makes it easy to automate all your software workflows, now with world-class CI/CD. Build, test, and deploy your code right from GitHub."}
+{"idx": 4, "title": "Issues · fiveai/understanding_safety_finetuning · GitHub", "date": "", "ddg_snippet": "fiveai / understanding_safety_finetuning Public Notifications You must be signed in to change notification settings Fork 0 Star 0 Code Projects Security Insights", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/issues", "content": "fiveai / understanding_safety_finetuning Public Notifications You must be signed in to change notification settings Fork 0 Star 0 Code Projects Security Insights"}
+{"idx": 5, "title": "Settings · Custom properties · fiveai/understanding_safety_finetuning ...", "date": "", "ddg_snippet": "Custom properties Custom properties allow you to decorate your repository with information such as compliance frameworks, data sensitivity, or project details.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/custom-properties", "content": "Custom properties Custom properties allow you to decorate your repository with information such as compliance frameworks, data sensitivity, or project details."}
+{"idx": 6, "title": "fiveai / understanding_safety_finetuning Public - GitHub", "date": "", "ddg_snippet": "Official Code for What Makes and Breaks Safety Fine-tuning? A Mechanistic Study (NeurIPS 2024) - understanding_safety_finetuning /trainer_pretrain.py at main · fiveai ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/blob/main/trainer_pretrain.py", "content": "Official Code for What Makes and Breaks Safety Fine-tuning? A Mechanistic Study (NeurIPS 2024) - understanding_safety_finetuning /trainer_pretrain.py at main · fiveai ..."}
+{"idx": 7, "title": "GitHub - fiveai/understanding_safety_finetuning: Official Code for What ...", "date": "", "ddg_snippet": "The official implementation of \"What Makes and Breaks Safety Fine-tuning? A Mechanistic Study\". This work is accepted to NeurIPS 2024. To better understand the underlying factors that make models safe via safety fine-tuning, we design a synthetic data generation framework that captures salient ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning", "content": "The official implementation of \"What Makes and Breaks Safety Fine-tuning? A Mechanistic Study\". This work is accepted to NeurIPS 2024. To better understand the underlying factors that make models safe via safety fine-tuning, we design a synthetic data generation framework that captures salient ..."}
+{"idx": 8, "title": "understanding_safety_finetuning/README.md at main · fiveai ...", "date": "", "ddg_snippet": "Abstract To better understand the underlying factors that make models safe via safety fine-tuning, we design a synthetic data generation framework that captures salient aspects of an unsafe input by modeling the interaction between the task the model is asked to perform (e.g., \"design\") versus the specific concepts the task is asked to be performed upon (e.g., a \"cycle\" vs. a \"bomb ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/blob/main/README.md", "content": "Abstract To better understand the underlying factors that make models safe via safety fine-tuning, we design a synthetic data generation framework that captures salient aspects of an unsafe input by modeling the interaction between the task the model is asked to perform (e.g., \"design\") versus the specific concepts the task is asked to be performed upon (e.g., a \"cycle\" vs. a \"bomb ..."}
+{"idx": 9, "title": "Releases · fiveai/understanding_safety_finetuning - GitHub", "date": "", "ddg_snippet": "You can create a release to package software, along with release notes and links to binary files, for other people to use. Learn more about releases in our docs ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/fiveai/understanding_safety_finetuning/releases", "content": "You can create a release to package software, along with release notes and links to binary files, for other people to use. Learn more about releases in our docs ..."}
diff --git a/data/sampled_jsons/siteopenreview.net_4uOEiitySn_A_Section_3.2_four-step_self-supervised_learning_process.jsonl b/data/sampled_jsons/siteopenreview.net_4uOEiitySn_A_Section_3.2_four-step_self-supervised_learning_process.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cc6e4c8f8d54bfda33f02f3deb3790beccbe22a7
--- /dev/null
+++ b/data/sampled_jsons/siteopenreview.net_4uOEiitySn_A_Section_3.2_four-step_self-supervised_learning_process.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Speech Self-Supervised Learning Using Diffusion Model ...", "date": "", "ddg_snippet": "Abstract While self-supervised learning (SSL) in speech has greatly reduced the reliance of speech process -ing systems on annotated corpora, the success of SSL still hinges on the availability of a large-scale unannotated corpus, which is still often impracti-cal for many low-resource languages or under pri-vacy concerns.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=ecnpYYHjt9", "content": "Abstract While self-supervised learning (SSL) in speech has greatly reduced the reliance of speech process -ing systems on annotated corpora, the success of SSL still hinges on the availability of a large-scale unannotated corpus, which is still often impracti-cal for many low-resource languages or under pri-vacy concerns."}
+{"idx": 1, "title": "Self-Supervised Visual Representation Learning for Medical ...", "date": "", "ddg_snippet": "In this study, we attempt to present a review of those methods and show how the self-supervised learning paradigm evolved over the years. Additionally, we also present an exhaustive review of the self-supervised methods applied to medical image analysis.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=3Wg1oErMcJ", "content": "In this study, we attempt to present a review of those methods and show how the self-supervised learning paradigm evolved over the years. Additionally, we also present an exhaustive review of the self-supervised methods applied to medical image analysis."}
+{"idx": 2, "title": "Mind Your Augmentation: The Key to Decoupling Dense Self ...", "date": "", "ddg_snippet": "Jan 16, 2024 · This paper presents a way for self-supervised learning tackling dense visual prediction tasks such as segmentation and detection. The authors propose a solution inspired by CutMix augmentation that combines visual signals from two images to create in context and out of context visual tokens.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=WQYHbr36Fo", "content": "Jan 16, 2024 · This paper presents a way for self-supervised learning tackling dense visual prediction tasks such as segmentation and detection. The authors propose a solution inspired by CutMix augmentation that combines visual signals from two images to create in context and out of context visual tokens."}
+{"idx": 3, "title": "Self-Supervised Graph Neural Networks for Improved...", "date": "", "ddg_snippet": "Jan 28, 2022 · Conducted new experiments comparing self-supervised pre-training to end-to-end training using the self-supervised prediction task as an auxiliary task in Section 3.2 “Comparison between self-supervised pre-training and auxiliary learning ” and Appendix L (Reviewer 6FoL).", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=k9bx1EfHI_-", "content": "Jan 28, 2022 · Conducted new experiments comparing self-supervised pre-training to end-to-end training using the self-supervised prediction task as an auxiliary task in Section 3.2 “Comparison between self-supervised pre-training and auxiliary learning ” and Appendix L (Reviewer 6FoL)."}
+{"idx": 4, "title": "Rethinking Uniformity in Self-Supervised Representation Learning", "date": "", "ddg_snippet": "Feb 1, 2023 · Self-supervised representation learning has achieved great success in many machine learning tasks. While many research efforts focus on learning better representations by preventing the model from the \\emph {collapse} problem, less attention has been drawn to analyzing the collapse degrees of representations.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=hFUlfiyf1oQ", "content": "Feb 1, 2023 · Self-supervised representation learning has achieved great success in many machine learning tasks. While many research efforts focus on learning better representations by preventing the model from the \\emph {collapse} problem, less attention has been drawn to analyzing the collapse degrees of representations."}
+{"idx": 5, "title": "SELF-S SET REPRESENTATION LEARNING FOR UNSUPERVISED META-LEARNING", "date": "", "ddg_snippet": "In this section , we empirically validate the effectiveness of our set representation learning framework on several downstream few-shot classification tasks, and compare our Set-SimCLR against UML baselines and instance-level self-supervised baselines in subsection 4 .1 and 4 .2, respectively.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=kIAx30hYi_p", "content": "In this section , we empirically validate the effectiveness of our set representation learning framework on several downstream few-shot classification tasks, and compare our Set-SimCLR against UML baselines and instance-level self-supervised baselines in subsection 4 .1 and 4 .2, respectively."}
+{"idx": 6, "title": "Self-Supervised Dataset Distillation for Transfer Learning", "date": "", "ddg_snippet": "Jan 16, 2024 · Our self-supervised dataset distillation method is based on the recent success of self-supervised learning in transfer learning . Self-supervised learning methods are known to learn features that transfer more effectively to downstream tasks than supervised ones [1, 2, 3].", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=h57gkDO2Yg", "content": "Jan 16, 2024 · Our self-supervised dataset distillation method is based on the recent success of self-supervised learning in transfer learning . Self-supervised learning methods are known to learn features that transfer more effectively to downstream tasks than supervised ones [1, 2, 3]."}
+{"idx": 7, "title": "S 4 RL: Surprisingly Simple Self - Supervision for Offline... | OpenReview", "date": "", "ddg_snippet": "Keywords: Offline Reinforcement Learning , Data Augmentation, Self - Supervised Learning . Abstract: Offline reinforcement learning proposes to learn policies from large collected datasets without interacting with the physical environment.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=8xC5NNej-l_", "content": "Keywords: Offline Reinforcement Learning , Data Augmentation, Self - Supervised Learning . Abstract: Offline reinforcement learning proposes to learn policies from large collected datasets without interacting with the physical environment."}
+{"idx": 8, "title": "Self - supervised Learning for Segmentation and Quantifica", "date": "", "ddg_snippet": "Experiments and Results Self - supervised models provide superior performance in limited data settingsImpact of pre-training data on self - supervised learning In this section , we provide the details of our deep learning framework, including the...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=izFnURFG3f", "content": "Experiments and Results Self - supervised models provide superior performance in limited data settingsImpact of pre-training data on self - supervised learning In this section , we provide the details of our deep learning framework, including the..."}
+{"idx": 9, "title": "Mugs: A Multi-Granular Self - Supervised Learning ... | OpenReview", "date": "", "ddg_snippet": "Keywords: multi-granular learning , contrastive learning , self - supervised learning .In this work, for the first time, we propose an effective MUlti-Granular Self - supervised learning (Mugs) framework to explicitly learn multi-granular visual features.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=ALcz2n3Wsdf", "content": "Keywords: multi-granular learning , contrastive learning , self - supervised learning .In this work, for the first time, we propose an effective MUlti-Granular Self - supervised learning (Mugs) framework to explicitly learn multi-granular visual features."}
diff --git a/data/sampled_jsons/siteopenreview.net_gQlxd3Mtru_LEnergy_Equation_10.jsonl b/data/sampled_jsons/siteopenreview.net_gQlxd3Mtru_LEnergy_Equation_10.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bbaaddb2fc258f913f3bb7bc88b13de3d0f2a111
--- /dev/null
+++ b/data/sampled_jsons/siteopenreview.net_gQlxd3Mtru_LEnergy_Equation_10.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Ordinary differential equations on graph networks - OpenReview", "date": "", "ddg_snippet": "To our knowledge, all existing graph networks have discrete depth. Inspired by neural ordinary differential equation (NODE) for data in the Euclidean domain, we extend the idea of continuous-depth models to graph data, and propose graph ordinary differential equation (GODE).", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=SJg9z6VFDr", "content": "To our knowledge, all existing graph networks have discrete depth. Inspired by neural ordinary differential equation (NODE) for data in the Euclidean domain, we extend the idea of continuous-depth models to graph data, and propose graph ordinary differential equation (GODE)."}
+{"idx": 1, "title": "Jingqiu Zhou - OpenReview", "date": "", "ddg_snippet": "A distributed linear equation solver for least square solutions Xuan Wang, Jingqiu Zhou, Shaoshuai Mou, Martin J. Corless 2017 (modified: 31 Oct 2022) CDC 2017", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/profile?id=~Jingqiu_Zhou1", "content": "A distributed linear equation solver for least square solutions Xuan Wang, Jingqiu Zhou, Shaoshuai Mou, Martin J. Corless 2017 (modified: 31 Oct 2022) CDC 2017"}
+{"idx": 2, "title": "LightGaussian: Unbounded 3D Gaussian Compression... | OpenReview", "date": "", "ddg_snippet": "Recent advances in real-time neural rendering using point-based techniques have enabled broader adoption of 3D representations.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=6AeIDnrTN2&noteId=riNuRFyEcq", "content": "Recent advances in real-time neural rendering using point-based techniques have enabled broader adoption of 3D representations."}
+{"idx": 3, "title": "DPM-Solver-v3: Improved Diffusion ODE Solver with... | OpenReview", "date": "", "ddg_snippet": "Experiments show that DPM-Solver-v3 achieves consistently better or comparable performance in both unconditional and conditional sampling with both pixel-space and latent-space DPMs, especially in 5$\\sim$ 10 NFEs.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=9fWKExmKa0", "content": "Experiments show that DPM-Solver-v3 achieves consistently better or comparable performance in both unconditional and conditional sampling with both pixel-space and latent-space DPMs, especially in 5$\\sim$ 10 NFEs."}
+{"idx": 4, "title": "ELoRA: Low-Rank Adaptation for Equivariant GNNs - OpenReview", "date": "", "ddg_snippet": "On the rMD17 organic dataset, ELoRA achieves a 25.5\\% improvement in energy prediction accuracy and a 23.7\\% improvement in force prediction accuracy compared to full-parameter fine-tuning. Similarly, across 10 inorganic datasets, ELoRA achieves average improvements of 12.3\\% and 14.4\\% in energy and force predictions, respectively.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=hcoxm3Vwgy", "content": "On the rMD17 organic dataset, ELoRA achieves a 25.5\\% improvement in energy prediction accuracy and a 23.7\\% improvement in force prediction accuracy compared to full-parameter fine-tuning. Similarly, across 10 inorganic datasets, ELoRA achieves average improvements of 12.3\\% and 14.4\\% in energy and force predictions, respectively."}
+{"idx": 5, "title": "Forum - OpenReview", "date": "", "ddg_snippet": "The server responded with the following message: Too many requests: You have made 73 requests, surpassing the limit of 60 requests. Please try again in 10 seconds (2025-09-20-9000053)", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=X41c4uB4k0", "content": "The server responded with the following message: Too many requests: You have made 73 requests, surpassing the limit of 60 requests. Please try again in 10 seconds (2025-09-20-9000053)"}
+{"idx": 6, "title": "G eneration with e nergy - based M odels", "date": "", "ddg_snippet": "For the energy network, a learning rate of 1 × 10 −4 and the number of training epochs of 20 are used.The sampling paths all start with ethane. An Adam optimizer Kingma & Ba (2014) is used with an initial learning rate of 3 × 10 −4.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qkTEaJ9orc1", "content": "For the energy network, a learning rate of 1 × 10 −4 and the number of training epochs of 20 are used.The sampling paths all start with ethane. An Adam optimizer Kingma & Ba (2014) is used with an initial learning rate of 3 × 10 −4."}
+{"idx": 7, "title": "Learning Stochastic Dynamics From Snap Shots Through Regularized ...", "date": "", "ddg_snippet": "An improved approach for calculating energy landscape of gene networks from moment equations . Chaos: An Interdisciplinary Journal of Nonlinear Science, 33(2), 2023.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=gQlxd3Mtru", "content": "An improved approach for calculating energy landscape of gene networks from moment equations . Chaos: An Interdisciplinary Journal of Nonlinear Science, 33(2), 2023."}
+{"idx": 8, "title": "How Gradient descent balances features: A dynamical analysis for...", "date": "", "ddg_snippet": "This paper investigates the fundamental regression task of learning k neurons (\\emph {a.k.a.} teachers) from Gaussian input, using two-layer ReLU neural networks with width m (\\emph {a.k.a.} students) and m, k = O (1), trained via gradient descent under proper initialization and a small step-size. Our analysis follows a three-phase structure: \\emph {alignment} after weak recovery, \\emph ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=25j2ZEgwTj", "content": "This paper investigates the fundamental regression task of learning k neurons (\\emph {a.k.a.} teachers) from Gaussian input, using two-layer ReLU neural networks with width m (\\emph {a.k.a.} students) and m, k = O (1), trained via gradient descent under proper initialization and a small step-size. Our analysis follows a three-phase structure: \\emph {alignment} after weak recovery, \\emph ..."}
+{"idx": 9, "title": "EGSDE: Unpaired Image-to-Image Translation via Energy-Guided...", "date": "", "ddg_snippet": "We propose energy-guided stochastic differential equations (EGSDE) that employs an energy function pretrained on both the source and target domains to guide the inference process of pretrained SDE for realistic and faithful unpaired I2I.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=xxgp42Qz6dL", "content": "We propose energy-guided stochastic differential equations (EGSDE) that employs an energy function pretrained on both the source and target domains to guide the inference process of pretrained SDE for realistic and faithful unpaired I2I."}
diff --git a/data/sampled_jsons/siteresearchgate.net_389091382_Table_1_dataset_statistics_train_interactions_density.jsonl b/data/sampled_jsons/siteresearchgate.net_389091382_Table_1_dataset_statistics_train_interactions_density.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b7e4d29b3cccf3e1af79737d2b32b3d9c2aba56e
--- /dev/null
+++ b/data/sampled_jsons/siteresearchgate.net_389091382_Table_1_dataset_statistics_train_interactions_density.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Statistics of the datasets Dataset 1 # Users # Items # Interactions Density", "date": "", "ddg_snippet": "Download scientific diagram | Statistics of the datasets Dataset 1 # Users # Items # Interactions Density from publication: Latent Structures Mining with Contrastive Modality Fusion for Multimedia ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Statistics-of-the-datasets-Dataset-1-Users-Items-Interactions-Density_tbl1_355841189", "content": "Download scientific diagram | Statistics of the datasets Dataset 1 # Users # Items # Interactions Density from publication: Latent Structures Mining with Contrastive Modality Fusion for Multimedia ..."}
+{"idx": 1, "title": "Dataset Statistics Dataset # Users # Items # Interactions Density", "date": "", "ddg_snippet": "Download scientific diagram | Dataset Statistics Dataset # Users # Items # Interactions Density from publication: Disentangling Past-Future Modeling in Sequential Recommendation via Dual Networks ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Dataset-Statistics-Dataset-Users-Items-Interactions-Density_tbl1_364771849", "content": "Download scientific diagram | Dataset Statistics Dataset # Users # Items # Interactions Density from publication: Disentangling Past-Future Modeling in Sequential Recommendation via Dual Networks ..."}
+{"idx": 2, "title": "Table 1 : Statistics of the three datasets. The first row of each...", "date": "", "ddg_snippet": "The detailed descriptions of the three datasets are shown in Table 1 . The first row of each dataset corresponds to the numbers of users, items and interactions , while the other rows correspond to ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Statistics-of-the-three-datasets-The-first-row-of-each-dataset-corresponds-to-the_tbl1_325964999", "content": "The detailed descriptions of the three datasets are shown in Table 1 . The first row of each dataset corresponds to the numbers of users, items and interactions , while the other rows correspond to ..."}
+{"idx": 3, "title": "Statistics of three Datasets. Density of dataset is #Interactions/(#U...", "date": "", "ddg_snippet": "The statistics of the datasets are summarized in Table 2 and the selected meta paths for all data sets are shown in Table 3.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Statistics-of-three-Datasets-Density-of-dataset-is-Interactions-U-sers-Items_tbl2_358423744", "content": "The statistics of the datasets are summarized in Table 2 and the selected meta paths for all data sets are shown in Table 3."}
+{"idx": 4, "title": "Comparison table with data environment of baselines. % Density Train ...", "date": "", "ddg_snippet": "Comparison table with data environment of baselines. % Density Train means the density of the training dataset in the corresponding methods. # Interaction Test means the average number of ratings ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Comparison-table-with-data-environment-of-baselines-Density-Train-means-the-density-of_tbl1_356125526", "content": "Comparison table with data environment of baselines. % Density Train means the density of the training dataset in the corresponding methods. # Interaction Test means the average number of ratings ..."}
+{"idx": 5, "title": "Statistics of the dataset. | Download Table - ResearchGate", "date": "", "ddg_snippet": "However, datasets that reflect user interaction traces in cross-lingual settings required to train and evaluate the user interaction models are mostly missing.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Statistics-of-the-dataset_tbl1_326655967", "content": "However, datasets that reflect user interaction traces in cross-lingual settings required to train and evaluate the user interaction models are mostly missing."}
+{"idx": 6, "title": "TABLE 1 Dataset Statistics: Amount and Density - ResearchGate", "date": "", "ddg_snippet": "Download Table | Dataset Statistics : Amount and Density from publication: Social Recommendation with Cross-Domain Transferable Knowledge | Recommender systems can suffer from data sparsity and ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Dataset-Statistics-Amount-and-Density_tbl1_277354369", "content": "Download Table | Dataset Statistics : Amount and Density from publication: Social Recommendation with Cross-Domain Transferable Knowledge | Recommender systems can suffer from data sparsity and ..."}
+{"idx": 7, "title": "Data set statistics. | Download Table - ResearchGate", "date": "", "ddg_snippet": "The social network of the law firm data set in this article is considered as the ground truth social network. Table 1 shows network statistics of the networks of the data sets used in this paper.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/Data-set-statistics_tbl1_308875452", "content": "The social network of the law firm data set in this article is considered as the ground truth social network. Table 1 shows network statistics of the networks of the data sets used in this paper."}
+{"idx": 8, "title": "The statistics of datasets. Dataset #User #Item #Interaction Sparsity", "date": "", "ddg_snippet": "Download scientific diagram | The statistics of datasets . Dataset #User #Item # Interaction Sparsity from publication: Shapley Value-driven Data Pruning for Recommender Systems | Recommender ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/The-statistics-of-datasets-Dataset-User-Item-Interaction-Sparsity_tbl1_392167770", "content": "Download scientific diagram | The statistics of datasets . Dataset #User #Item # Interaction Sparsity from publication: Shapley Value-driven Data Pruning for Recommender Systems | Recommender ..."}
+{"idx": 9, "title": "descriptive statistics of the train set data - ResearchGate", "date": "", "ddg_snippet": "By organizing and aggregating the data, the pandas function can generate a table similar to Table 1 that contains an overall summary of the data.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/descriptive-statistics-of-the-train-set-data_tbl1_368489443", "content": "By organizing and aggregating the data, the pandas function can generate a table similar to Table 1 that contains an overall summary of the data."}
diff --git a/data/sampled_jsons/sitethecvf.com_Instant_Gaussian_Stream_Equation_6_motion_interpolation.jsonl b/data/sampled_jsons/sitethecvf.com_Instant_Gaussian_Stream_Equation_6_motion_interpolation.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2c932f7048a12a5fd2f8f5894dc4faea024fe411
--- /dev/null
+++ b/data/sampled_jsons/sitethecvf.com_Instant_Gaussian_Stream_Equation_6_motion_interpolation.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "CVPR Poster DiET-GS: Diffusion Prior and Event Stream ...", "date": "", "ddg_snippet": "In this paper, we present DiET-GS, a diffusion prior and event stream -assisted motion deblurring 3DGS. Our framework effectively leverages blur-free event ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/poster/32624", "content": "In this paper, we present DiET-GS, a diffusion prior and event stream -assisted motion deblurring 3DGS. Our framework effectively leverages blur-free event ..."}
+{"idx": 1, "title": "CVPR 2025 Awards - The Computer Vision Foundation", "date": "", "ddg_snippet": "Instant Gaussian Stream : Fast and Generalizable Streaming of Dynamic Scene Reconstruction via Gaussian Splatting ... interpolation, for each Gaussian we ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/awards_detail", "content": "Instant Gaussian Stream : Fast and Generalizable Streaming of Dynamic Scene Reconstruction via Gaussian Splatting ... interpolation, for each Gaussian we ..."}
+{"idx": 2, "title": "DynaMoDe-NeRF: Motion-aware Deblurring Neural Radiance ...", "date": "", "ddg_snippet": "by A Kumar · 2025 · Cited by 2 — This enforces multiview consistency and motion aware de- blurring by treating motion as a global cue. It also enables continuous blur kernel spatial location ... 11 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Kumar_DynaMoDe-NeRF_Motion-aware_Deblurring_Neural_Radiance_Field_for_Dynamic_Scenes_CVPR_2025_paper.pdf", "content": "by A Kumar · 2025 · Cited by 2 — This enforces multiview consistency and motion aware de- blurring by treating motion as a global cue. It also enables continuous blur kernel spatial location ... 11 pages"}
+{"idx": 3, "title": "Unified Particles for Versatile Motion Synthesis and Rendering", "date": "", "ddg_snippet": "by Y Feng · 2025 · Cited by 15 — Versatile motion synthesis and rendering. Gaussian Splashing (GSP) is a unified framework combining 3D Gaussian Splatting. (3DGS) and position-based dynamics. 12 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Feng_Gaussian_Splashing_Unified_Particles_for_Versatile_Motion_Synthesis_and_Rendering_CVPR_2025_paper.pdf", "content": "by Y Feng · 2025 · Cited by 15 — Versatile motion synthesis and rendering. Gaussian Splashing (GSP) is a unified framework combining 3D Gaussian Splatting. (3DGS) and position-based dynamics. 12 pages"}
+{"idx": 4, "title": "Gaussian Splashing: Unified Particles for Versatile Motion ...", "date": "", "ddg_snippet": "Abstract: We demonstrate the feasibility of integrating physics-based animations of solids and fluids with 3D Gaussian Splatting (3DGS) to create novel ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/poster/33536", "content": "Abstract: We demonstrate the feasibility of integrating physics-based animations of solids and fluids with 3D Gaussian Splatting (3DGS) to create novel ..."}
+{"idx": 5, "title": "Full-DoF Egomotion Estimation for Event Cameras Using ...", "date": "", "ddg_snippet": "by J Zhao · 2025 · Cited by 4 — We demonstrate the possibility of recovering full-. DoF egomotion parameters for both angular and linear velocities without requiring extra sensor measurements ... 10 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Zhao_Full-DoF_Egomotion_Estimation_for_Event_Cameras_Using_Geometric_Solvers_CVPR_2025_paper.pdf", "content": "by J Zhao · 2025 · Cited by 4 — We demonstrate the possibility of recovering full-. DoF egomotion parameters for both angular and linear velocities without requiring extra sensor measurements ... 10 pages"}
+{"idx": 6, "title": "Track: Poster Session 4 - CVPR", "date": "", "ddg_snippet": "14 Jun 2025 — In this paper, we propose Instant Gaussian Stream (IGS), a fast and ... Extensive experiments show that our method outperforms prior arts in both ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/session/35268", "content": "14 Jun 2025 — In this paper, we propose Instant Gaussian Stream (IGS), a fast and ... Extensive experiments show that our method outperforms prior arts in both ..."}
+{"idx": 7, "title": "Track: Poster Session 2 - CVPR", "date": "", "ddg_snippet": "13 Jun 2025 — In this work, we introduce Monocular and Generalizable Gaussian Talking Head Animation (MGGTalk), which requires monocular datasets and ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/session/35266", "content": "13 Jun 2025 — In this work, we introduce Monocular and Generalizable Gaussian Talking Head Animation (MGGTalk), which requires monocular datasets and ..."}
+{"idx": 8, "title": "Track: Poster Session 5 - CVPR", "date": "", "ddg_snippet": "15 Jun 2025 — To address this, we introduce GIFStream, a novel 4D Gaussian representation using a canonical space and a deformation field enhanced with time- ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/session/35269", "content": "15 Jun 2025 — To address this, we introduce GIFStream, a novel 4D Gaussian representation using a canonical space and a deformation field enhanced with time- ..."}
+{"idx": 9, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}
diff --git a/data/sampled_jsons/stochastic_localization_samplers_generative_models.jsonl b/data/sampled_jsons/stochastic_localization_samplers_generative_models.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..fec647d76020976ae68f382a346dbe533fc2b75c
--- /dev/null
+++ b/data/sampled_jsons/stochastic_localization_samplers_generative_models.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Track: Deep Generative Model 1", "date": "", "ddg_snippet": "We also demonstrate our improved sampler for training deep energy-based models on high dimensional discrete image data.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2021/session/11978", "content": "We also demonstrate our improved sampler for training deep energy-based models on high dimensional discrete image data."}
+{"idx": 1, "title": "53 questions with answers in STOCHASTIC MODELS | Science topic", "date": "", "ddg_snippet": "Probabilistic Model : A probabilistic model is a type of stochastic model that explicitly uses probability distributions to describe uncertainty.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/topic/Stochastic-Models", "content": "Probabilistic Model : A probabilistic model is a type of stochastic model that explicitly uses probability distributions to describe uncertainty."}
+{"idx": 2, "title": "ICML 2020 Papers", "date": "", "ddg_snippet": "Robust One-Bit Recovery via ReLU Generative Networks: Near-Optimal Statistical Rate and Global Landscape Analysis ... Optimization: Models , Insights ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2020/papers.html?filter=keywords", "content": "Robust One-Bit Recovery via ReLU Generative Networks: Near-Optimal Statistical Rate and Global Landscape Analysis ... Optimization: Models , Insights ..."}
+{"idx": 3, "title": "ICML 2020 Papers", "date": "", "ddg_snippet": "Robust One-Bit Recovery via ReLU Generative Networks: Near-Optimal Statistical Rate and Global Landscape Analysis ... Optimization: Models , Insights ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2020/papers.html", "content": "Robust One-Bit Recovery via ReLU Generative Networks: Near-Optimal Statistical Rate and Global Landscape Analysis ... Optimization: Models , Insights ..."}
+{"idx": 4, "title": "[2502.00921] Blink of an eye: a simple theory for feature", "date": "", "ddg_snippet": "In this work we develop a simple, unifying theory to explain this phenomenon using the formalism of stochastic localization samplers .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2502.00921", "content": "In this work we develop a simple, unifying theory to explain this phenomenon using the formalism of stochastic localization samplers ."}
+{"idx": 5, "title": "From discrete-time policies to continuous-time diffusion", "date": "", "ddg_snippet": "We study the problem of training neural stochastic differential equations, ordiffusion models , to sample from a Boltzmann distribution without access ...", "subpage_snippet": "", "source": "deeplearn.org", "link": "https://deeplearn.org/arxiv/566343/from-discrete-time-policies-to-continuous-time-diffusion-samplers:-asymptotic-equivalences-and-faster-training", "content": "We study the problem of training neural stochastic differential equations, ordiffusion models , to sample from a Boltzmann distribution without access ..."}
+{"idx": 6, "title": "Downloads", "date": "", "ddg_snippet": "... Model for Understanding the ... Bundle Networks: Fiber Bundles, Local Trivializations, and a Generative Approach to Exploring Many-to-one Maps", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/Downloads/2022", "content": "... Model for Understanding the ... Bundle Networks: Fiber Bundles, Local Trivializations, and a Generative Approach to Exploring Many-to-one Maps"}
+{"idx": 7, "title": "ICLR 2024 Schedule", "date": "", "ddg_snippet": "Navigating the Design Space of Equivariant Diffusion-Based Generative Models for De Novo 3D Molecule Generation ... Models : Reducing Model Size ...", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2024/calendar", "content": "Navigating the Design Space of Equivariant Diffusion-Based Generative Models for De Novo 3D Molecule Generation ... Models : Reducing Model Size ..."}
+{"idx": 8, "title": "Edge-Selector Model Applied for Local Search Neighborhood for", "date": "", "ddg_snippet": "The main of our method is an edge solution selector model , which classifies solution edges to identify prohibited moves during the local search ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.14071v1", "content": "The main of our method is an edge solution selector model , which classifies solution edges to identify prohibited moves during the local search ..."}
+{"idx": 9, "title": "CS 8803: Counting and Sampling (Fall 2024) – Zongchen Chen", "date": "", "ddg_snippet": "... local-to-global induction, trickle-down theorem Oct 2, Oct 7, Oct 9: Spectral independence, optimal relaxation time, universality, disagreement ...", "subpage_snippet": "", "source": "sites.gatech.edu", "link": "https://sites.gatech.edu/zongchenchen/cs8803fall24/", "content": "... local-to-global induction, trickle-down theorem Oct 2, Oct 7, Oct 9: Spectral independence, optimal relaxation time, universality, disagreement ..."}
diff --git a/data/sampled_jsons/structure_constant_spurious_correlation_Schubert_polynomials_dataset_year_2024.jsonl b/data/sampled_jsons/structure_constant_spurious_correlation_Schubert_polynomials_dataset_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0ccd37da75b3b38cc6c1bfc467efd75464b6b4bc
--- /dev/null
+++ b/data/sampled_jsons/structure_constant_spurious_correlation_Schubert_polynomials_dataset_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "PDF Boolean structure constants", "date": "", "ddg_snippet": "In this paper, we provide an explicit formula for the (equivariant) Schubert structure constants cw uv across all Lie types when the elements u, v, w are boolean. In particular, in type A, all Schubert structure constants on boolean elements are either 0 or 1. Keywords: Schubert calculus, Boolean, structure constant , Dynkin diagram", "subpage_snippet": "", "source": "www.mat.univie.ac.at", "link": "https://www.mat.univie.ac.at/~slc/wpapers/FPSAC2025/66/Boolean+structure+constants.pdf", "content": "In this paper, we provide an explicit formula for the (equivariant) Schubert structure constants cw uv across all Lie types when the elements u, v, w are boolean. In particular, in type A, all Schubert structure constants on boolean elements are either 0 or 1. Keywords: Schubert calculus, Boolean, structure constant , Dynkin diagram"}
+{"idx": 1, "title": "Schubert calculus from polyhedral parametrizations of Demazure crystals", "date": "", "ddg_snippet": "In the case G=SLn+1(C), Lascoux-Schützenberger [38]gave a specific choice {Sw|w∈W}of representatives, called Schubert polynomials , which have good combinatorial properties. Since Swrepresents [Xw]for w∈W, we know the structure constant cu,vwby computing the product Su⋅Svof Schubert polynomials in the coinvariant algebra of W.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0001870822000172", "content": "In the case G=SLn+1(C), Lascoux-Schützenberger [38]gave a specific choice {Sw|w∈W}of representatives, called Schubert polynomials , which have good combinatorial properties. Since Swrepresents [Xw]for w∈W, we know the structure constant cu,vwby computing the product Su⋅Svof Schubert polynomials in the coinvariant algebra of W."}
+{"idx": 2, "title": "unum-cloud/ann-arxiv-2m · Datasets at Hugging Face", "date": "", "ddg_snippet": "Structures on a graph defined by means of semicubes, and Djokovi\\'{c}'s and Winkler's relations play an important role in the theory of partial cubes.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/datasets/unum-cloud/ann-arxiv-2m", "content": "Structures on a graph defined by means of semicubes, and Djokovi\\'{c}'s and Winkler's relations play an important role in the theory of partial cubes."}
+{"idx": 3, "title": "mteb/raw_arxiv · Datasets at Hugging Face", "date": "", "ddg_snippet": "Structures on a graph defined by means of semicubes, and Djokovi\\'{c}'s and Winkler's relations play an important role in the theory of partial cubes.", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/datasets/mteb/raw_arxiv", "content": "Structures on a graph defined by means of semicubes, and Djokovi\\'{c}'s and Winkler's relations play an important role in the theory of partial cubes."}
+{"idx": 4, "title": "ICML 2024 Schedule", "date": "", "ddg_snippet": "10:30] LSEnet: Lorentz Structural ... 11:15] Pruned Pivot: Correlation Clustering Algorithm for Dynamic, Parallel, and Local Computation Models", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2024/calendar", "content": "10:30] LSEnet: Lorentz Structural ... 11:15] Pruned Pivot: Correlation Clustering Algorithm for Dynamic, Parallel, and Local Computation Models"}
+{"idx": 5, "title": "The Complexity of Learning Sparse Superposed Features with", "date": "", "ddg_snippet": "... composition across attention layers (Yang and Hu, 2021b ) , and revealing how models develop structured representations (Elhage et al., 2022 ) .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.05407v4", "content": "... composition across attention layers (Yang and Hu, 2021b ) , and revealing how models develop structured representations (Elhage et al., 2022 ) ."}
+{"idx": 6, "title": "Towards Principled Evaluations of Sparse Autoencoders for", "date": "", "ddg_snippet": "... to the indirect object identification (IOI) task using GPT-2 Small, with sparse autoencoders (SAEs) trained on either the IOI or OpenWebText datasets ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2405.08366v3", "content": "... to the indirect object identification (IOI) task using GPT-2 Small, with sparse autoencoders (SAEs) trained on either the IOI or OpenWebText datasets ..."}
+{"idx": 7, "title": "Sunbeam: an extensible pipeline for analyzing metagenomic", "date": "", "ddg_snippet": "... because existing tools for analyzing nucleotide sequence complexity [ 35 , 36 , 37 ] did not meet our needs in terms of speed, removal of spurious ...", "subpage_snippet": "", "source": "microbiomejournal.biomedcentral.com", "link": "https://microbiomejournal.biomedcentral.com/articles/10.1186/s40168-019-0658-x", "content": "... because existing tools for analyzing nucleotide sequence complexity [ 35 , 36 , 37 ] did not meet our needs in terms of speed, removal of spurious ..."}
+{"idx": 8, "title": "Interpreting wide-band neural activity using convolutional", "date": "", "ddg_snippet": "Neural activity is high dimensional and often sparse, while the available datasets are typically incomplete, being both temporally and spatially ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/articles/66551", "content": "Neural activity is high dimensional and often sparse, while the available datasets are typically incomplete, being both temporally and spatially ..."}
+{"idx": 9, "title": "ESD - Multiscale fractal dimension analysis of a reduced order", "date": "", "ddg_snippet": "... established concepts like the box-counting dimension (e.g., Steinhaus , 1954 ; Mandelbrot , 1967 ; Ott , 2002 ) , generalized correlation ...", "subpage_snippet": "", "source": "esd.copernicus.org", "link": "https://esd.copernicus.org/articles/12/837/2021/", "content": "... established concepts like the box-counting dimension (e.g., Steinhaus , 1954 ; Mandelbrot , 1967 ; Ott , 2002 ) , generalized correlation ..."}
diff --git a/data/sampled_jsons/text-to-video_retrieval_MSVD_12_frames_OR_8_frames_preprocessing_experimental_setup_year_2024.jsonl b/data/sampled_jsons/text-to-video_retrieval_MSVD_12_frames_OR_8_frames_preprocessing_experimental_setup_year_2024.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c19a34d0159f2aad1451fcd84dacbc4be258cc43
--- /dev/null
+++ b/data/sampled_jsons/text-to-video_retrieval_MSVD_12_frames_OR_8_frames_preprocessing_experimental_setup_year_2024.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Unmasked Teacher: Towards Training-Efficient Video Foundation...", "date": "", "ddg_snippet": "Table 14: Text - to - video retrieval on MSRVTT, DiDeMo, AcitivityNet, LSMDC, and MSVD . “ #Pairs” denotes the number of pre -training pairs. Models pre -trained with large-scale pairs are noted in gray.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2303.16058", "content": "Table 14: Text - to - video retrieval on MSRVTT, DiDeMo, AcitivityNet, LSMDC, and MSVD . “ #Pairs” denotes the number of pre -training pairs. Models pre -trained with large-scale pairs are noted in gray."}
+{"idx": 1, "title": "Multi-modal Video", "date": "", "ddg_snippet": "Text - to - video retrieval is another important area of video understanding research.[ 12 ] suggested to “diuse” in the latent space instead of the pixel space to reduce the cost of training (aka. latent diusion model (LDM) or Stable Diusion).", "subpage_snippet": "", "source": "trepo.tuni.fi", "link": "https://trepo.tuni.fi/bitstream/handle/10024/147432/978-952-03-2872-6.pdf?sequence=2&isAllowed=y", "content": "Text - to - video retrieval is another important area of video understanding research.[ 12 ] suggested to “diuse” in the latent space instead of the pixel space to reduce the cost of training (aka. latent diusion model (LDM) or Stable Diusion)."}
+{"idx": 2, "title": "(PDF) HiTeA: Hierarchical Temporal-Aware Video -Language Pre -training", "date": "", "ddg_snippet": "size video frames to 224 ×224. During fine-tuning, follow-. ing [25,28], we randomly sample 12 frames for text - to - video . retrieval , 16 frames for video question answering and video. captions. We perform uniform sampling during inference.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/366789557_HiTeA_Hierarchical_Temporal-Aware_Video-Language_Pre-training", "content": "size video frames to 224 ×224. During fine-tuning, follow-. ing [25,28], we randomly sample 12 frames for text - to - video . retrieval , 16 frames for video question answering and video. captions. We perform uniform sampling during inference."}
+{"idx": 3, "title": "VideoCLIP-XL: Advancing Long Description Understanding for Video", "date": "", "ddg_snippet": "For fine-tuned setting of text - video retrieval on traditional benchmarks, we tune our pre -trained VideoCLIP-XL with the vanilla text - video con-trastive learning loss on each training set of the evaluated benchmarks. During both training and testing, we sample 12 frames .", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2024.emnlp-main.898.pdf", "content": "For fine-tuned setting of text - video retrieval on traditional benchmarks, we tune our pre -trained VideoCLIP-XL with the vanilla text - video con-trastive learning loss on each training set of the evaluated benchmarks. During both training and testing, we sample 12 frames ."}
+{"idx": 4, "title": "Manfrotto AVENGER 12\"x 12 \" frames Η1200M Modular Frames size...", "date": "", "ddg_snippet": "Click to enlarge.Manfrotto AVENGER 8 “x 8 ” frames Η800M Modular Frames size :244cm x 244cm. €30,00. Γρήγορη προβολή.", "subpage_snippet": "", "source": "makedonltd.gr", "link": "https://makedonltd.gr/product/manfrotto-avenger-12-quot-x-12-quot-frames-i1200m-modular-frames-size-366cm-x-366cm/", "content": "Click to enlarge.Manfrotto AVENGER 8 “x 8 ” frames Η800M Modular Frames size :244cm x 244cm. €30,00. Γρήγορη προβολή."}
+{"idx": 5, "title": "picture frames custom frames art to frames – Excel Image Group", "date": "", "ddg_snippet": "with letters of name 16×24 picture cheap 9× 12 frames frame and mat white collage picture frames 8 by 20 frame 7 x 38 30×26 20×28 white frame 16×16 frame matted to 12×12 22×55 frame 22×22 white frame acrylic floating frame custom canvas frames online what size frame.", "subpage_snippet": "", "source": "excelimagegroup.com", "link": "https://excelimagegroup.com/picture-frames-custom-frames-art-to-frames/", "content": "with letters of name 16×24 picture cheap 9× 12 frames frame and mat white collage picture frames 8 by 20 frame 7 x 38 30×26 20×28 white frame 16×16 frame matted to 12×12 22×55 frame 22×22 white frame acrylic floating frame custom canvas frames online what size frame."}
+{"idx": 6, "title": "Discover videos related to 12 Frames on TikTok.", "date": "", "ddg_snippet": "12 frames per second #animations.TikTok video from Kota! (@kota_vwv_bear): \"dont hate on me plz i swear i listen to actual rock/metal to.", "subpage_snippet": "", "source": "www.tiktok.com", "link": "https://www.tiktok.com/discover/12-Frames", "content": "12 frames per second #animations.TikTok video from Kota! (@kota_vwv_bear): \"dont hate on me plz i swear i listen to actual rock/metal to."}
+{"idx": 7, "title": "2025 Online Picture Framing vs. Local Frame Shops: — Modern...", "date": "", "ddg_snippet": "12 x 12 Frames .With larger workforces and readily available inventory, these e-commerce platforms can mass-produce custom frames and ship them quickly, catering to customers who value speed. Local frame shops, while capable of producing high-quality custom frames , often face...", "subpage_snippet": "", "source": "www.modernmemorydesign.com", "link": "https://www.modernmemorydesign.com/blogs/news/online-picture-framing-vs-local-frame-shops-efficiency-trends-and-convenience", "content": "12 x 12 Frames .With larger workforces and readily available inventory, these e-commerce platforms can mass-produce custom frames and ship them quickly, catering to customers who value speed. Local frame shops, while capable of producing high-quality custom frames , often face..."}
+{"idx": 8, "title": "Preloaders.net - Free loading animations", "date": "", "ddg_snippet": "Setting up loading animations Please, wait...A snakes from balls moving around. Download. 12 Frames . 128×128. Fading lines. Free SVG.", "subpage_snippet": "", "source": "icons8.com", "link": "https://icons8.com/preloaders/en/free", "content": "Setting up loading animations Please, wait...A snakes from balls moving around. Download. 12 Frames . 128×128. Fading lines. Free SVG."}
+{"idx": 9, "title": "What is an *****on ones***** animation | StudyX", "date": "", "ddg_snippet": "12 frames per one second of film: This would be \"on twos\" animation, where each drawing is held for two frames , resulting in 12 drawings per second. This is a common technique to reduce workload while maintaining acceptable motion quality.", "subpage_snippet": "", "source": "studyx.ai", "link": "https://studyx.ai/questions/4ltnfwt/what-is-an-on-ones-animation-composed-of-12-frames-per-one-second-of-film-6-frames-per", "content": "12 frames per one second of film: This would be \"on twos\" animation, where each drawing is held for two frames , resulting in 12 drawings per second. This is a common technique to reduce workload while maintaining acceptable motion quality."}
diff --git a/data/sampled_jsons/vh9yEPLeyD_Can_We_Leave_Deepfake_Data_Behind_in_Training_Deepfake_Detector_Section_4.1.jsonl b/data/sampled_jsons/vh9yEPLeyD_Can_We_Leave_Deepfake_Data_Behind_in_Training_Deepfake_Detector_Section_4.1.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9048549b6138f1054147cbddad5a9de6fdc869bf
--- /dev/null
+++ b/data/sampled_jsons/vh9yEPLeyD_Can_We_Leave_Deepfake_Data_Behind_in_Training_Deepfake_Detector_Section_4.1.jsonl
@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Italian Chicken Breasts - Easy Baked Dinner - COOKtheSTORY", "date": "", "ddg_snippet": "Jun 23, 2023 · This Italian chicken recipe really is one of those dishes that comes together in minutes, looks impressive, and tastes fabulous! Scroll down to read more about how it all comes together or click here to jump straight down to the recipe.", "subpage_snippet": "", "source": "cookthestory.com", "link": "https://cookthestory.com/italian-chicken-breast/", "content": "Jun 23, 2023 · This Italian chicken recipe really is one of those dishes that comes together in minutes, looks impressive, and tastes fabulous! Scroll down to read more about how it all comes together or click here to jump straight down to the recipe."}
+{"idx": 1, "title": "Italian Baked Chicken Breast Recipe - The Mediterranean Dish", "date": "", "ddg_snippet": "Sep 23, 2021 · Tender, juicy baked chicken breast recipe, covered in delicious Italian flavors! A few tips make all the difference in this must-try chicken!", "subpage_snippet": "", "source": "www.themediterraneandish.com", "link": "https://www.themediterraneandish.com/easy-italian-baked-chicken-recipe/", "content": "Sep 23, 2021 · Tender, juicy baked chicken breast recipe, covered in delicious Italian flavors! A few tips make all the difference in this must-try chicken!"}
+{"idx": 2, "title": "Tender Italian Baked Chicken Recipe", "date": "", "ddg_snippet": "Jun 5, 2024 · Try this Italian baked chicken recipe for moist, flavorful results in just 30 minutes.", "subpage_snippet": "", "source": "www.allrecipes.com", "link": "https://www.allrecipes.com/recipe/52005/tender-italian-baked-chicken/", "content": "Jun 5, 2024 · Try this Italian baked chicken recipe for moist, flavorful results in just 30 minutes."}
+{"idx": 3, "title": "Baked Chicken with Italian Seasoning - Laura Fuentes", "date": "", "ddg_snippet": "Apr 10, 2025 · Drooling alert! This must-try recipe for baked chicken with Italian seasoning makes chicken breasts taste super flavorful and juicy.", "subpage_snippet": "", "source": "www.laurafuentes.com", "link": "https://www.laurafuentes.com/baked-chicken-with-italian-seasoning/", "content": "Apr 10, 2025 · Drooling alert! This must-try recipe for baked chicken with Italian seasoning makes chicken breasts taste super flavorful and juicy."}
+{"idx": 4, "title": "Italian Baked Chicken - Barefeet In The Kitchen", "date": "", "ddg_snippet": "Aug 26, 2024 · Italian baked chicken thighs with a crisp and flavorful skin are possible with less than 5 minutes worth of effort! This post may contain affiliate links. Please read our disclosure policy. Find out how easy it is to make delicious, crisp-skinned chicken in the oven with Italian Baked Chicken .", "subpage_snippet": "", "source": "barefeetinthekitchen.com", "link": "https://barefeetinthekitchen.com/crispy-italian-oven-chicken-recipe/", "content": "Aug 26, 2024 · Italian baked chicken thighs with a crisp and flavorful skin are possible with less than 5 minutes worth of effort! This post may contain affiliate links. Please read our disclosure policy. Find out how easy it is to make delicious, crisp-skinned chicken in the oven with Italian Baked Chicken ."}
+{"idx": 5, "title": "Baked Italian Chicken - Healthy Fitness Meals", "date": "", "ddg_snippet": "Apr 10, 2025 · My baked Italian Chicken recipe with veggies is a simple one-pan meal that’s both satisfying and so good. Seasoned chicken is pan-seared and then topped with olives, tomatoes, artichokes, sun-dried tomatoes, capers, and plenty of mozzarella cheese. Bake until bubbly and cheesy for the perfect weeknight dinner that tastes gourmet!", "subpage_snippet": "", "source": "healthyfitnessmeals.com", "link": "https://healthyfitnessmeals.com/baked-italian-chicken/", "content": "Apr 10, 2025 · My baked Italian Chicken recipe with veggies is a simple one-pan meal that’s both satisfying and so good. Seasoned chicken is pan-seared and then topped with olives, tomatoes, artichokes, sun-dried tomatoes, capers, and plenty of mozzarella cheese. Bake until bubbly and cheesy for the perfect weeknight dinner that tastes gourmet!"}
+{"idx": 6, "title": "Italian Baked Chicken (Easy Sheet Pan Recipe) - The Diet Chef", "date": "", "ddg_snippet": "Oct 15, 2023 · Chicken fillets and colorful veggies are all tossed in a savory Italian marinade and baked to golden perfection in one sheet pan, then topped with cheesy parmesan and fresh parsley. It's the easiest and most satisfying weeknight dinner and will fit seamlessly into your weekly rotation.", "subpage_snippet": "", "source": "www.thedietchefs.com", "link": "https://www.thedietchefs.com/italian-baked-chicken/", "content": "Oct 15, 2023 · Chicken fillets and colorful veggies are all tossed in a savory Italian marinade and baked to golden perfection in one sheet pan, then topped with cheesy parmesan and fresh parsley. It's the easiest and most satisfying weeknight dinner and will fit seamlessly into your weekly rotation."}
+{"idx": 7, "title": "Baked Italian Chicken Breast Recipe - JamiesItalian.sg", "date": "", "ddg_snippet": "May 13, 2025 · This baked Italian chicken breast recipe is a delightful and easy-to-make dish that brings the flavors of Italy right to your dinner table. With a blend of aromatic herbs and a touch of garlic, this recipe ensures juicy and flavorful chicken breasts that are perfect for any weeknight meal or special occasion.", "subpage_snippet": "", "source": "jamiesitalian.sg", "link": "https://jamiesitalian.sg/baked-italian-chicken-breast-recipe/", "content": "May 13, 2025 · This baked Italian chicken breast recipe is a delightful and easy-to-make dish that brings the flavors of Italy right to your dinner table. With a blend of aromatic herbs and a touch of garlic, this recipe ensures juicy and flavorful chicken breasts that are perfect for any weeknight meal or special occasion."}
+{"idx": 8, "title": "Baked Italian Chicken Breast with Brie: A Decadent and Delicious...", "date": "", "ddg_snippet": "May 24, 2025 · Craving a dinner that’s both elegant and comforting? Look no further than this Baked Italian Chicken Breast with Brie recipe. It’s surprisingly easy to make, yet the combination of juicy, flavorful chicken, creamy melted brie, and vibrant Italian herbs creates a truly unforgettable dish.", "subpage_snippet": "", "source": "italianchef.org", "link": "https://italianchef.org/baked-italian-chicken-breast-with-brie-a-decadent-and-delicious-dinner/", "content": "May 24, 2025 · Craving a dinner that’s both elegant and comforting? Look no further than this Baked Italian Chicken Breast with Brie recipe. It’s surprisingly easy to make, yet the combination of juicy, flavorful chicken, creamy melted brie, and vibrant Italian herbs creates a truly unforgettable dish."}
+{"idx": 9, "title": "29+ Mouthwatering Italian Baked Chicken Breast Recipes You Need...", "date": "", "ddg_snippet": "Mar 26, 2025 · From classic chicken Parmesan to zesty lemon herb variations, there are countless ways to elevate a simple chicken breast with the flavors of Italy. In this article, we’ve compiled over 29 Italian baked chicken breast recipes that will inspire your next meal.", "subpage_snippet": "", "source": "chefsbliss.com", "link": "https://chefsbliss.com/italian-baked-chicken-breast-recipes/", "content": "Mar 26, 2025 · From classic chicken Parmesan to zesty lemon herb variations, there are countless ways to elevate a simple chicken breast with the flavors of Italy. In this article, we’ve compiled over 29 Italian baked chicken breast recipes that will inspire your next meal."}