Add files using upload-large-folder tool

data/sampled_jsons/ACL_Anthology_CVE-Bench_paper_'Insufficient_Exploration'_definition_year_2023-2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "CVE - Bench : A Benchmark for AI Agents’ Ability to Exploit Real-World...", "date": "", "ddg_snippet": "In CVE - Bench , we collect 40 Common Vulnerabilities and Exposures (CVEs) in the National Vulnerability Database (Booth et al., 2013) .Table 5: Frequency of common failure modes of agents. Insufficient exploration is a key bottleneck for all agents.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.17332", "content": "In CVE - Bench , we collect 40 Common Vulnerabilities and Exposures (CVEs) in the National Vulnerability Database (Booth et al., 2013) .Table 5: Frequency of common failure modes of agents. Insufficient exploration is a key bottleneck for all agents."}
+{"idx": 1, "title": "Pre-Training Curriculum for Multi-Token Prediction in... - ACL Anthology", "date": "", "ddg_snippet": "Ansar Aynetdinov and Alan Akbik. 2025. Pre-Training Curriculum for Multi-Token Prediction in Language Models. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers ), pages 25573–25588, Vienna, Austria.", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.acl-long.1243/", "content": "Ansar Aynetdinov and Alan Akbik. 2025. Pre-Training Curriculum for Multi-Token Prediction in Language Models. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers ), pages 25573–25588, Vienna, Austria."}
+{"idx": 2, "title": "Metadata correction for 2021.tacl-1.68 · Issue #5987...", "date": "", "ddg_snippet": "acl -org / acl - anthology Public. Notifications You must be signed in to change notification settings. Fork 355.Assignees. anthology -assist.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/acl-org/acl-anthology/issues/5987", "content": "acl -org / acl - anthology Public. Notifications You must be signed in to change notification settings. Fork 355.Assignees. anthology -assist."}
+{"idx": 3, "title": "Английский язык 6 класс Spotlight Английский в фокусе Ваулина.", "date": "", "ddg_snippet": "1b Who are you? (страница 8) Номер 1 a) Look at the cards. Which is a credit card? an identity card? a membership card? a driving licence? b) What information from the list is on each card? full name home address nationality identific...", "subpage_snippet": "", "source": "Reshalka.com", "link": "https://Reshalka.com/uchebniki/6-klass/english/vaulina1/10", "content": "1b Who are you? (страница 8) Номер 1 a) Look at the cards. Which is a credit card? an identity card? a membership card? a driving licence? b) What information from the list is on each card? full name home address nationality identific..."}
+{"idx": 4, "title": "Melania Meme", "date": "", "ddg_snippet": "Melania Meme token that support her initiatives.", "subpage_snippet": "", "source": "melaniameme.com", "link": "https://melaniameme.com/", "content": "Melania Meme token that support her initiatives."}
+{"idx": 5, "title": "PostgreSQL: Documentation: 17: 21.5. Predefined Roles", "date": "", "ddg_snippet": "PostgreSQL provides a set of predefined roles that provide access to certain, commonly needed, privileged capabilities and information. Administrators (including roles that have the CREATEROLE privilege) can GRANT these roles to users and/or other ro...", "subpage_snippet": "", "source": "www.postgresql.org", "link": "https://www.postgresql.org/docs/current/predefined-roles.html", "content": "PostgreSQL provides a set of predefined roles that provide access to certain, commonly needed, privileged capabilities and information. Administrators (including roles that have the CREATEROLE privilege) can GRANT these roles to users and/or other ro..."}
+{"idx": 6, "title": "Litematica - Minecraft Mod", "date": "", "ddg_snippet": "A client-side schematic mod with extra features for creative mode work.", "subpage_snippet": "", "source": "modrinth.com", "link": "https://modrinth.com/mod/litematica", "content": "A client-side schematic mod with extra features for creative mode work."}
+{"idx": 7, "title": "24 Exposures (2013) G XA 3 11111 — Video by Yaşar Demiral | VK", "date": "", "ddg_snippet": "Watch 24 Exposures (2013) G XA 3 11111 1 hr. 17 min 15 s from 16 June 2023 online in HD for free in the VK catalog without signing up!", "subpage_snippet": "", "source": "vk.com", "link": "https://vk.com/video587582593_456239113", "content": "Watch 24 Exposures (2013) G XA 3 11111 1 hr. 17 min 15 s from 16 June 2023 online in HD for free in the VK catalog without signing up!"}
+{"idx": 8, "title": "osu-pps by grumd - osu! farm pp maps and beatmap recommendations", "date": "", "ddg_snippet": "osu! farm pp maps and beatmap recommendations...", "subpage_snippet": "", "source": "osu-pps.com", "link": "https://osu-pps.com/", "content": "osu! farm pp maps and beatmap recommendations..."}
+{"idx": 9, "title": "Video to GIF Converter - FreeConvert.com", "date": "", "ddg_snippet": "Video to GIF Converter. Easily create high-quality GIF animations from your video files online for free. Convert MP4, FLV, MOV, MKV, and more to GIF.", "subpage_snippet": "", "source": "www.freeconvert.com", "link": "https://www.freeconvert.com/convert/video-to-gif", "content": "Video to GIF Converter. Easily create high-quality GIF animations from your video files online for free. Convert MP4, FLV, MOV, MKV, and more to GIF."}

data/sampled_jsons/AERO_Enhancing_Sharding_Blockchain_Deep_Reinforcement_Learning_Account_Migration_reward_function_equ.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Deep Reinforcement Learning Versus Evolution Strategies", "date": "", "ddg_snippet": "by AY Majid · 2021 · Cited by 105 — Abstract— Deep Reinforcement Learning (DRL) and Evolution. Strategies (ESs) have surpassed human-level control in many.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2110.01411", "content": "by AY Majid · 2021 · Cited by 105 — Abstract— Deep Reinforcement Learning (DRL) and Evolution. Strategies (ESs) have surpassed human-level control in many."}
+{"idx": 1, "title": "Sustainable business decision modelling with blockchain ...", "date": "", "ddg_snippet": "by G Wickremasinghe · 2025 — This includes machine learning techniques such as supervised, unsupervised, reinforcement learning , partial differential equations integration ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S2096720925000399", "content": "by G Wickremasinghe · 2025 — This includes machine learning techniques such as supervised, unsupervised, reinforcement learning , partial differential equations integration ..."}
+{"idx": 2, "title": "BlockEmulator: An Emulator Enabling to Test Blockchain ...", "date": "", "ddg_snippet": "by H Huang · 2025 · Cited by 33 — 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 33 — We developed BlockEmulator, which is designed as an experimental platform, particularly for emulating blockchain sharding mechanisms."}
+{"idx": 3, "title": "Reinforcement Learning: An Introduction | Guide books", "date": "", "ddg_snippet": "In Reinforcement Learning , Richard Sutton and Andrew Barto provide a clear and simple account of the field's key ideas and algorithms. This second edition has ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/book/10.5555/3312046", "content": "In Reinforcement Learning , Richard Sutton and Andrew Barto provide a clear and simple account of the field's key ideas and algorithms. This second edition has ..."}
+{"idx": 4, "title": "Blockchain Security and Its Application in Internet of Things", "date": "", "ddg_snippet": "The blockchain model initiates transactions , and then machine learning models are used to classify these transactions as malicious or legitimate. This is a ...", "subpage_snippet": "", "source": "mdpi-res.com", "link": "https://mdpi-res.com/bookfiles/book/10958/Blockchain_Security_and_Its_Application_in_Internet_of_Things.pdf?v=1748528140", "content": "The blockchain model initiates transactions , and then machine learning models are used to classify these transactions as malicious or legitimate. This is a ..."}
+{"idx": 5, "title": "Deep Reinforcement Learning Versus Evolution Strategies", "date": "", "ddg_snippet": "reward function that defines the immediate reward r that the agent observes after taking the action a and the environment transition from s to s′. The total ...", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/pdf/deep-reinforcement-learning-versus-evolution-strategies-a-17xrmnwwgy.pdf", "content": "reward function that defines the immediate reward r that the agent observes after taking the action a and the environment transition from s to s′. The total ..."}
+{"idx": 6, "title": "Paper Digest: WWW 2025 Papers & Highlights", "date": "", "ddg_snippet": "30 Apr 2025 — Highlight: We propose AERO , a deep reinforcement learning framework to facilitate efficient account migration in sharding blockchains. ...", "subpage_snippet": "", "source": "www.paperdigest.org", "link": "https://www.paperdigest.org/2025/04/www-2025-papers-highlights/", "content": "30 Apr 2025 — Highlight: We propose AERO , a deep reinforcement learning framework to facilitate efficient account migration in sharding blockchains. ..."}
+{"idx": 7, "title": "Poster Session | International World Wide ...", "date": "", "ddg_snippet": "AERO: Enhancing Sharding Blockchain via Deep Reinforcement Learning for Account Migration . Posterboard-12, Research-307, 812, A Cooperative Multi-Agent ...", "subpage_snippet": "", "source": "www2025.thewebconf.org", "link": "https://www2025.thewebconf.org/poster-session", "content": "AERO: Enhancing Sharding Blockchain via Deep Reinforcement Learning for Account Migration . Posterboard-12, Research-307, 812, A Cooperative Multi-Agent ..."}
+{"idx": 8, "title": "International Conference on Computer Application and ...", "date": "", "ddg_snippet": "25 Apr 2025 — This article not only provides theoretical basis and practical guidance for the application of deep learning in data mining, but also provides ...", "subpage_snippet": "", "source": "spie.org", "link": "https://spie.org/Publications/Proceedings/Volume/13562", "content": "25 Apr 2025 — This article not only provides theoretical basis and practical guidance for the application of deep learning in data mining, but also provides ..."}
+{"idx": 9, "title": "Programmable Money, Smart Contracts, and Stablecoins", "date": "", "ddg_snippet": "My aim in this post is to demystify the key building blocks ( blockchains , smart contracts, stablecoins) and their interplay with advanced AI, ...", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/@adnanmasood/programmable-money-smart-contracts-and-stablecoins-a-leadership-guide-for-banking-finance-ec0791847977", "content": "My aim in this post is to demystify the key building blocks ( blockchains , smart contracts, stablecoins) and their interplay with advanced AI, ..."}

data/sampled_jsons/AGM-Net_training_dataset_implementation_details_Instant_Gaussian_Stream_year_2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "US20190379976A1 - Systems and methods for using multiple", "date": "", "ddg_snippet": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US20190379976A1/en", "content": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means"}
+{"idx": 1, "title": "US11485517B1 - System and method for communicating a", "date": "", "ddg_snippet": "2021-11-15 Assigned to BETA AIR, LLC reassignment BETA AIR, LLC ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US11485517B1/en", "content": "2021-11-15 Assigned to BETA AIR, LLC reassignment BETA AIR, LLC ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 2, "title": "US20190378383A1 - Systems and methods for transient processing", "date": "", "ddg_snippet": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US20190378383A1/en", "content": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means"}
+{"idx": 3, "title": "US20190379342A1 - Systems and methods using predictive analysis", "date": "", "ddg_snippet": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US20190379342A1/en", "content": "... OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ... hardware implementation of neural networks, neurons or parts of neurons using electronic means"}
+{"idx": 4, "title": "US10854051B2 - Systems and methods for transient processing of", "date": "", "ddg_snippet": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US10854051B2/en", "content": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 5, "title": "US10657779B2 - Systems and methods using predictive analysis of", "date": "", "ddg_snippet": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US10657779B2/en", "content": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 6, "title": "US20190379977A1 - Systems and methods for generating haptic", "date": "", "ddg_snippet": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US20190379977A1/en", "content": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 7, "title": "US11120673B2 - Systems and methods for generating haptic output", "date": "", "ddg_snippet": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US11120673B2/en", "content": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 8, "title": "US10777050B2 - Systems and methods for using multiple actuators", "date": "", "ddg_snippet": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS ).", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US10777050B2/en", "content": "2019-06-21 Assigned to LOFELT GMBH reassignment LOFELT GMBH ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS )."}
+{"idx": 9, "title": "WO2019234191A1 - Systems and methods for generating haptic", "date": "", "ddg_snippet": "Audio frames from an audio source are intercepted by the haptic conversion system to generate a haptic signal, which is passed on through an actuator ...", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/WO2019234191A1/en", "content": "Audio frames from an audio source are intercepted by the haptic conversion system to generate a haptic signal, which is passed on through an actuator ..."}

data/sampled_jsons/Active_Learning_definition_machine_learning.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Machine learning - Wikipedia", "date": "", "ddg_snippet": "Reinforcement learning is an area of machine learning concerned with how software agents ought to take actions in an environment so as to maximise some notion of cumulative reward.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Machine_learning", "content": "Reinforcement learning is an area of machine learning concerned with how software agents ought to take actions in an environment so as to maximise some notion of cumulative reward."}
+{"idx": 1, "title": "Active Learning Definition | DeepAI", "date": "", "ddg_snippet": "Active learning is a form of semi-supervised machine learning where the algorithm chooses which data to learn from and queries a teacher for guidance.", "subpage_snippet": "", "source": "deepai.org", "link": "https://deepai.org/machine-learning-glossary-and-terms/active-learning", "content": "Active learning is a form of semi-supervised machine learning where the algorithm chooses which data to learn from and queries a teacher for guidance."}
+{"idx": 2, "title": "Active Learning for AI: How Machines Learn to Learn", "date": "", "ddg_snippet": "Active learning is an innovative practice in the world of data that allows machines to learn on their own. It’s a different path from traditional, supervised machine learning algorithms that learn passively.By definition , active learning is the exact opposite of passive learning .", "subpage_snippet": "", "source": "www.linkedin.com", "link": "https://www.linkedin.com/pulse/active-learning-ai-how-machines-learn-tiffany-perkins-munn-ph-d-", "content": "Active learning is an innovative practice in the world of data that allows machines to learn on their own. It’s a different path from traditional, supervised machine learning algorithms that learn passively.By definition , active learning is the exact opposite of passive learning ."}
+{"idx": 3, "title": "Active Learning : Definition , Benefits, and Uses | Ultralytics", "date": "", "ddg_snippet": "Discover active learning , a cost-effective machine learning method that boosts accuracy with fewer labels. Learn how it transforms AI training!", "subpage_snippet": "", "source": "www.ultralytics.com", "link": "https://www.ultralytics.com/glossary/active-learning", "content": "Discover active learning , a cost-effective machine learning method that boosts accuracy with fewer labels. Learn how it transforms AI training!"}
+{"idx": 4, "title": "Active Learning in Machine Learning [Guide & Examples]", "date": "", "ddg_snippet": "Active learning is a type of machine learning where the model is trained on only the most relevant data. Explore the benefits and limitations of the framework.", "subpage_snippet": "", "source": "www.v7labs.com", "link": "https://www.v7labs.com/blog/active-learning-guide", "content": "Active learning is a type of machine learning where the model is trained on only the most relevant data. Explore the benefits and limitations of the framework."}
+{"idx": 5, "title": "Active Learning … Learning with the model | by Anjana Yadav | Medium", "date": "", "ddg_snippet": "Active Learning is a semi-supervised learning where we effectively select the most important data points that can efficiently represent the complete population distribution. Thus along with the confidence of the model we can also check for similarities in the data.", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/analytics-vidhya/active-learning-learning-with-the-model-5727d2474a90", "content": "Active Learning is a semi-supervised learning where we effectively select the most important data points that can efficiently represent the complete population distribution. Thus along with the confidence of the model we can also check for similarities in the data."}
+{"idx": 6, "title": "Weakly supervised learning", "date": "", "ddg_snippet": "Active Learning (principle). Definition : • Active Component: ask queries to an oracle • Improve the performance of a classier • Minimizing the cost of obtaining labeled. data.", "subpage_snippet": "", "source": "egc2020.sciencesconf.org", "link": "https://egc2020.sciencesconf.org/data/pages/e_EGC_2020_VLemaire.pdf", "content": "Active Learning (principle). Definition : • Active Component: ask queries to an oracle • Improve the performance of a classier • Minimizing the cost of obtaining labeled. data."}
+{"idx": 7, "title": "On Interactive Machine Learning – Gaurav Trivedi", "date": "", "ddg_snippet": "Active Learning : Active learning algorithms try to optimize for the number of training examples. Such an algorithm would ask an oracle to give labels such that it can achieve higher accuracy with smallest number of queries.", "subpage_snippet": "", "source": "www.trivedigaurav.com", "link": "https://www.trivedigaurav.com/blog/on-interactive-machine-learning/", "content": "Active Learning : Active learning algorithms try to optimize for the number of training examples. Such an algorithm would ask an oracle to give labels such that it can achieve higher accuracy with smallest number of queries."}
+{"idx": 8, "title": "LV 706.315 Interactive Machine Learning (iML) - human-centered.ai", "date": "", "ddg_snippet": "Definition of interactive Machine Learning (iML): We define iML-approaches as algorithms that can interact with both computational and human agents *) and can optimize its learning behaviour trough this interaction. *) Such agents are called in Active Learning “oracles” (see e.g.: Settles...", "subpage_snippet": "", "source": "human-centered.ai", "link": "https://human-centered.ai/lv-706-315-interactive-machine-learning/", "content": "Definition of interactive Machine Learning (iML): We define iML-approaches as algorithms that can interact with both computational and human agents *) and can optimize its learning behaviour trough this interaction. *) Such agents are called in Active Learning “oracles” (see e.g.: Settles..."}
+{"idx": 9, "title": "1st Lecture Prof. Massimo Vergassola - Active learning : definition ...", "date": "", "ddg_snippet": "Please find below more details. A mini-course on decision-making and active learning Machine Learning has come of age and is found everywhere, in applications and basic sciences alike.Lecture 1. Active learning : definition and examples - 21st of May, 14h30-16h00.", "subpage_snippet": "", "source": "www-en.fisica.uniroma2.it", "link": "https://www-en.fisica.uniroma2.it/events/1st-lecture-prof-massimo-vergassola-active-learning-definition-and-examples/", "content": "Please find below more details. A mini-course on decision-making and active learning Machine Learning has come of age and is found everywhere, in applications and basic sciences alike.Lecture 1. Active learning : definition and examples - 21st of May, 14h30-16h00."}

data/sampled_jsons/Algorithm_2_Tanh_Demon_adaptive_temperature_τ_=_c__std(R)_arXiv_2410.05760.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Training-free Diffusion Model Alignment with Sampling Demons", "date": "", "ddg_snippet": "Algorithm 2 Tanh Demon with Adaptive Temperature . 1: Input: A list of ODE reward estimate.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.05760v1", "content": "Algorithm 2 Tanh Demon with Adaptive Temperature . 1: Input: A list of ODE reward estimate."}
+{"idx": 1, "title": "6 The Metropolis algorithm — Applying Maths in the Chemical...", "date": "", "ddg_snippet": "6. 2 Basis of the Algorithm . 6.3 Average energy, heat capacity and displacement of a harmonic oscillator. 6.4 Maxwell distribution.", "subpage_snippet": "", "source": "applying-maths-book.com", "link": "https://applying-maths-book.com/chapter-12/monte-carlo-D.html", "content": "6. 2 Basis of the Algorithm . 6.3 Average energy, heat capacity and displacement of a harmonic oscillator. 6.4 Maxwell distribution."}
+{"idx": 2, "title": "Чем отличаются поршня STD A, STD B, STD C ? | Форум", "date": "", "ddg_snippet": "Программа показывает три типа стандартных поршней STD A, STD B, и STD C . Картинку я видел.Коротыш писал(а): Ставь STD B. Почему именно В, обоснуй !?", "subpage_snippet": "", "source": "pajero4x4.ru", "link": "https://pajero4x4.ru/bbs/phpBB2/viewtopic.php?t=130503", "content": "Программа показывает три типа стандартных поршней STD A, STD B, и STD C . Картинку я видел.Коротыш писал(а): Ставь STD B. Почему именно В, обоснуй !?"}
+{"idx": 3, "title": "Функция getline в C ++ — чтение строк с ввода - Советы от...", "date": "", "ddg_snippet": "В C ++ есть несколько вариантов функции getline, и каждый имеет свои особенности. Основные версии: std ::getline для std ::string — самый популярный вариант для работы со строками.", "subpage_snippet": "", "source": "arenda-server.cloud", "link": "https://arenda-server.cloud/blog/funkcija-getline-v-c-chtenie-strok-s-vvoda/", "content": "В C ++ есть несколько вариантов функции getline, и каждый имеет свои особенности. Основные версии: std ::getline для std ::string — самый популярный вариант для работы со строками."}
+{"idx": 4, "title": "C ++ | Минимальный и максимальный элементы", "date": "", "ddg_snippet": "Функции std ::min_element и std ::max_element возвращают минимальный и макисмальный элементы соответственно из некоторого диапазона. В качестве коллекции элементов может выступать контейнер или массив.", "subpage_snippet": "", "source": "metanit.com", "link": "https://metanit.com/cpp/tutorial/16.1.php", "content": "Функции std ::min_element и std ::max_element возвращают минимальный и макисмальный элементы соответственно из некоторого диапазона. В качестве коллекции элементов может выступать контейнер или массив."}
+{"idx": 5, "title": "LSTM - нейронная сеть с долгой краткосрочной памятью", "date": "", "ddg_snippet": "Все рекуррентные нейронные сети имеют форму цепочки повторяющихся модулей нейронной сети. В стандартных РНС этот повторяющийся модуль имеет простую структуру, например, один слой tanh . пример lSTM.", "subpage_snippet": "", "source": "neurohive.io", "link": "https://neurohive.io/ru/osnovy-data-science/lstm-nejronnaja-set/", "content": "Все рекуррентные нейронные сети имеют форму цепочки повторяющихся модулей нейронной сети. В стандартных РНС этот повторяющийся модуль имеет простую структуру, например, один слой tanh . пример lSTM."}
+{"idx": 6, "title": "How to use min() function in C ++ - CodeSpeedy", "date": "", "ddg_snippet": "#include #include using namespace std ; int main() {.« C ++ program to convert temperature from Celsius to Kelvin.", "subpage_snippet": "", "source": "www.codespeedy.com", "link": "https://www.codespeedy.com/how-to-use-min-function-in-cpp/", "content": "#include #include using namespace std ; int main() {.« C ++ program to convert temperature from Celsius to Kelvin."}
+{"idx": 7, "title": "Строки и символы в C / C ++ и Arduino. Урок | AlexGyver Technologies", "date": "", "ddg_snippet": "В Си есть стандартная библиотека string.h с функциями для работы со строками, она разобрана в справочнике. Работать с ней не рекомендуется - это неудобно и небезопасно. Используйте более высокоуровневые инструменты, такие как String (Arduino) или std ::string ( C ++...", "subpage_snippet": "", "source": "alexgyver.ru", "link": "https://alexgyver.ru/lessons/cstring/", "content": "В Си есть стандартная библиотека string.h с функциями для работы со строками, она разобрана в справочнике. Работать с ней не рекомендуется - это неудобно и небезопасно. Используйте более высокоуровневые инструменты, такие как String (Arduino) или std ::string ( C ++..."}
+{"idx": 8, "title": "Точки пересечения графика функции с осью | Онлайн калькулятор", "date": "", "ddg_snippet": "\\operatorname{th} x : tanh [x] или Tanh [x]. \\operatorname{cth} x : coth[x] или Coth[x]. \\operatorname{sech} x : sech[x] или Sech[x].", "subpage_snippet": "", "source": "allcalc.ru", "link": "https://allcalc.ru/node/676", "content": "\\operatorname{th} x : tanh [x] или Tanh [x]. \\operatorname{cth} x : coth[x] или Coth[x]. \\operatorname{sech} x : sech[x] или Sech[x]."}
+{"idx": 9, "title": "Калькулятор тригонометрических уравнений- бесплатный онлайн...", "date": "", "ddg_snippet": "sinh. cosh. tanh . coth.", "subpage_snippet": "", "source": "ru.symbolab.com", "link": "https://ru.symbolab.com/solver/trigonometric-equation-calculator", "content": "sinh. cosh. tanh . coth."}

data/sampled_jsons/Bansal_2010_constructive_algorithms_discrepancy_minimization_contribution.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Constructive Algorithms for Discrepancy Minimization", "date": "", "ddg_snippet": "by N Bansal · 2010 · Cited by 220 — In this paper we give the first polynomial time algorithms for discrepancy minimization that achieve bounds similar to those known existentially ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1002.2259", "content": "by N Bansal · 2010 · Cited by 220 — In this paper we give the first polynomial time algorithms for discrepancy minimization that achieve bounds similar to those known existentially ..."}
+{"idx": 1, "title": "Constructive ℓ₂-Discrepancy Minimization with Additive ...", "date": "", "ddg_snippet": "8 Sept 2025 — Our algorithm is based on the framework on Bansal and Garg, together with a new analysis involving ( i ) (i) additional linear and spectral ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.21423v2", "content": "8 Sept 2025 — Our algorithm is based on the framework on Bansal and Garg, together with a new analysis involving ( i ) (i) additional linear and spectral ..."}
+{"idx": 2, "title": "ICML Poster Discrepancy Minimization in Input-Sparsity Time", "date": "", "ddg_snippet": "Bansal's algorithm has approximation guarantees in terms of the hereditary discrepancy (Lov´asz et al., 1986). ... where A is the set of all matrices obtained ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/45157", "content": "Bansal's algorithm has approximation guarantees in terms of the hereditary discrepancy (Lov´asz et al., 1986). ... where A is the set of all matrices obtained ..."}
+{"idx": 3, "title": "Discrepancy Minimization in Input-Sparsity Time", "date": "", "ddg_snippet": "by Y Deng · Cited by 31 — This paper proposes an improved randomized algorithm for discrepancy minimization problem for real-valued matrices m*nmatrices A with m = poly(n).", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=TmJvacopmV", "content": "by Y Deng · Cited by 31 — This paper proposes an improved randomized algorithm for discrepancy minimization problem for real-valued matrices m*nmatrices A with m = poly(n)."}
+{"idx": 4, "title": "FOCS 2010 Accepted Papers - Stanford CS Theory", "date": "", "ddg_snippet": "In this paper we give the first polynomial time algorithms for discrepancy minimization , that achieve bounds similar to those known existentially using the so- ...", "subpage_snippet": "", "source": "theory.stanford.edu", "link": "http://theory.stanford.edu/focs2010/accabs.html", "content": "In this paper we give the first polynomial time algorithms for discrepancy minimization , that achieve bounds similar to those known existentially using the so- ..."}
+{"idx": 5, "title": "Discrepancy Minimization via a Self-Balancing Walk", "date": "", "ddg_snippet": "10 Jan 2024 — 4. N. Bansal, Constructive algorithms for discrepancy minimization , in 51th Annual IEEE Symposium on Foundations of Computer Science, FOCS 2010, ...", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/10.1137/21M1442450", "content": "10 Jan 2024 — 4. N. Bansal, Constructive algorithms for discrepancy minimization , in 51th Annual IEEE Symposium on Foundations of Computer Science, FOCS 2010, ..."}
+{"idx": 6, "title": "Constructive Discrepancy Minimization for Convex Sets", "date": "", "ddg_snippet": "by T Rothvoß · Cited by 110 — Bansal , “ Constructive algorithms for discrepancy mini- mization,” in FOCS, pp. 3–10, 2010 . [6] S. Lovett and R. Meka, “ Constructive discrepancy minimiza-.", "subpage_snippet": "", "source": "ieee-focs.org", "link": "https://ieee-focs.org/FOCS-2014-Papers/6517a140.pdf", "content": "by T Rothvoß · Cited by 110 — Bansal , “ Constructive algorithms for discrepancy mini- mization,” in FOCS, pp. 3–10, 2010 . [6] S. Lovett and R. Meka, “ Constructive discrepancy minimiza-."}
+{"idx": 7, "title": "Semidefinite optimization in discrepancy theory", "date": "", "ddg_snippet": "by N Bansal · 2012 · Cited by 8 — Abstract Recently, there have been several new developments in discrepancy theory based on connections to semidefinite programming.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/s10107-012-0546-7.pdf", "content": "by N Bansal · 2012 · Cited by 8 — Abstract Recently, there have been several new developments in discrepancy theory based on connections to semidefinite programming."}
+{"idx": 8, "title": "Fast Discrepancy Minimization with Hereditary Guarantees", "date": "", "ddg_snippet": "by K Green Larsen · 2023 · Cited by 3 — Giving such an algorithm is the main focus of this work. Our Contribution . In this work, we give a fast algorithm for finding low- discrepancy colorings when the.", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/pdf/10.1137/1.9781611977554.ch11", "content": "by K Green Larsen · 2023 · Cited by 3 — Giving such an algorithm is the main focus of this work. Our Contribution . In this work, we give a fast algorithm for finding low- discrepancy colorings when the."}
+{"idx": 9, "title": "Discrepancy Without Partial Colorings", "date": "", "ddg_snippet": "by NJA Harvey · Cited by 17 — [4] Nikhil Bansal. Constructive algorithms for discrepancy minimization . In Proceedings of the 51st. Annual IEEE Symposium on Foundations of Computer Science ( ...", "subpage_snippet": "", "source": "www2.isye.gatech.edu", "link": "https://www2.isye.gatech.edu/~msingh94/publications/discrepancy1.pdf", "content": "by NJA Harvey · Cited by 17 — [4] Nikhil Bansal. Constructive algorithms for discrepancy minimization . In Proceedings of the 51st. Annual IEEE Symposium on Foundations of Computer Science ( ..."}

data/sampled_jsons/DART_CVPR_2025_Table_2_MIMIC-CXR_F1_score_0.469.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "DART: Disease-aware Image-Text Alignment and Self-correcting ...", "date": "", "ddg_snippet": "Table 2 presents a comparison of the clinical eficacy (CE) metrics between our proposed framework and state-of-the-art meth-ods on the MIMIC-CXR dataset, evaluated by F1 score , pre-cision, and recall.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Park_DART_Disease-aware_Image-Text_Alignment_and_Self-correcting_Re-alignment_for_Trustworthy_Radiology_CVPR_2025_paper.pdf", "content": "Table 2 presents a comparison of the clinical eficacy (CE) metrics between our proposed framework and state-of-the-art meth-ods on the MIMIC-CXR dataset, evaluated by F1 score , pre-cision, and recall."}
+{"idx": 1, "title": "CVPR 2025 Statistics - Paper Copilot", "date": "", "ddg_snippet": "For example, if a paper received ratings of 3, 4, and 5, its average score is 4 — and this average is used in the distribution. Suppose the Accept tier contains submissions with reviewer averages: {4.0, 3.1, 3.6}.", "subpage_snippet": "", "source": "papercopilot.com", "link": "https://papercopilot.com/statistics/cvpr-statistics/cvpr-2025-statistics/", "content": "For example, if a paper received ratings of 3, 4, and 5, its average score is 4 — and this average is used in the distribution. Suppose the Accept tier contains submissions with reviewer averages: {4.0, 3.1, 3.6}."}
+{"idx": 2, "title": "Visual Question Answering evaluation dataset for MIMIC CXR", "date": "", "ddg_snippet": "Jan 28, 2025 · This dataset provides 224 VQAs for 40 test set cases, and 111 VQAs for 23 validation set cases of the MIMIC CXR dataset.", "subpage_snippet": "", "source": "physionet.org", "link": "https://physionet.org/content/vqa-evaluation-mimic-cxr/", "content": "Jan 28, 2025 · This dataset provides 224 VQAs for 40 test set cases, and 111 VQAs for 23 validation set cases of the MIMIC CXR dataset."}
+{"idx": 3, "title": "Under review as a conference paper at ICLR 2025", "date": "", "ddg_snippet": "Table 2 : Comparison of our model with other baselines on MIMIC -R3G-test-A. B@n, M, R-L represent the NLG metrics BLEU, METEOR, and ROUGE-L respectively. P, R, F1 represent the CE metrics CheXpert precision, recall, and F1 - score respectively.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=i4pGIOlH8l", "content": "Table 2 : Comparison of our model with other baselines on MIMIC -R3G-test-A. B@n, M, R-L represent the NLG metrics BLEU, METEOR, and ROUGE-L respectively. P, R, F1 represent the CE metrics CheXpert precision, recall, and F1 - score respectively."}
+{"idx": 4, "title": "CVPR 2025 Open Access Repository", "date": "", "ddg_snippet": "In this study, we propose a Disease-aware image-text Alignment and self-correcting Re-alignment for Trustworthy radiology report generation ( DART ) framework. In the first stage, we generate initial reports based on image-to-text retrieval with disease-matching, embedding both images and texts in a shared embedding space through contrastive ...", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/html/Park_DART_Disease-aware_Image-Text_Alignment_and_Self-correcting_Re-alignment_for_Trustworthy_Radiology_CVPR_2025_paper.html", "content": "In this study, we propose a Disease-aware image-text Alignment and self-correcting Re-alignment for Trustworthy radiology report generation ( DART ) framework. In the first stage, we generate initial reports based on image-to-text retrieval with disease-matching, embedding both images and texts in a shared embedding space through contrastive ..."}
+{"idx": 5, "title": "Paper Digest: CVPR 2025 Papers & Highlights", "date": "", "ddg_snippet": "Jun 7, 2025 · Note: CVPR - 2025 accepts more than 2,800 papers, this page only includes 500 of them selected by our daily paper digest algorithm. Interested users can choose to read All 2,800 CVPR - 2025 papers in a separate page. To search for papers presented at CVPR - 2025 on a specific topic, please make use of the search by venue ( CVPR - 2025 ) service.", "subpage_snippet": "", "source": "resources.paperdigest.org", "link": "https://resources.paperdigest.org/2025/06/cvpr-2025-papers-highlights/", "content": "Jun 7, 2025 · Note: CVPR - 2025 accepts more than 2,800 papers, this page only includes 500 of them selected by our daily paper digest algorithm. Interested users can choose to read All 2,800 CVPR - 2025 papers in a separate page. To search for papers presented at CVPR - 2025 on a specific topic, please make use of the search by venue ( CVPR - 2025 ) service."}
+{"idx": 6, "title": "Intelligent Information and Database Systems", "date": "", "ddg_snippet": "ACIIDS 2025 took place in Kitakyushu, Japan, from April 23 to April 25, 2025 . The ACIIDS conference series has a long history. The initial two ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/content/pdf/10.1007/978-981-96-6005-6.pdf", "content": "ACIIDS 2025 took place in Kitakyushu, Japan, from April 23 to April 25, 2025 . The ACIIDS conference series has a long history. The initial two ..."}
+{"idx": 7, "title": "Innovation in Medicine and Healthcare: Proceedings of ...", "date": "", "ddg_snippet": "Table 2 shows the mean SUS score for each mode. The questionnaire revealed that participants preferred Dialogue Mode, despite the fact that Discursive Mode ...", "subpage_snippet": "", "source": "dokumen.pub", "link": "https://dokumen.pub/innovation-in-medicine-and-healthcare-proceedings-of-11th-kes-inmed-2023-9819933102-9789819933105.html", "content": "Table 2 shows the mean SUS score for each mode. The questionnaire revealed that participants preferred Dialogue Mode, despite the fact that Discursive Mode ..."}
+{"idx": 8, "title": "Computer Vision – ECCV 2022 Workshops: Tel Aviv, Israel ...", "date": "", "ddg_snippet": "Examining the results ( Table 2 and Fig. 2 ), we make the following observations. (1) In Table 2 , our baseline (BASE) exhibits inconsistent performance ...", "subpage_snippet": "", "source": "dokumen.pub", "link": "https://dokumen.pub/computer-vision-eccv-2022-workshops-tel-aviv-israel-october-2327-2022-proceedings-part-iv-3031250680-9783031250682.html", "content": "Examining the results ( Table 2 and Fig. 2 ), we make the following observations. (1) In Table 2 , our baseline (BASE) exhibits inconsistent performance ..."}
+{"idx": 9, "title": "Two-stage Conditional Chest X-ray Radiology Report Generation", "date": "", "ddg_snippet": "Rows 16 and 27 in Table 1 and row 8 in Table 2 present our best results with a dual encoder of Vision Transformer (pre-trained from CLIP [26]) and a Bio Clinical BERT [ 2 ] fine-tuned on MIMIC-CXR using contrastive loss.", "subpage_snippet": "", "source": "www.cse.cuhk.edu.hk", "link": "https://www.cse.cuhk.edu.hk/~qdou/public/medneurips2022/64.pdf", "content": "Rows 16 and 27 in Table 1 and row 8 in Table 2 present our best results with a dual encoder of Vision Transformer (pre-trained from CLIP [26]) and a Bio Clinical BERT [ 2 ] fine-tuned on MIMIC-CXR using contrastive loss."}

data/sampled_jsons/DART_Disease-aware_Image-Text_Alignment_Self-correcting_Re-alignment_Equation_(7)_correction_loss_year_2023.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Vacuum hose diagram question | Dodge Dart Forum", "date": "", "ddg_snippet": "Jun 7 , 2025 · Having issues with my 2013 dodge dart 1.4l. Keep getting underboost code and a sligh shudder when turbo kicks in. New to this type of vehicle and not at all familiar with turbos in general. Someone suggested I replace the boost selenoid and they took it off. Got a new one but now have no idea...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/vacuum-hose-diagram-question.71323/", "content": "Jun 7 , 2025 · Having issues with my 2013 dodge dart 1.4l. Keep getting underboost code and a sligh shudder when turbo kicks in. New to this type of vehicle and not at all familiar with turbos in general. Someone suggested I replace the boost selenoid and they took it off. Got a new one but now have no idea..."}
+{"idx": 1, "title": "Recall Shifter Bushing Replacement - 2013 Dart", "date": "", "ddg_snippet": "Jul 20, 2024 · 2013 1.4 dart manual, 87k miles, and accidentally discovered that my bushing was holding on by a thread while investigating oil on top of my transmission (looks like my vaccuum pump gasket is rotting away).", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/recall-shifter-bushing-replacement-2013-dart.70634/", "content": "Jul 20, 2024 · 2013 1.4 dart manual, 87k miles, and accidentally discovered that my bushing was holding on by a thread while investigating oil on top of my transmission (looks like my vaccuum pump gasket is rotting away)."}
+{"idx": 2, "title": "2.4 multiair variable valve timing actuator - Dodge Dart Forum", "date": "", "ddg_snippet": "Nov 9, 2019 · Dodge- Dart .org is a forum dedicated to the 2013-2016 Dodge Dart . Join and participate in discussions about maintenance, performance mods, and mechanical issues. Get the latest tips, news, browse the classifieds, and more!", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/2-4-multiair-variable-valve-timing-actuator.65832/", "content": "Nov 9, 2019 · Dodge- Dart .org is a forum dedicated to the 2013-2016 Dodge Dart . Join and participate in discussions about maintenance, performance mods, and mechanical issues. Get the latest tips, news, browse the classifieds, and more!"}
+{"idx": 3, "title": "No Crank, No Start - Dodge Dart Forum", "date": "", "ddg_snippet": "Feb 16, 2023 · Y'all helped me spectacularly last time I had issues with my Dart I figured I'd come back for round 2. So since the last time I asked for help my Dart has started refusing to start. No crank, a singular click, and nothing more. The Dashboard comes to life, the AC works fine, and the Radio works...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/no-crank-no-start.69518/", "content": "Feb 16, 2023 · Y'all helped me spectacularly last time I had issues with my Dart I figured I'd come back for round 2. So since the last time I asked for help my Dart has started refusing to start. No crank, a singular click, and nothing more. The Dashboard comes to life, the AC works fine, and the Radio works..."}
+{"idx": 4, "title": "UConnect Bluetooth Issue | Dodge Dart Forum", "date": "", "ddg_snippet": "Jan 2, 2025 · I have attached some images of the messages I've been getting on my phone when trying to connect, these messages appear just after I tap uconnect on my phone. I have tried many things already such as the temperature button soft reset procedure and the corner of the screen soft reset procedure...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/uconnect-bluetooth-issue.71014/", "content": "Jan 2, 2025 · I have attached some images of the messages I've been getting on my phone when trying to connect, these messages appear just after I tap uconnect on my phone. I have tried many things already such as the temperature button soft reset procedure and the corner of the screen soft reset procedure..."}
+{"idx": 5, "title": "Coolant hose connector - Dodge Dart Forum", "date": "", "ddg_snippet": "May 10, 2024 · Dodge- Dart .org is a forum dedicated to the 2013-2016 Dodge Dart . Join and participate in discussions about maintenance, performance mods, and mechanical issues. Get the latest tips, news, browse the classifieds, and more!", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/coolant-hose-connector.70491/", "content": "May 10, 2024 · Dodge- Dart .org is a forum dedicated to the 2013-2016 Dodge Dart . Join and participate in discussions about maintenance, performance mods, and mechanical issues. Get the latest tips, news, browse the classifieds, and more!"}
+{"idx": 6, "title": "Transmission Shudder - Dodge Dart Forum", "date": "", "ddg_snippet": "Aug 15, 2018 · 2016 dart gt blacktop. 10k miles. First owner. This has been going on for almost a year now. At about 2000rpm (coasting) if I lightly press the throttle the entire car will jerk violently, the feeling can be compared to almost stalling out a manual transmission. Most the time it's when it...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/transmission-shudder.62830/", "content": "Aug 15, 2018 · 2016 dart gt blacktop. 10k miles. First owner. This has been going on for almost a year now. At about 2000rpm (coasting) if I lightly press the throttle the entire car will jerk violently, the feeling can be compared to almost stalling out a manual transmission. Most the time it's when it..."}
+{"idx": 7, "title": "Dart Wiring Diagrams | Page 5 | Dodge Dart Forum", "date": "", "ddg_snippet": "Jan 28, 2024 · Search Wiring Diagrams Use the following link to search for wiring diagrams for the dart .", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/dart-wiring-diagrams.50274/page-5", "content": "Jan 28, 2024 · Search Wiring Diagrams Use the following link to search for wiring diagrams for the dart ."}
+{"idx": 8, "title": "Front strut replacement - Dodge Dart Forum", "date": "", "ddg_snippet": "May 24, 2020 · Front struts are gone, will replace with KYB's struts part #'s 334982 and 334981. Rockauto.com has them for $83 each rightnow compared to the next cheapest autozone $127 so definitely gonna pull the trigger, however I need to know, do I need to replace the strut mounts with KYB's also (sm5811...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/front-strut-replacement.66659/", "content": "May 24, 2020 · Front struts are gone, will replace with KYB's struts part #'s 334982 and 334981. Rockauto.com has them for $83 each rightnow compared to the next cheapest autozone $127 so definitely gonna pull the trigger, however I need to know, do I need to replace the strut mounts with KYB's also (sm5811..."}
+{"idx": 9, "title": "Recirculation Door Actuator Location - Dodge Dart Forum", "date": "", "ddg_snippet": "Jun 10, 2024 · The recirculation door actuator (1) is a reversible 12 volt Direct Current (DC) servo motor. The recirculation door actuator is located on the bottom of the HVAC air inlet housing, behind the instrument panel. The recirculation door actuator is contained within a black molded plastic housing with an integral wire connector receptacle (4). Three mounting tabs (3) allow the actuator to be ...", "subpage_snippet": "", "source": "www.dodge-dart.org", "link": "https://www.dodge-dart.org/threads/recirculation-door-actuator-location.70551/", "content": "Jun 10, 2024 · The recirculation door actuator (1) is a reversible 12 volt Direct Current (DC) servo motor. The recirculation door actuator is located on the bottom of the HVAC air inlet housing, behind the instrument panel. The recirculation door actuator is contained within a black molded plastic housing with an integral wire connector receptacle (4). Three mounting tabs (3) allow the actuator to be ..."}

data/sampled_jsons/Do_Not_Trust_What_They_Tell_Exposing_Malicious_Accomplices_in_Tor_via_Anomalous_Circuit_Detection_fu.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Do Not Trust What They Tell: Exposing Malicious Accomplices ...", "date": "", "ddg_snippet": "Apr 22, 2025 · 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": "Apr 22, 2025 · 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": "Do Not Trust What They Tell: Exposing Malicious Accomplices ...", "date": "", "ddg_snippet": "We propose a technique to determine the position of a con-trolled Middle node within a specific type of circuit by clas-sifying the circuit section that includes the node. Through validation, the accuracy of the circuit section classification reaches 100%.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qcnePVejeV", "content": "We propose a technique to determine the position of a con-trolled Middle node within a specific type of circuit by clas-sifying the circuit section that includes the node. Through validation, the accuracy of the circuit section classification reaches 100%."}
+{"idx": 2, "title": "Defending Against Malicious Cyber Activity Originating from Tor", "date": "", "ddg_snippet": "Malicious cyber actors use Tor to mask their identity when engaging in malicious cyber activity impacting the confidentiality, integrity, and availability of an organization’s information systems and data.", "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": "Malicious cyber actors use Tor to mask their identity when engaging in malicious cyber activity impacting the confidentiality, integrity, and availability of an organization’s information systems and data."}
+{"idx": 3, "title": "dblp: Do Not Trust What They Tell: Exposing Malicious ...", "date": "", "ddg_snippet": "Bibliographic details on Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection .", "subpage_snippet": "", "source": "dblp.org", "link": "https://dblp.org/rec/conf/www/Yao0LDGW25", "content": "Bibliographic details on Do Not Trust What They Tell : Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection ."}
+{"idx": 4, "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...", "date": "", "ddg_snippet": "Mar 13, 2025 · 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中的潜在恶意同谋节点。 原文标题： Do Not Trust What They Tell : Expo sing Mali cious Acc ompli ces in Tor via Ano mal ous Circuit Det ection . 1. 论文内容概述.", "subpage_snippet": "", "source": "www.secrss.com", "link": "https://www.secrss.com/articles/76608", "content": "Mar 13, 2025 · 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中的潜在恶意同谋节点。 原文标题： Do Not Trust What They Tell : Expo sing Mali cious Acc ompli ces in Tor via Ano mal ous Circuit Det ection . 1. 论文内容概述."}
+{"idx": 5, "title": "Spoiled Onions: Exposing Malicious Tor Exit Relays", "date": "", "ddg_snippet": "In this paper, we seek to expose such malicious exit relays and document their actions. First, we monitored the Tor network after developing two fast and modular exit relay scanners—one for credential sniffing and one for active MitM attacks.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/chapter/10.1007/978-3-319-08506-7_16", "content": "In this paper, we seek to expose such malicious exit relays and document their actions. First, we monitored the Tor network after developing two fast and modular exit relay scanners—one for credential sniffing and one for active MitM attacks."}
+{"idx": 6, "title": "Detecting Malicious Users Behind Circuit-Based Anonymity ...", "date": "", "ddg_snippet": "Nov 13, 2020 · Abstract: This project addresses the issue of detecting intruders from hiding behind privacy-protecting anonymity networks. The freely available Tor and the SOCKS proxy services have been popular tools that provide circuit -based anonymous connections to network users.", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/9258912", "content": "Nov 13, 2020 · Abstract: This project addresses the issue of detecting intruders from hiding behind privacy-protecting anonymity networks. The freely available Tor and the SOCKS proxy services have been popular tools that provide circuit -based anonymous connections to network users."}
+{"idx": 7, "title": "Ming Yang - Home - ACM Digital Library", "date": "", "ddg_snippet": "Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection ... full texts may be downloaded from an ACM full - text ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/profile/81318492158", "content": "Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection ... full texts may be downloaded from an ACM full - text ..."}
+{"idx": 8, "title": "ACM TheWebConf 2025 Conference", "date": "", "ddg_snippet": "... Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection · pdf icon. Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/group?id=ACM.org/TheWebConf/2025/Conference", "content": "... Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor via Anomalous Circuit Detection · pdf icon. Yixuan Yao, Ming Yang, Zixia Liu, Kai Dong ..."}
+{"idx": 9, "title": "Paper Digest: WWW 2025 Papers & Highlights", "date": "", "ddg_snippet": "30 Apr 2025 — Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor Via Anomalous Circuit Detection ... text information. Wei Li; Jiawen ...", "subpage_snippet": "", "source": "www.paperdigest.org", "link": "https://www.paperdigest.org/2025/04/www-2025-papers-highlights/", "content": "30 Apr 2025 — Do Not Trust What They Tell: Exposing Malicious Accomplices in Tor Via Anomalous Circuit Detection ... text information. Wei Li; Jiawen ..."}

data/sampled_jsons/E91gjsccP1_HtmlRAG_Beautiful_Soup_[50]_primary_function_year_2023-2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Облегчаем себе жизнь с помощью BeautifulSoup 4 / Хабр", "date": "", "ddg_snippet": "Переходим в редактор кода и импортируем наши библиотеки: from bs4 import BeautifulSoup import requests.Самое время воспользоваться BeautifulSoup 4 и скормить ему наш page, указав в кавычках как он нам поможет ' html .parcer'", "subpage_snippet": "", "source": "habr.com", "link": "https://habr.com/ru/articles/544828/", "content": "Переходим в редактор кода и импортируем наши библиотеки: from bs4 import BeautifulSoup import requests.Самое время воспользоваться BeautifulSoup 4 и скормить ему наш page, указав в кавычках как он нам поможет ' html .parcer'"}
+{"idx": 1, "title": "BeautifulSoup - уроки по парсингу сайтов в Python", "date": "", "ddg_snippet": "Парсинг сайтов при помощи библиотеки BeautifulSoup . Уроки как при помощи BeautifulSoup менять HTML элементы.", "subpage_snippet": "", "source": "python-scripts.com", "link": "https://python-scripts.com/beautifulsoup-parsing", "content": "Парсинг сайтов при помощи библиотеки BeautifulSoup . Уроки как при помощи BeautifulSoup менять HTML элементы."}
+{"idx": 2, "title": "Implementing Web Scraping in Python with BeautifulSoup", "date": "", "ddg_snippet": "Jul 26, 2025 · BeautifulSoup is a Python library used for web scraping. It helps parse HTML and XML documents making it easy to navigate and extract specific parts of a webpage. This article explains the steps of web scraping using BeautifulSoup. Steps involved in web scraping Send an HTTP Request: Use the requests library to send a request to the webpage URL and get the HTML content in response. Parse the ...", "subpage_snippet": "", "source": "www.geeksforgeeks.org", "link": "https://www.geeksforgeeks.org/python/implementing-web-scraping-python-beautiful-soup/", "content": "Jul 26, 2025 · BeautifulSoup is a Python library used for web scraping. It helps parse HTML and XML documents making it easy to navigate and extract specific parts of a webpage. This article explains the steps of web scraping using BeautifulSoup. Steps involved in web scraping Send an HTTP Request: Use the requests library to send a request to the webpage URL and get the HTML content in response. Parse the ..."}
+{"idx": 3, "title": "The Complete BeautifulSoup Cheatsheet with Examples", "date": "", "ddg_snippet": "Oct 4, 2023 · This cheatsheet covers the full BeautifulSoup 4 API with practical examples. It provides a comprehensive guide to web scraping and HTML parsing using Python's BeautifulSoup library.", "subpage_snippet": "", "source": "proxiesapi.com", "link": "https://proxiesapi.com/articles/the-complete-beautifulsoup-cheatsheet-with-examples", "content": "Oct 4, 2023 · This cheatsheet covers the full BeautifulSoup 4 API with practical examples. It provides a comprehensive guide to web scraping and HTML parsing using Python's BeautifulSoup library."}
+{"idx": 4, "title": "Beautiful Soup Documentation — Beautiful Soup 4.13.0 ...", "date": "", "ddg_snippet": "Beautiful Soup is a Python library for parsing HTML and XML documents, offering tools to navigate, search, and modify parse trees.", "subpage_snippet": "", "source": "www.crummy.com", "link": "https://www.crummy.com/software/BeautifulSoup/bs4/doc/", "content": "Beautiful Soup is a Python library for parsing HTML and XML documents, offering tools to navigate, search, and modify parse trees."}
+{"idx": 5, "title": "Beautiful Soup: Python Web Scraping Code Explained | AnSoup", "date": "", "ddg_snippet": "Sep 9, 2025 · Learn web scraping with BeautifulSoup in Python. Understand the code and start extracting data from websites like a pro.", "subpage_snippet": "", "source": "ansoup.com", "link": "https://ansoup.com/article/what-does-beautiful-soup-code-look-like", "content": "Sep 9, 2025 · Learn web scraping with BeautifulSoup in Python. Understand the code and start extracting data from websites like a pro."}
+{"idx": 6, "title": "BeautifulSoup Cheatsheet with Code Samples | ScrapingAnt", "date": "", "ddg_snippet": "Aug 19, 2024 · Explore the core concepts and advanced features of BeautifulSoup with detailed code samples and explanations to help you get started with web scraping and HTML parsing in Python.", "subpage_snippet": "", "source": "scrapingant.com", "link": "https://scrapingant.com/blog/beautifulsoup-cheatsheet", "content": "Aug 19, 2024 · Explore the core concepts and advanced features of BeautifulSoup with detailed code samples and explanations to help you get started with web scraping and HTML parsing in Python."}
+{"idx": 7, "title": "BeautifulSoup Cheat Sheet Python - DEV Community", "date": "", "ddg_snippet": "Oct 14, 2024 · BeautifulSoup Cheat Sheet Python Installtion pip install beautifulsoup4 ... Tagged with python, scrape, beautifulsoup.", "subpage_snippet": "", "source": "dev.to", "link": "https://dev.to/shahidkhans/beautifulsoup-cheat-sheet-python-5ep3", "content": "Oct 14, 2024 · BeautifulSoup Cheat Sheet Python Installtion pip install beautifulsoup4 ... Tagged with python, scrape, beautifulsoup."}
+{"idx": 8, "title": "HtmlRAG : HTML is Better Than Plain Text for Modeling Retrieved...", "date": "", "ddg_snippet": "Soup [ 50 ] to parse the concatenated HTML document to a single.748 Vanilla HTML ; (2) Plain Text: The plain text extracted with an on-. 749 the-self package BeautifulSoup [ 50 ]; (3) Markdown: The Markdown. 750 converted by an on-the-self converter Markdownify [2]. Additional.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=E91gjsccP1", "content": "Soup [ 50 ] to parse the concatenated HTML document to a single.748 Vanilla HTML ; (2) Plain Text: The plain text extracted with an on-. 749 the-self package BeautifulSoup [ 50 ]; (3) Markdown: The Markdown. 750 converted by an on-the-self converter Markdownify [2]. Additional."}
+{"idx": 9, "title": "HtmlRAG : HTML is Better Than Plain Text for Modeling Retrieved...", "date": "", "ddg_snippet": "”. 3.3 Granularity-Adjustable Block Tree Building. To prune all retrieved HTML documents as a whole, we first con-catenate all retrieved HTML documents together, and use Beautiful Soup [59] to parse the concatenated HTML document to a single DOM tree.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2411.02959", "content": "”. 3.3 Granularity-Adjustable Block Tree Building. To prune all retrieved HTML documents as a whole, we first con-catenate all retrieved HTML documents together, and use Beautiful Soup [59] to parse the concatenated HTML document to a single DOM tree."}

data/sampled_jsons/ELITE_paper_Figure_3_filtering_threshold_s_score_2502.04757.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "ELITE: Enhanced Language-Image Toxicity Evaluation for ...", "date": "", "ddg_snippet": "by W Lee · 2025 — 2) Filtering : Integrate only image-text pairs where at least two out of three model responses assign an ELITE evaluator score of 10 or higher. 3 ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2502.04757", "content": "by W Lee · 2025 — 2) Filtering : Integrate only image-text pairs where at least two out of three model responses assign an ELITE evaluator score of 10 or higher. 3 ..."}
+{"idx": 1, "title": "ELITE: Enhanced Language-Image Toxicity Evaluation for Safety Enhancing NLP Model Performance Through Data Filtering Justice Integrity Report - Sept. 2025 News Filtering & evaluation - NGS - variant analysis - GitHub Pages Decoding Digital Labor: A Topic Modeling Analysis of ... - MDPI Confidence Score: The Forgotten Dimension of Object Detection ... Similarity Score for Information Filtering Thresholds", "date": "", "ddg_snippet": "Feb 7, 2025 · Therefore, we found that existing benchmarks have low levels of harmfulness, ambiguous data, and limited diversity in image-text pair combinations. To address these issues, we propose the ELITE benchmark, a high-quality safety evaluation benchmark for VLMs, underpinned by our enhanced evaluation method, the ELITE evaluator. In this paper , we address the challenge of identifying and filtering low-quality text data to improve NLP model finetuning performance. We propose a data cleaning and filtering methodology based on four carefully chosen criteria: relevance, to ensure the texts are closely related to the core characteristics of the dataset; informativeness, to filter out sentences with little valuable ... Editor' s Choice: Scroll below for our monthly blend of mainstream and September 2025 news and views Note: Excerpts are from the authors' words except for subheads and occasional \"Editor' s notes\" such as this. Sept. 17 New York Times, Trump Administration: Trump Administration Live Updates: Kennedy Is Undermining Trust in Public Health, Former C.D.C. Officials Testify, Sheryl Gay Stolberg, Sept ... First, filtering thresholds are usually different for SNPs and INDELs. You can extract all the SNP records in our trio vcf like this: Exercise: Check out the documentation of gatk SelectVariants, and: 1. Figure out what you’ll need to fill in at --select-typeif you want to select only INDELS. 2. Generate a vcf with only the SNPs and a second vcf wi... See full list on sib-swiss.github.io The command gatk VariantFiltration enables you to filter for both the INFO field (per variant) and FORMAT field (per genotype). For now we’re only interested in filtering variants. Below you can find the command to hard- filter the SNP variants on some sensible thresholds (that are explained here). Exercise:Run the filtering command above. Did it af... See full list on sib-swiss.github.io A command with sensible parametersto do a first iteration of hard filtering the INDELs would be: Exercise:Run the command and figure out how many variants are filtered out. See full list on sib-swiss.github.io Now that we have filtered the INDELs and SNPs separately, we can merge them again with this command: Exercise:Run the command to merge the vcfs. See full list on sib-swiss.github.io 4 days ago · The cosine similarity threshold for semantic filtering was set at 0.40, which offered a balanced trade-off between coverage and thematic relevance. Threshold values between 0.35 and 0.45 were additionally tested, and the resulting thematic structure remained stable, confirming the robustness of the chosen parameter. When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ... This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2502.04757", "content": "Feb 7, 2025 · Therefore, we found that existing benchmarks have low levels of harmfulness, ambiguous data, and limited diversity in image-text pair combinations. To address these issues, we propose the ELITE benchmark, a high-quality safety evaluation benchmark for VLMs, underpinned by our enhanced evaluation method, the ELITE evaluator. In this paper , we address the challenge of identifying and filtering low-quality text data to improve NLP model finetuning performance. We propose a data cleaning and filtering methodology based on four carefully chosen criteria: relevance, to ensure the texts are closely related to the core characteristics of the dataset; informativeness, to filter out sentences with little valuable ... Editor' s Choice: Scroll below for our monthly blend of mainstream and September 2025 news and views Note: Excerpts are from the authors' words except for subheads and occasional \"Editor' s notes\" such as this. Sept. 17 New York Times, Trump Administration: Trump Administration Live Updates: Kennedy Is Undermining Trust in Public Health, Former C.D.C. Officials Testify, Sheryl Gay Stolberg, Sept ... First, filtering thresholds are usually different for SNPs and INDELs. You can extract all the SNP records in our trio vcf like this: Exercise: Check out the documentation of gatk SelectVariants, and: 1. Figure out what you’ll need to fill in at --select-typeif you want to select only INDELS. 2. Generate a vcf with only the SNPs and a second vcf wi... See full list on sib-swiss.github.io The command gatk VariantFiltration enables you to filter for both the INFO field (per variant) and FORMAT field (per genotype). For now we’re only interested in filtering variants. Below you can find the command to hard- filter the SNP variants on some sensible thresholds (that are explained here). Exercise:Run the filtering command above. Did it af... See full list on sib-swiss.github.io A command with sensible parametersto do a first iteration of hard filtering the INDELs would be: Exercise:Run the command and figure out how many variants are filtered out. See full list on sib-swiss.github.io Now that we have filtered the INDELs and SNPs separately, we can merge them again with this command: Exercise:Run the command to merge the vcfs. See full list on sib-swiss.github.io 4 days ago · The cosine similarity threshold for semantic filtering was set at 0.40, which offered a balanced trade-off between coverage and thematic relevance. Threshold values between 0.35 and 0.45 were additionally tested, and the resulting thematic structure remained stable, confirming the robustness of the chosen parameter. When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ... This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ..."}
+{"idx": 2, "title": "Enhancing NLP Model Performance Through Data Filtering", "date": "", "ddg_snippet": "In this paper , we address the challenge of identifying and filtering low-quality text data to improve NLP model finetuning performance. We propose a data cleaning and filtering methodology based on four carefully chosen criteria: relevance, to ensure the texts are closely related to the core characteristics of the dataset; informativeness, to filter out sentences with little valuable ...", "subpage_snippet": "", "source": "www2.eecs.berkeley.edu", "link": "https://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-170.pdf", "content": "In this paper , we address the challenge of identifying and filtering low-quality text data to improve NLP model finetuning performance. We propose a data cleaning and filtering methodology based on four carefully chosen criteria: relevance, to ensure the texts are closely related to the core characteristics of the dataset; informativeness, to filter out sentences with little valuable ..."}
+{"idx": 3, "title": "Filtering & evaluation - NGS - variant analysis - GitHub Pages Decoding Digital Labor: A Topic Modeling Analysis of ... - MDPI Confidence Score: The Forgotten Dimension of Object Detection ... Similarity Score for Information Filtering Thresholds", "date": "", "ddg_snippet": "First, filtering thresholds are usually different for SNPs and INDELs. You can extract all the SNP records in our trio vcf like this: Exercise: Check out the documentation of gatk SelectVariants, and: 1. Figure out what you’ll need to fill in at --select-typeif you want to select only INDELS. 2. Generate a vcf with only the SNPs and a second vcf wi... See full list on sib-swiss.github.io The command gatk VariantFiltration enables you to filter for both the INFO field (per variant) and FORMAT field (per genotype). For now we’re only interested in filtering variants. Below you can find the command to hard- filter the SNP variants on some sensible thresholds (that are explained here). Exercise:Run the filtering command above. Did it af... See full list on sib-swiss.github.io A command with sensible parametersto do a first iteration of hard filtering the INDELs would be: Exercise:Run the command and figure out how many variants are filtered out. See full list on sib-swiss.github.io Now that we have filtered the INDELs and SNPs separately, we can merge them again with this command: Exercise:Run the command to merge the vcfs. See full list on sib-swiss.github.io 4 days ago · The cosine similarity threshold for semantic filtering was set at 0.40, which offered a balanced trade-off between coverage and thematic relevance. Threshold values between 0.35 and 0.45 were additionally tested, and the resulting thematic structure remained stable, confirming the robustness of the chosen parameter. When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ... This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ...", "subpage_snippet": "", "source": "sib-swiss.github.io", "link": "https://sib-swiss.github.io/NGS-variants-training/2021.9/day2/filtering_evaluation/", "content": "First, filtering thresholds are usually different for SNPs and INDELs. You can extract all the SNP records in our trio vcf like this: Exercise: Check out the documentation of gatk SelectVariants, and: 1. Figure out what you’ll need to fill in at --select-typeif you want to select only INDELS. 2. Generate a vcf with only the SNPs and a second vcf wi... See full list on sib-swiss.github.io The command gatk VariantFiltration enables you to filter for both the INFO field (per variant) and FORMAT field (per genotype). For now we’re only interested in filtering variants. Below you can find the command to hard- filter the SNP variants on some sensible thresholds (that are explained here). Exercise:Run the filtering command above. Did it af... See full list on sib-swiss.github.io A command with sensible parametersto do a first iteration of hard filtering the INDELs would be: Exercise:Run the command and figure out how many variants are filtered out. See full list on sib-swiss.github.io Now that we have filtered the INDELs and SNPs separately, we can merge them again with this command: Exercise:Run the command to merge the vcfs. See full list on sib-swiss.github.io 4 days ago · The cosine similarity threshold for semantic filtering was set at 0.40, which offered a balanced trade-off between coverage and thematic relevance. Threshold values between 0.35 and 0.45 were additionally tested, and the resulting thematic structure remained stable, confirming the robustness of the chosen parameter. When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ... This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ..."}
+{"idx": 4, "title": "Confidence Score: The Forgotten Dimension of Object Detection ...", "date": "", "ddg_snippet": "When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8271464/", "content": "When deploying a model for object detection, a confidence score threshold is chosen to filter out false positives and ensure that a predicted bounding box has a certain minimum score . To achieve state-of-the-art performance on benchmark datasets, ..."}
+{"idx": 5, "title": "Similarity Score for Information Filtering Thresholds", "date": "", "ddg_snippet": "This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/Similarity-Score-for-Information-Filtering-Lai-Soh/3320f3cdb9654beb72dc2b500c56d562d151172d", "content": "This paper proposes a new algorithm to calculate an index called document similarity score based on elements of the document, which can be used to improve the efficiency of information filtering and retrieval. The rapid growth ofthe on-line information has led to the development of many techniques for information filtering . 7he tremendous growth in the amount of information available and the ..."}
+{"idx": 6, "title": "Enhanced Language-Image Toxicity Evaluation for Safety", "date": "", "ddg_snippet": "24 Jul 2025 — Specifically, on the ELITE evaluator's [0-25] point scale, we set a threshold determined by human judgment. ELITE evaluator score s ≥ 10 s 10 s ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.04757v3", "content": "24 Jul 2025 — Specifically, on the ELITE evaluator's [0-25] point scale, we set a threshold determined by human judgment. ELITE evaluator score s ≥ 10 s 10 s ..."}
+{"idx": 7, "title": "Suhyun Kim", "date": "", "ddg_snippet": "The ELITE evaluator explicitly incorporates a toxicity score to accurately assess harmfulness in multimodal contexts, where VLMs often provide specific, ...", "subpage_snippet": "", "source": "www.scilit.com", "link": "https://www.scilit.com/scholars/5633958", "content": "The ELITE evaluator explicitly incorporates a toxicity score to accurately assess harmfulness in multimodal contexts, where VLMs often provide specific, ..."}
+{"idx": 8, "title": "Frontier Risk Evaluation for National Security and Public Safety", "date": "", "ddg_snippet": "Full results are in Figure 3 . . Second, each adversarial prompt is paired with a benign prompt to help measure over-refusal, balancing safeguard robustness ...", "subpage_snippet": "", "source": "static.scale.com", "link": "https://static.scale.com/uploads/654197dc94d34f66c0f5184e/FORTRESS__Scale_+(5).pdf", "content": "Full results are in Figure 3 . . Second, each adversarial prompt is paired with a benign prompt to help measure over-refusal, balancing safeguard robustness ..."}
+{"idx": 9, "title": "davanstrien/arxiv-cs-2024-sample · Datasets at ...", "date": "", "ddg_snippet": "2502.04757 . ELITE : Enhanced Language-Image Toxicity Evaluation for Safety ... The ELITE evaluator explicitly incorporates a toxicity score to accurately ...", "subpage_snippet": "", "source": "huggingface.co", "link": "https://huggingface.co/datasets/davanstrien/arxiv-cs-2024-sample/viewer", "content": "2502.04757 . ELITE : Enhanced Language-Image Toxicity Evaluation for Safety ... The ELITE evaluator explicitly incorporates a toxicity score to accurately ..."}

data/sampled_jsons/EventPS_Real-Time_Photometric_Stereo_Using_an_Event_Camera_Yu_mean_angular_error_13.66.jsonl

@@ -0,0 +1,4 @@
+{"idx": 0, "title": "EventPS : Real - Time Photometric Stereo Using an Event Camera", "date": "", "ddg_snippet": "Photometric stereo . Assuming an object illuminated by. an ideal distant light source, the radiance of the light source.In this paper, we propose EventPS , a novel real - time PS approach using a single event camera .", "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": "Photometric stereo . Assuming an object illuminated by. an ideal distant light source, the radiance of the light source.In this paper, we propose EventPS , a novel real - time PS approach using a single event camera ."}
+{"idx": 1, "title": "Robust Photometric Stereo Based on Event Interval Profile", "date": "", "ddg_snippet": "... mean angular error of 8.12◦ in the presence of non-Lambertian effects. 2 ... Yu et al. [67] proposed EventPS under the assumption that the movement of ...", "subpage_snippet": "", "source": "cvpr.thecvf.com", "link": "https://cvpr.thecvf.com/virtual/2025/poster/33500", "content": "... mean angular error of 8.12◦ in the presence of non-Lambertian effects. 2 ... Yu et al. [67] proposed EventPS under the assumption that the movement of ..."}
+{"idx": 2, "title": "Cas-FNE: Cascaded Face Normal Estimation", "date": "", "ddg_snippet": "B. Yu , J. Ren, J. Han, F. Wang, J. Liang, and B. Shi, “ EventPS : Real - time photometric stereo using an event camera ,” in Proc. IEEE/CVF Conf.X. Huang and S. Belongie, “Arbitrary style transfer in real - time with adaptive instance normalization,” in Proc. IEEE Int.", "subpage_snippet": "", "source": "www.ieee-jas.net", "link": "https://www.ieee-jas.net/article/doi/10.1109/JAS.2024.124899", "content": "B. Yu , J. Ren, J. Han, F. Wang, J. Liang, and B. Shi, “ EventPS : Real - time photometric stereo using an event camera ,” in Proc. IEEE/CVF Conf.X. Huang and S. Belongie, “Arbitrary style transfer in real - time with adaptive instance normalization,” in Proc. IEEE Int."}
+{"idx": 3, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}

data/sampled_jsons/EventPS_Real-time_photometric_stereo_using_event_camera_Yu_et_al.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 - YouTube", "date": "", "ddg_snippet": "This is the video of the following work: Bohan Yu , Jieji Ren, Jin Han, Feishi Wang, Jinxiu Liang, and Boxin Shi. EventPS : Real-time photometric stereo using ...", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=-TTp0zJKNPU", "content": "This is the video of the following work: Bohan Yu , Jieji Ren, Jin Han, Feishi Wang, Jinxiu Liang, and Boxin Shi. EventPS : Real-time photometric stereo using ..."}
+{"idx": 3, "title": "Yu Bohan from Peking University: EventPS - Real-time Photometric Stereo ...", "date": "", "ddg_snippet": "The paper \" EventPS : Real-Time Photometric Stereo using an Event Camera \" is the first to use the unique properties of event cameras to achieve real-time photometric stereo .", "subpage_snippet": "", "source": "inf.news", "link": "https://inf.news/en/tech/4be4c17d4830518b9004d13cb772438e.html", "content": "The paper \" EventPS : Real-Time Photometric Stereo using an Event Camera \" is the first to use the unique properties of event cameras to achieve real-time photometric stereo ."}
+{"idx": 4, "title": "Event Fusion Photometric Stereo Network - arXiv.org", "date": "", "ddg_snippet": "To alleviate the limitations of the darkroom environment and to use essential light information, we employ an event cam-era with a high dynamic range and low latency. This is the rst study that uses an event camera for the photometric stereo task, which works on con-tinuous light sources and ambient light environment.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2303.00308", "content": "To alleviate the limitations of the darkroom environment and to use essential light information, we employ an event cam-era with a high dynamic range and low latency. This is the rst study that uses an event camera for the photometric stereo task, which works on con-tinuous light sources and ambient light environment."}
+{"idx": 5, "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": 6, "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": 7, "title": "\"EventPS: Real-Time Photometric Stereo Using an Event Camera.\" - dblp", "date": "", "ddg_snippet": "Bibliographic details on EventPS : Real-Time Photometric Stereo Using an Event Camera .", "subpage_snippet": "", "source": "dblp.org", "link": "https://dblp.org/rec/conf/cvpr/YuRHWLS24", "content": "Bibliographic details on EventPS : Real-Time Photometric Stereo Using an Event Camera ."}
+{"idx": 8, "title": "PS-EIP: Robust Photometric Stereo Based on Event Interval Profile", "date": "", "ddg_snippet": "Recently, the energy-efficient photometric stereo method using an event camera ( EventPS [67]) 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": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/document/11095199", "content": "Recently, the energy-efficient photometric stereo method using an event camera ( EventPS [67]) 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": 9, "title": "Track: Orals 3C Medical and Physics-based vision", "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/session/32097", "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."}

data/sampled_jsons/FlowDec_paper_results_FAD_score_FlowDec-75m_7.5_kbps.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "FlowDec : A flow -based full-band general audio codec with high...", "date": "", "ddg_snippet": "FlowDec - 75 m : 75 Hz, multi-bitrate.In the main paper , we showed objective metrics result visually. For completeness, we list the exact numbers of metric values in Table 8. Table 8: Mean ± 95% confidence interval of all metrics shown visually in Fig.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.01485v1", "content": "FlowDec - 75 m : 75 Hz, multi-bitrate.In the main paper , we showed objective metrics result visually. For completeness, we list the exact numbers of metric values in Table 8. Table 8: Mean ± 95% confidence interval of all metrics shown visually in Fig."}
+{"idx": 1, "title": "FlowDec | A flow-based full-band general audio codec with ...", "date": "", "ddg_snippet": "We show that FlowDec is a competitive alternative to the recent GAN-dominated stream of neural codecs, achieving FAD scores better than those of the established GAN-based codec DAC and listening test scores that are on par, and producing qualitatively more natural reconstructions for speech and harmonic structures in music.", "subpage_snippet": "", "source": "sp-uhh.github.io", "link": "https://sp-uhh.github.io/FlowDec/", "content": "We show that FlowDec is a competitive alternative to the recent GAN-dominated stream of neural codecs, achieving FAD scores better than those of the established GAN-based codec DAC and listening test scores that are on par, and producing qualitatively more natural reconstructions for speech and harmonic structures in music."}
+{"idx": 2, "title": "GitHub - facebookresearch/FlowDec: An neural full-band audio ...", "date": "", "ddg_snippet": "Mar 3, 2025 · An neural full-band audio codec for general audio sampled at 48 kHz with 7.5 kps or 4.5 kbps . - facebookresearch/ FlowDec", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/FlowDec", "content": "Mar 3, 2025 · An neural full-band audio codec for general audio sampled at 48 kHz with 7.5 kps or 4.5 kbps . - facebookresearch/ FlowDec"}
+{"idx": 3, "title": "FlowDec:An neural full-band audio codec for general audio ...", "date": "", "ddg_snippet": "FlowDec FlowDec (ICLR 2025) is a full-band audio codec for general audio sampled at 48 kHz that combines non-adversarial codec training with a stochastic postfilter based on a novel conditional flow matching method.", "subpage_snippet": "", "source": "gitcode.com", "link": "https://gitcode.com/gh_mirrors/fl/FlowDec/overview", "content": "FlowDec FlowDec (ICLR 2025) is a full-band audio codec for general audio sampled at 48 kHz that combines non-adversarial codec training with a stochastic postfilter based on a novel conditional flow matching method."}
+{"idx": 4, "title": "FlowDec/index.html at main · sp-uhh/FlowDec · GitHub", "date": "", "ddg_snippet": "We show that FlowDec is a competitive alternative to the recent GAN-dominated stream of neural codecs, achieving FAD scores better than those of the established GAN-based codec DAC and listening test scores that are on par, and producing qualitatively more natural reconstructions for speech and harmonic structures in music.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/sp-uhh/FlowDec/blob/main/index.html", "content": "We show that FlowDec is a competitive alternative to the recent GAN-dominated stream of neural codecs, achieving FAD scores better than those of the established GAN-based codec DAC and listening test scores that are on par, and producing qualitatively more natural reconstructions for speech and harmonic structures in music."}
+{"idx": 5, "title": "fad_pytorch - drscotthawley.github.io", "date": "", "ddg_snippet": "fad _gen supports local data read or WebDataset (audio data stored in S3 buckets) runs on CPU, CUDA, or MPS Instructions: This is designed to be run as 3 command-line scripts in succession. The latter 2 ( fad _embed and fad _ score ) are probably what most people will want: fad _gen: produces directories of real & fake audio (given real data).", "subpage_snippet": "", "source": "drscotthawley.github.io", "link": "https://drscotthawley.github.io/fad_pytorch/index.html", "content": "fad _gen supports local data read or WebDataset (audio data stored in S3 buckets) runs on CPU, CUDA, or MPS Instructions: This is designed to be run as 3 command-line scripts in succession. The latter 2 ( fad _embed and fad _ score ) are probably what most people will want: fad _gen: produces directories of real & fake audio (given real data)."}
+{"idx": 6, "title": "FlowDec : A flow -based full-band general audio codec... | OpenReview", "date": "", "ddg_snippet": "TL;DR: FlowDec is a flow -based postfilter codec for general audio without adversarial training, and a competitive alternative to current GAN-based SOTA codecs.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=uxDFlPGRLX", "content": "TL;DR: FlowDec is a flow -based postfilter codec for general audio without adversarial training, and a competitive alternative to current GAN-based SOTA codecs."}
+{"idx": 7, "title": "GitHub - hammerlab/ flowdec : TensorFlow Deconvolution for...", "date": "", "ddg_snippet": "pip install flowdec [tf_gpu].Deconvolution Validation - This notebook aggregates results from Flowdec and DeconvolutionLab2 applied to several reference datasets and verifies that deconvolved volumes are very nearly identical.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/hammerlab/flowdec", "content": "pip install flowdec [tf_gpu].Deconvolution Validation - This notebook aggregates results from Flowdec and DeconvolutionLab2 applied to several reference datasets and verifies that deconvolved volumes are very nearly identical."}
+{"idx": 8, "title": "A arXiv:2503.01485v1 [cs.SD] 3 Mar 2025", "date": "", "ddg_snippet": "We showed that FlowDec achieves state-of-the-art FAD scores for the coding task and, in a listening test, performs on par with the current state-of-the-art GAN-based codec DAC (Kumar et al., 2024) at bitrates between 4.5 and 7.5kbit/s.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.01485v1", "content": "We showed that FlowDec achieves state-of-the-art FAD scores for the coding task and, in a listening test, performs on par with the current state-of-the-art GAN-based codec DAC (Kumar et al., 2024) at bitrates between 4.5 and 7.5kbit/s."}
+{"idx": 9, "title": "(PDF) Objective Measures of Perceptual Audio Quality Reviewed: An...", "date": "", "ddg_snippet": "FlowDec : A flow -based full-band general audio codec with high perceptual quality.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/350475455_Objective_Measures_of_Perceptual_Audio_Quality_Reviewed_An_Evaluation_of_Their_Application_Domain_Dependence", "content": "FlowDec : A flow -based full-band general audio codec with high perceptual quality."}

data/sampled_jsons/Gradient-enhanced_physics-informed_neural_networks_for_forward_and_inverse_PDE_problems_Yu_et_al._20_year_2022.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Gradient-enhanced physics-informed neural networks for ...", "date": "", "ddg_snippet": "by J Yu · 2021 · Cited by 657 — ... Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , by Jeremy Yu and 3 other authors. View PDF . Abstract ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2111.02801", "content": "by J Yu · 2021 · Cited by 657 — ... Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , by Jeremy Yu and 3 other authors. View PDF . Abstract ..."}
+{"idx": 1, "title": "Gradient-enhanced physics-informed neural networks for ...", "date": "", "ddg_snippet": "by J Yu · 2022 · Cited by 657 — Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ... et al . Physics-informed neural networks: A deep learning ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0045782522001438", "content": "by J Yu · 2022 · Cited by 657 — Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ... et al . Physics-informed neural networks: A deep learning ..."}
+{"idx": 2, "title": "Gradient-enhanced physics-informed neural networks for ...", "date": "", "ddg_snippet": "by M Mohammadian · 2023 · Cited by 28 — Yu J. et al . Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput. Methods Appl. Mech. Engrg. ( 2022 ). DinhM.H. et al .", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0378779623004406", "content": "by M Mohammadian · 2023 · Cited by 28 — Yu J. et al . Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput. Methods Appl. Mech. Engrg. ( 2022 ). DinhM.H. et al ."}
+{"idx": 3, "title": "Gradient-enhanced fractional physics-informed neural ...", "date": "", "ddg_snippet": "Yu , L. Lu, X. H. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , Comput. Methods Appl ...", "subpage_snippet": "", "source": "www.aimspress.com", "link": "https://www.aimspress.com/article/doi/10.3934/math.20241332?viewType=HTML", "content": "Yu , L. Lu, X. H. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , Comput. Methods Appl ..."}
+{"idx": 4, "title": "Gradient-regularized physics informed neural networks for acou", "date": "", "ddg_snippet": "Yu , J., L. Lu, X. Meng, and G. E. Karniadakis, 2022 , Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems : Computer ...", "subpage_snippet": "", "source": "library.seg.org", "link": "https://library.seg.org/doi/pdf/10.1190/image2024-4101489.1?download=true", "content": "Yu , J., L. Lu, X. Meng, and G. E. Karniadakis, 2022 , Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems : Computer ..."}
+{"idx": 5, "title": "Physics-informed neural networks for PDE problems", "date": "", "ddg_snippet": "by K Luo · 2025 · Cited by 5 — Yu J, Lu L, Meng X, Karniadakis GE ( 2022 ) Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s10462-025-11322-7", "content": "by K Luo · 2025 · Cited by 5 — Yu J, Lu L, Meng X, Karniadakis GE ( 2022 ) Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput ..."}
+{"idx": 6, "title": "A Second-Order Network Structure Based on Gradient ...", "date": "", "ddg_snippet": "by K Sun · 2023 · Cited by 13 — Yu J., Lu L., Meng X.H., Karniadakis G.E. Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput. Methods Appl ...", "subpage_snippet": "", "source": "pmc.ncbi.nlm.nih.gov", "link": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10137436/", "content": "by K Sun · 2023 · Cited by 13 — Yu J., Lu L., Meng X.H., Karniadakis G.E. Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems . Comput. Methods Appl ..."}
+{"idx": 7, "title": "Gradient-enhanced fractional physics-informed neural ...", "date": "", "ddg_snippet": "by S Yuan · 2024 · Cited by 3 — Yu , L. Lu, X. H. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , Comput. Methods Appl ...", "subpage_snippet": "", "source": "www.aimspress.com", "link": "https://www.aimspress.com/article/doi/10.3934/math.20241332", "content": "by S Yuan · 2024 · Cited by 3 — Yu , L. Lu, X. H. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems , Comput. Methods Appl ..."}
+{"idx": 8, "title": "Physics-Informed Neural Networks", "date": "", "ddg_snippet": "Yu , Jeremy et al . Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ,. Computer Methods in Applied Mechanics and ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/media/neurips-2024/Slides/93144.pdf", "content": "Yu , Jeremy et al . Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ,. Computer Methods in Applied Mechanics and ..."}
+{"idx": 9, "title": "arXiv:2305.08310v1 [math.NA] 15 May 2023", "date": "", "ddg_snippet": "by S Lin · 2023 · Cited by 18 — [33] J. Yu , L. Lu, X. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ,.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2305.08310", "content": "by S Lin · 2023 · Cited by 18 — [33] J. Yu , L. Lu, X. Meng, G. E. Karniadakis, Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems ,."}

data/sampled_jsons/Hoeffding_inequality_formula.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Understanding Hoeffding ’s Inequality — Part 2 | by Helene | Medium", "date": "", "ddg_snippet": "Hoeffding ’s Inequality provides a bound on the probability that the sum of bounded independent random variables deviates from its expected value by more than a certain amount.", "subpage_snippet": "", "source": "helenedk.medium.com", "link": "https://helenedk.medium.com/understanding-hoeffdings-inequality-part-2-21044a334c1c", "content": "Hoeffding ’s Inequality provides a bound on the probability that the sum of bounded independent random variables deviates from its expected value by more than a certain amount."}
+{"idx": 1, "title": "Hoeffding 's Inequality - Statistics How To", "date": "", "ddg_snippet": "Hoeffding ’s inequality is useful for bounding quantities that are hard to compute. It gives an upper bound on the probability that the sum of a set of random variables deviates from the...", "subpage_snippet": "", "source": "www.statisticshowto.com", "link": "https://www.statisticshowto.com/hoeffdings-inequality/", "content": "Hoeffding ’s inequality is useful for bounding quantities that are hard to compute. It gives an upper bound on the probability that the sum of a set of random variables deviates from the..."}
+{"idx": 2, "title": "What is Hoeffding inequality . Learn its proof using Hoeffding ’s lemma.", "date": "", "ddg_snippet": "The Hoeffding Inequality gives the upper bound on the probability that the sum Sn deviates from its expected value E[Sn] by more than a certain amount.", "subpage_snippet": "", "source": "mathmonks.com", "link": "https://mathmonks.com/inequalities/hoeffding-inequality", "content": "The Hoeffding Inequality gives the upper bound on the probability that the sum Sn deviates from its expected value E[Sn] by more than a certain amount."}
+{"idx": 3, "title": "Concentration bounds. Part 4: Hoeffding inequality | Aleksandar Petrov", "date": "", "ddg_snippet": "Hoeffding ’s inequality bounds the second one. We will see how this works out in an example later.This is also called Taylor’s formula and is closely related to the Mean Value Theorem.", "subpage_snippet": "", "source": "blog.p-petrov.com", "link": "https://blog.p-petrov.com/2021-03-08/concentration-bounds-4-hoeffding", "content": "Hoeffding ’s inequality bounds the second one. We will see how this works out in an example later.This is also called Taylor’s formula and is closely related to the Mean Value Theorem."}
+{"idx": 4, "title": "probability - Applying Hoeffding 's Inequality in Real Life Situations...", "date": "", "ddg_snippet": "More generally, Hoeffding 's inequality is an example of a concentration inequality (that is, an instance of what is known as the concentration of measure phenomenon).", "subpage_snippet": "", "source": "math.stackexchange.com", "link": "https://math.stackexchange.com/questions/4836210/applying-hoeffdings-inequality-in-real-life-situations", "content": "More generally, Hoeffding 's inequality is an example of a concentration inequality (that is, an instance of what is known as the concentration of measure phenomenon)."}
+{"idx": 5, "title": "(Azuma-) Hoeffding Inequality | Math by Matt", "date": "", "ddg_snippet": "(Azuma-) Hoeffding Inequality . Posted on November 26, 2012 by Matt R.We can use Hoeffding ’s inequality to obtain a special case of the Azuma- Hoeffding concentration inequality .", "subpage_snippet": "", "source": "matthewhr.wordpress.com", "link": "https://matthewhr.wordpress.com/2012/11/26/azuma-hoeffding-inequality/", "content": "(Azuma-) Hoeffding Inequality . Posted on November 26, 2012 by Matt R.We can use Hoeffding ’s inequality to obtain a special case of the Azuma- Hoeffding concentration inequality ."}
+{"idx": 6, "title": "pr.probability - Hoeffding 's inequality for vector valued... - MathOverflow", "date": "", "ddg_snippet": "Is there a version of Hoeffding 's inequality for vector valued random variables? This seems to be hard to find and I wonder why.", "subpage_snippet": "", "source": "mathoverflow.net", "link": "https://mathoverflow.net/questions/186097/hoeffdings-inequality-for-vector-valued-random-variables", "content": "Is there a version of Hoeffding 's inequality for vector valued random variables? This seems to be hard to find and I wonder why."}
+{"idx": 7, "title": "Azuma- Hoeffding Inequality | Eventually Almost Everywhere", "date": "", "ddg_snippet": "Azuma- Hoeffding Inequality . Posted on November 23, 2015 by dominicyeo.where the final inequality follows by directly comparing the Taylor series. We’ll use this shortly.", "subpage_snippet": "", "source": "eventuallyalmosteverywhere.wordpress.com", "link": "https://eventuallyalmosteverywhere.wordpress.com/2015/11/23/azuma-hoeffding-inequality/", "content": "Azuma- Hoeffding Inequality . Posted on November 23, 2015 by dominicyeo.where the final inequality follows by directly comparing the Taylor series. We’ll use this shortly."}
+{"idx": 8, "title": "Hoeffding ’s Inequality and Martingales", "date": "", "ddg_snippet": "In this section we present Hoeffding ’s Inequality and its proof. To do so, we first go through the Hoeffding ’s Lemma and Markov’s Inequality .", "subpage_snippet": "", "source": "yangchenye.github.io", "link": "https://yangchenye.github.io/files/math/Hoeffding_s_Inequality_and_Martingales.pdf", "content": "In this section we present Hoeffding ’s Inequality and its proof. To do so, we first go through the Hoeffding ’s Lemma and Markov’s Inequality ."}
+{"idx": 9, "title": "Sample Size Estimation for Machine Learning Models Using...", "date": "", "ddg_snippet": "We can use this bound, set by Hoeffding ’s inequality to calculate the sample size needed to attain this small error. All we have to do is little algebra to solve for n.", "subpage_snippet": "", "source": "malishoaib.wordpress.com", "link": "https://malishoaib.wordpress.com/2017/09/08/sample-size-estimation-for-machine-learning-models-using-hoeffdings-inequality/", "content": "We can use this bound, set by Hoeffding ’s inequality to calculate the sample size needed to attain this small error. All we have to do is little algebra to solve for n."}

data/sampled_jsons/ICML_2025_PMLR_267_42nd_International_Conference_Machine_Learning.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "2025 Conference", "date": "", "ddg_snippet": "2025 Forty-Second International Conference on Machine Learning Vancouver Convention Center Sun. July 13th through Sat. July 19th", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/", "content": "2025 Forty-Second International Conference on Machine Learning Vancouver Convention Center Sun. July 13th through Sat. July 19th"}
+{"idx": 1, "title": "ICML 2025 Call for Papers", "date": "", "ddg_snippet": "The 42nd International Conference on Machine Learning ( ICML 2025 ) will be held in Vancouver, Canada, July 13–19, and is planned to be an in-person conference.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/Conferences/2025/CallForPapers", "content": "The 42nd International Conference on Machine Learning ( ICML 2025 ) will be held in Vancouver, Canada, July 13–19, and is planned to be an in-person conference."}
+{"idx": 2, "title": "Submission and Formatting Instructions for International ...", "date": "", "ddg_snippet": "Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). • Do not include ... 7 pages", "subpage_snippet": "", "source": "media.icml.cc", "link": "https://media.icml.cc/Conferences/ICML2025/Styles/example_paper.pdf", "content": "Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). • Do not include ... 7 pages"}
+{"idx": 3, "title": "Proceedings of Machine Learning Research | The ...", "date": "", "ddg_snippet": "The Proceedings of Machine Learning Research is a series that publishes machine learning research papers presented at conferences and workshops.", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/", "content": "The Proceedings of Machine Learning Research is a series that publishes machine learning research papers presented at conferences and workshops."}
+{"idx": 4, "title": "gdmnl/Unifews: The original code for ICML 2025 paper \" ...", "date": "", "ddg_snippet": "Unifews: You Need Fewer Operations for Efficient Graph Neural Networks. In Proceedings of the 42nd International Conference on Machine Learning , PMLR 267 , 2025 .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/gdmnl/Unifews", "content": "Unifews: You Need Fewer Operations for Efficient Graph Neural Networks. In Proceedings of the 42nd International Conference on Machine Learning , PMLR 267 , 2025 ."}
+{"idx": 5, "title": "Speaker Identity Unlearning for Zero-Shot Text-to-Speech", "date": "", "ddg_snippet": "by T Kim · 2025 · Cited by 1 — Proceedings of the 42nd International Conference on Machine Learning ( ICML 2025 ), Vancouver, Canada. PMLR 267 , 2025 . Authors Jinju Kim and ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2507.20140", "content": "by T Kim · 2025 · Cited by 1 — Proceedings of the 42nd International Conference on Machine Learning ( ICML 2025 ), Vancouver, Canada. PMLR 267 , 2025 . Authors Jinju Kim and ..."}
+{"idx": 6, "title": "Machine Learning Jun 2025", "date": "", "ddg_snippet": "1 Jun 2025 — ... 42nd International Conference on Machine Learning ( ICML 2025 ), Vancouver, Canada. Proceedings of Machine Learning Research, Vol. 267 , 2025 .", "subpage_snippet": "", "source": "www.arxiv.org", "link": "https://www.arxiv.org/list/cs.LG/2025-06?skip=25&show=2000", "content": "1 Jun 2025 — ... 42nd International Conference on Machine Learning ( ICML 2025 ), Vancouver, Canada. Proceedings of Machine Learning Research, Vol. 267 , 2025 ."}
+{"idx": 7, "title": "Offline Learning for Combinatorial Multi-armed Bandits", "date": "", "ddg_snippet": "by X Liu · Cited by 4 — Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). the past decade ...", "subpage_snippet": "", "source": "www.microsoft.com", "link": "https://www.microsoft.com/en-us/research/wp-content/uploads/2025/05/icml25_offlineCMAB2.pdf", "content": "by X Liu · Cited by 4 — Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). the past decade ..."}
+{"idx": 8, "title": "XGBoost iterations - ICML 2025 CO-BUILD contest", "date": "", "ddg_snippet": "Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). maximize ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/attachment?id=7mJt4y42LF&name=pdf", "content": "Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). maximize ..."}
+{"idx": 9, "title": "A Theory of Deep Learning Must Include Compositional ...", "date": "", "ddg_snippet": "by DA Danhofer · Cited by 1 — Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). DeepSeek-AI et al., ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=A0HtZM0MpZ", "content": "by DA Danhofer · Cited by 1 — Proceedings of the 42nd International Conference on Machine . Learning , Vancouver, Canada. PMLR 267 , 2025 . Copyright 2025 by the author(s). DeepSeek-AI et al., ..."}

data/sampled_jsons/IMhoJgWANP_Digital_Disparities_median_website_size_developed_developing_countries_Figure_1.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "PDF Digital Disparities: A Comparative Web Measurement Study Across ...", "date": "", "ddg_snippet": "Figure 1(a) shows that the median website size in developed countries is 3.35 MB, i.e., slightly larger than the 2.99 MB median size in developing countries , both of which exceed the global median of 2.6 MB as per the HTTP archive [28].", "subpage_snippet": "", "source": "www.staicu.org", "link": "https://www.staicu.org/publications/www2025.pdf", "content": "Figure 1(a) shows that the median website size in developed countries is 3.35 MB, i.e., slightly larger than the 2.99 MB median size in developing countries , both of which exceed the global median of 2.6 MB as per the HTTP archive [28]."}
+{"idx": 1, "title": "Widening Digital Gap between Developed, Developing States Threatening ...", "date": "", "ddg_snippet": "A widening digital divide and severely lagging Internet-use in developing countries threaten to leave those States in the technological wake and preclude progress on the Sustainable Development Goals (SDGs), a senior United Nations official and Member States told the Second Committee (Economic and Financial) today as it took up information and communications technology (ICT) questions.", "subpage_snippet": "", "source": "press.un.org", "link": "https://press.un.org/en/2023/gaef3587.doc.htm", "content": "A widening digital divide and severely lagging Internet-use in developing countries threaten to leave those States in the technological wake and preclude progress on the Sustainable Development Goals (SDGs), a senior United Nations official and Member States told the Second Committee (Economic and Financial) today as it took up information and communications technology (ICT) questions."}
+{"idx": 2, "title": "Unequal Internet: study highlights differences between websites from ...", "date": "", "ddg_snippet": "Differences in digitalization between developing and developed countries involve a variety of factors. \"Previous studies have primarily examined macro-level indicators, such as the availability of Internet, smartphone usage, or general technological infrastructure,\" CISPA researcher Masudul Bhuiyan explains. These studies show that only 60 percent of the population in developing countries ...", "subpage_snippet": "", "source": "nachrichten.idw-online.de", "link": "https://nachrichten.idw-online.de/2025/04/29/unequal-internet-study-highlights-differences-between-websites-from-developing-and-developed-countries", "content": "Differences in digitalization between developing and developed countries involve a variety of factors. \"Previous studies have primarily examined macro-level indicators, such as the availability of Internet, smartphone usage, or general technological infrastructure,\" CISPA researcher Masudul Bhuiyan explains. These studies show that only 60 percent of the population in developing countries ..."}
+{"idx": 3, "title": "Websites from developing and developed countries differ", "date": "", "ddg_snippet": "These studies show that only 60 percent of the population in developing countries are online, whereas in developed countries , the figure is 93 percent. Conversely, people in developing countries ...", "subpage_snippet": "", "source": "www.linkedin.com", "link": "https://www.linkedin.com/pulse/unequal-internet-study-highlights-differences-between-websites-from-u7uje", "content": "These studies show that only 60 percent of the population in developing countries are online, whereas in developed countries , the figure is 93 percent. Conversely, people in developing countries ..."}
+{"idx": 4, "title": "Measuring digital development: Facts and Figures 2024", "date": "", "ddg_snippet": "Landlocked Developing Countries (LLDCs) face unique constraints, but digital technologies offer powerful tools to overcome them. They can unlock access to essential services, strengthen resilience, and support integration into the global economy. This special edition of ITU's Facts and Figures series presents a comprehensive overview of digital progress in LLDCs, with key data and analysis ...", "subpage_snippet": "", "source": "www.itu.int", "link": "https://www.itu.int/en/ITU-D/Statistics/Pages/facts/default.aspx/", "content": "Landlocked Developing Countries (LLDCs) face unique constraints, but digital technologies offer powerful tools to overcome them. They can unlock access to essential services, strengthen resilience, and support integration into the global economy. This special edition of ITU's Facts and Figures series presents a comprehensive overview of digital progress in LLDCs, with key data and analysis ..."}
+{"idx": 5, "title": "Digital Disparities: A Comparative Web Measurement Study Across ...", "date": "", "ddg_snippet": "in developed countries is 3.35 MB, slightly larger than the 2.99 i.e., MB median size in developing countries , both of which exceed the global median of 2.6 MB as per the HTTP archive [22].", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=IMhoJgWANP", "content": "in developed countries is 3.35 MB, slightly larger than the 2.99 i.e., MB median size in developing countries , both of which exceed the global median of 2.6 MB as per the HTTP archive [22]."}
+{"idx": 6, "title": "Figure 1. The digital divide between the world, developed and...", "date": "", "ddg_snippet": "Download scientific diagram | The digital divide between the world, developed and developing countries in term of internet users per 100 inhabitants from publication: Internet and the Arab World ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/figure/The-digital-divide-between-the-world-developed-and-developing-countries-in-term-of_fig1_220413609", "content": "Download scientific diagram | The digital divide between the world, developed and developing countries in term of internet users per 100 inhabitants from publication: Internet and the Arab World ..."}
+{"idx": 7, "title": "Digital Progress and Trends Report - World Bank Group", "date": "", "ddg_snippet": "This report tracks global progress of digitalization and countries' production and use of digital technologies, from digital jobs, digital services exports, and app development to internet use, affordability, quality, and more. The report also highlights policy shifts and debates, with a focus on developing countries .", "subpage_snippet": "", "source": "www.worldbank.org", "link": "https://www.worldbank.org/en/publication/digital-progress-and-trends-report", "content": "This report tracks global progress of digitalization and countries' production and use of digital technologies, from digital jobs, digital services exports, and app development to internet use, affordability, quality, and more. The report also highlights policy shifts and debates, with a focus on developing countries ."}
+{"idx": 8, "title": "Unequal Internet: study highlights differences between websites from ...", "date": "", "ddg_snippet": "Masudul Bhuiyan compared 200,000 websites from 20 developing and developed countries to find out abou security and data privacy differences.", "subpage_snippet": "", "source": "cispa.de", "link": "https://cispa.de/en/bhuiyan-unequal-internet", "content": "Masudul Bhuiyan compared 200,000 websites from 20 developing and developed countries to find out abou security and data privacy differences."}
+{"idx": 9, "title": "Measuring the impact of the digital economy in developing countries: A ...", "date": "", "ddg_snippet": "This paper explores the definition, measurement, role, and impacts of the digital economy across various economies. It also examines the involvement of governments and telecommunication regulators in assessing the digital economy and identifies future directions for developing countries .", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S2405844023048624", "content": "This paper explores the definition, measurement, role, and impacts of the digital economy across various economies. It also examines the involvement of governments and telecommunication regulators in assessing the digital economy and identifies future directions for developing countries ."}

data/sampled_jsons/Jin_et_al._2024b_PH3_method_language_models.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Large language model - Wikipedia", "date": "", "ddg_snippet": "Machine learningand data mining. v. t. e. A large language model is a language model trained with self-supervised machine learning on a vast amount of text, designed for natural language processing tasks, especially language generation.", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Large_language_model", "content": "Machine learningand data mining. v. t. e. A large language model is a language model trained with self-supervised machine learning on a vast amount of text, designed for natural language processing tasks, especially language generation."}
+{"idx": 1, "title": "DeepSeek - Wikipedia", "date": "", "ddg_snippet": "Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence company that develops large language models . Based in Hangzhou, Zhejiang, Deepseek is owned and funded by ...", "subpage_snippet": "", "source": "en.m.wikipedia.org", "link": "https://en.m.wikipedia.org/wiki/DeepSeek", "content": "Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence company that develops large language models . Based in Hangzhou, Zhejiang, Deepseek is owned and funded by ..."}
+{"idx": 2, "title": "Evaluating the Role of Large Language Models Detection...", "date": "", "ddg_snippet": "Ding, Jun-En, et al . (2024). Large language multimodal models for new-onset type 2 diabetes prediction using five-year cohort electronic health records. Scientific Reports, 14(1), 20774.", "subpage_snippet": "", "source": "abjar.vandanapublications.com", "link": "https://abjar.vandanapublications.com/index.php/ojs/article/view/75", "content": "Ding, Jun-En, et al . (2024). Large language multimodal models for new-onset type 2 diabetes prediction using five-year cohort electronic health records. Scientific Reports, 14(1), 20774."}
+{"idx": 3, "title": "Beyond Bradley-Terry Models : A General... | Read Paper on Bytez", "date": "", "ddg_snippet": "(2024) and Furuta et al . (2024) proposed robust preference optimization methods to handle noise and soft labels, respectively.", "subpage_snippet": "", "source": "bytez.com", "link": "https://bytez.com/docs/icml/45103/paper", "content": "(2024) and Furuta et al . (2024) proposed robust preference optimization methods to handle noise and soft labels, respectively."}
+{"idx": 4, "title": "Fairness in Large Language Models in Three Hours", "date": "", "ddg_snippet": "Large Language Models (LLMs) have demonstrated remarkable success across various domains but often lack fairness considerations, potentially leading to discriminatory outcomes against marginalized populations.Chu et al . ( 2024 b ) Zhibo Chu, Zichong Wang, and Wenbin Zhang.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2408.00992v2", "content": "Large Language Models (LLMs) have demonstrated remarkable success across various domains but often lack fairness considerations, potentially leading to discriminatory outcomes against marginalized populations.Chu et al . ( 2024 b ) Zhibo Chu, Zichong Wang, and Wenbin Zhang."}
+{"idx": 5, "title": "By My Eyes: Grounding Multimodal Large Language Models", "date": "", "ddg_snippet": "HuggingGPT (Shen et al ., 2024 ) and Chameleon (Lu et al ., 2024 ) in-tegrated multiple expert models to enhance func-tionalities. Recently, Data Interpreter (Hong et al ., 2024 ) enabled LLMs to analyze data and build task-specific models for data interpretation.", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2024.emnlp-main.133.pdf", "content": "HuggingGPT (Shen et al ., 2024 ) and Chameleon (Lu et al ., 2024 ) in-tegrated multiple expert models to enhance func-tionalities. Recently, Data Interpreter (Hong et al ., 2024 ) enabled LLMs to analyze data and build task-specific models for data interpretation."}
+{"idx": 6, "title": "(PDF) Not All Jokes Land: Evaluating Large Language Models ...", "date": "", "ddg_snippet": "2024;Scherbakov et al ., 2024 ;Shen et al ., 2024 ). An example of automated writing is the use of. LLM agents to write an email, a document, or a.tion, (Tikhonov and Shtykovskiy, 2024 b ) evaluates. Vision- language models on humor, (Chen et al .", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/392371334_Not_All_Jokes_Land_Evaluating_Large_Language_Models_Understanding_of_Workplace_Humor", "content": "2024;Scherbakov et al ., 2024 ;Shen et al ., 2024 ). An example of automated writing is the use of. LLM agents to write an email, a document, or a.tion, (Tikhonov and Shtykovskiy, 2024 b ) evaluates. Vision- language models on humor, (Chen et al ."}
+{"idx": 7, "title": "What Are Large Language Models (LLMs)? | IBM", "date": "", "ddg_snippet": "Large language models (LLMs) are a category of deep learning models trained on immense amounts of data, making them capable of understanding and generating natural language and other types of content to perform a wide range of tasks.", "subpage_snippet": "", "source": "www.ibm.com", "link": "https://www.ibm.com/think/topics/large-language-models", "content": "Large language models (LLMs) are a category of deep learning models trained on immense amounts of data, making them capable of understanding and generating natural language and other types of content to perform a wide range of tasks."}
+{"idx": 8, "title": "Parenting: Optimizing Knowledge Selection of", "date": "", "ddg_snippet": "... Language Models (LLMs) have demonstrated exceptional capabilities, achieving state-of-the-art performance on various tasks Brown ( 2020 ); Chowdhery ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.10360v3", "content": "... Language Models (LLMs) have demonstrated exceptional capabilities, achieving state-of-the-art performance on various tasks Brown ( 2020 ); Chowdhery ..."}
+{"idx": 9, "title": "An open platform for evaluating AI through human preference", "date": "", "ddg_snippet": "Compare answers across top AI models , share your feedback and power our public leaderboard. Inputs are processed by third -party AI and responses may be inaccurate.", "subpage_snippet": "", "source": "lmarena.ai", "link": "https://lmarena.ai/", "content": "Compare answers across top AI models , share your feedback and power our public leaderboard. Inputs are processed by third -party AI and responses may be inaccurate."}

data/sampled_jsons/Leveraging_Per-Instance_Privacy_for_Machine_Unlearning_arXiv_2024_publication_date.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "by NM Sepahvand · 2025 — Abstract page for arXiv paper 2505.18786: Leveraging Per-Instance Privacy for Machine Unlearning . ... 2024 ), obtaining a better utility ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.18786", "content": "by NM Sepahvand · 2025 — Abstract page for arXiv paper 2505.18786: Leveraging Per-Instance Privacy for Machine Unlearning . ... 2024 ), obtaining a better utility ..."}
+{"idx": 1, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning . Download PDF. Nazanin ... ( 2024 ) into the unlearning setting, this paper builds the bound of ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0A4Y9qRnu9&noteId=Zd6KsMzKb8", "content": "Leveraging Per-Instance Privacy for Machine Unlearning . Download PDF. Nazanin ... ( 2024 ) into the unlearning setting, this paper builds the bound of ..."}
+{"idx": 2, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning . Naz Sepahvand ... arXiv :2411.04388, 2024 . Barbulescu, G.-O. and Triantafillou, P. To each ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46697", "content": "Leveraging Per-Instance Privacy for Machine Unlearning . Naz Sepahvand ... arXiv :2411.04388, 2024 . Barbulescu, G.-O. and Triantafillou, P. To each ..."}
+{"idx": 3, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "24 May 2025 — Leveraging Per-Instance Privacy for Machine Unlearning . Report issue ... 2024b ) , or updating out-of- date information. Report issue for ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.18786v1", "content": "24 May 2025 — Leveraging Per-Instance Privacy for Machine Unlearning . Report issue ... 2024b ) , or updating out-of- date information. Report issue for ..."}
+{"idx": 4, "title": "TPDP 2025 – Theory and Practice of Differential Privacy", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning Nazanin Mohammadi ... Submissions: Authors are invited to submit a short abstract of new work or work ...", "subpage_snippet": "", "source": "tpdp.journalprivacyconfidentiality.org", "link": "https://tpdp.journalprivacyconfidentiality.org/2025/", "content": "Leveraging Per-Instance Privacy for Machine Unlearning Nazanin Mohammadi ... Submissions: Authors are invited to submit a short abstract of new work or work ..."}
+{"idx": 5, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning . cs.LG26 May 2025. Nazanin Mohammadi Sepahvand, Anvith Thudi, Berivan Isik, Ashmita Bhattacharyya ...", "subpage_snippet": "", "source": "chatpaper.com", "link": "https://chatpaper.com/pt/chatpaper/paper/141960", "content": "Leveraging Per-Instance Privacy for Machine Unlearning . cs.LG26 May 2025. Nazanin Mohammadi Sepahvand, Anvith Thudi, Berivan Isik, Ashmita Bhattacharyya ..."}
+{"idx": 6, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "2024 · arXiv · [5] Machine Unlearning. 2021 · [6] ... 2024 · arXiv · [30] Deep ... Leveraging Per-Instance Privacy for Machine Unlearning | Read Paper on Bytez.", "subpage_snippet": "", "source": "bytez.com", "link": "https://bytez.com/docs/icml/46697/references?_c=eyJ2IjoxLCJyZWxhdGVkIjpbImNvZGUiLCJyZWZlcmVuY2VzIiwiY29uZmVyZW5jZSJdfQ==", "content": "2024 · arXiv · [5] Machine Unlearning. 2021 · [6] ... 2024 · arXiv · [30] Deep ... Leveraging Per-Instance Privacy for Machine Unlearning | Read Paper on Bytez."}
+{"idx": 7, "title": "ICML 2025 Papers", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning · Conditioning Diffusions Using Malliavin Calculus · MindCustomer: Multi-Context Image Generation Blended ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/papers.html", "content": "Leveraging Per-Instance Privacy for Machine Unlearning · Conditioning Diffusions Using Malliavin Calculus · MindCustomer: Multi-Context Image Generation Blended ..."}
+{"idx": 8, "title": "Anvith Thudi", "date": "", "ddg_snippet": "Leveraging Per-Instance Privacy for Machine Unlearning . NM Sepahvand, A Thudi, B Isik, A Bhattacharyya, N Papernot, ... Proceedings of the 42nd International ...", "subpage_snippet": "", "source": "scholar.google.com", "link": "https://scholar.google.com/citations?user=bTEybH0AAAAJ&hl=en", "content": "Leveraging Per-Instance Privacy for Machine Unlearning . NM Sepahvand, A Thudi, B Isik, A Bhattacharyya, N Papernot, ... Proceedings of the 42nd International ..."}
+{"idx": 9, "title": "Computer Science May 2025", "date": "", "ddg_snippet": "17 May 2025 — Comments: Accepted manuscript, published in ICIC 2024 . Final version at ... Title: Leveraging Per-Instance Privacy for Machine Unlearning .", "subpage_snippet": "", "source": "www.arxiv.org", "link": "https://www.arxiv.org/list/cs/2025-05?skip=8775&show=2000", "content": "17 May 2025 — Comments: Accepted manuscript, published in ICIC 2024 . Final version at ... Title: Leveraging Per-Instance Privacy for Machine Unlearning ."}

data/sampled_jsons/MHaSq1LlTe_Signed_Laplacians_Constrained_Graph_Clustering_Algorithm_1_self-loop_weights.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Cluster analysis - Wikipedia", "date": "", "ddg_snippet": "Signed graph models: Every path in a signed graph has a sign from the product of the signs on the edges. Under the assumptions of balance theory, edges may change sign and result in a bifurcated graph .", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/Cluster_analysis", "content": "Signed graph models: Every path in a signed graph has a sign from the product of the signs on the edges. Under the assumptions of balance theory, edges may change sign and result in a bifurcated graph ."}
+{"idx": 1, "title": "Signed Laplacians for Constrained Graph Clustering - OpenReview", "date": "", "ddg_snippet": "In this section, we present a constrained graph clustering algorithm called CC++. At a high level, our algorithm con-sists of the following: in the preprocessing step, we adjust the edge weights of G and add self -loops to the vertices of G, such that both of G and H have the same degree se-quence.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=MHaSq1LlTe", "content": "In this section, we present a constrained graph clustering algorithm called CC++. At a high level, our algorithm con-sists of the following: in the preprocessing step, we adjust the edge weights of G and add self -loops to the vertices of G, such that both of G and H have the same degree se-quence."}
+{"idx": 2, "title": "On Spectral Properties of Signed Laplacians with Connections ...", "date": "", "ddg_snippet": "Abstract— Signed graphs have appeared in a broad variety of applications, ranging from social networks to biological networks, from distributed control and computation to power systems. In this paper, we investigate spectral properties of signed Laplacians for undirected signed graphs. We find conditions on the negative weights under which a signed Laplacian is positive semidefinite via the ...", "subpage_snippet": "", "source": "eeqiu.people.ust.hk", "link": "http://eeqiu.people.ust.hk/wp-content/uploads/2021/09/On-Spectral-Properties-of-Signed-Laplacians-with-Connections-to-Eventual-Positivity.pdf", "content": "Abstract— Signed graphs have appeared in a broad variety of applications, ranging from social networks to biological networks, from distributed control and computation to power systems. In this paper, we investigate spectral properties of signed Laplacians for undirected signed graphs. We find conditions on the negative weights under which a signed Laplacian is positive semidefinite via the ..."}
+{"idx": 3, "title": "Chapter 4 Signed Graphs - University of Pennsylvania", "date": "", "ddg_snippet": "Chapter 4 Signed Graphs 4. 1 Signed Graphs and Signed Laplacians Intuitively, in a weighted graph , an edge with a positive weight denotes similarity or proximity of its endpoints. For many reasons, it is desirable to allow edges labeled with negative weights , the intuition being that a nega-tive weight indicates dissimilarity or distance.", "subpage_snippet": "", "source": "www.cis.upenn.edu", "link": "https://www.cis.upenn.edu/~cis5150/cis515-15-spectral-clust-chap4.pdf", "content": "Chapter 4 Signed Graphs 4. 1 Signed Graphs and Signed Laplacians Intuitively, in a weighted graph , an edge with a positive weight denotes similarity or proximity of its endpoints. For many reasons, it is desirable to allow edges labeled with negative weights , the intuition being that a nega-tive weight indicates dissimilarity or distance."}
+{"idx": 4, "title": "Lecture: Signed Graph Clustering 4mm Foundations of Data ...", "date": "", "ddg_snippet": "Clustering of the empirical correlation matrix of 1500 time series (stocks contained in the S&P 1500 index) Compute the bottom k = 10 eigenvectors of L, and run a standard machine learning clustering algorithm (k-means++) to recover k clusters.", "subpage_snippet": "", "source": "www.stats.ox.ac.uk", "link": "https://www.stats.ox.ac.uk/~cucuring/SignedClustering_CDT.pdf", "content": "Clustering of the empirical correlation matrix of 1500 time series (stocks contained in the S&P 1500 index) Compute the bottom k = 10 eigenvectors of L, and run a standard machine learning clustering algorithm (k-means++) to recover k clusters."}
+{"idx": 5, "title": "Spectral Clustering of Signed Graphs via Matrix Power Means", "date": "", "ddg_snippet": "Abstract Signed graphs encode positive (attractive) and negative (repulsive) relations between nodes. We extend spectral clustering to signed graphs via the one-parameter family of Signed Power Mean Laplacians , defined as the matrix power mean of normalized standard and signless Laplacians of positive and negative edges.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1905.06230", "content": "Abstract Signed graphs encode positive (attractive) and negative (repulsive) relations between nodes. We extend spectral clustering to signed graphs via the one-parameter family of Signed Power Mean Laplacians , defined as the matrix power mean of normalized standard and signless Laplacians of positive and negative edges."}
+{"idx": 6, "title": "ICML Poster Signed Laplacians for Constrained Graph Clustering", "date": "", "ddg_snippet": "In this work, we establish a Cheeger-type inequality that relates the solution of the constrained clustering problem to the spectral properties of G and H. To reduce computational complexity, we utilise the signed Laplacian of H, streamlining calculations while maintaining accuracy.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/45552", "content": "In this work, we establish a Cheeger-type inequality that relates the solution of the constrained clustering problem to the spectral properties of G and H. To reduce computational complexity, we utilise the signed Laplacian of H, streamlining calculations while maintaining accuracy."}
+{"idx": 7, "title": "Signed Laplacians for Constrained Graph Clustering", "date": "", "ddg_snippet": "May 1 , 2025 · This paper considers the constrained clustering problems over two graphs. This paper establishes the Cheeger inequality for the proposed algorithm for constrained clustering , which can be a counterpart of Cheeger inequality to the standard spectral clustering over a graph . The proposed algorithm improves spectral clustering , in a scenario that is challenigin for spectral clustering . This paper ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=MHaSq1LlTe", "content": "May 1 , 2025 · This paper considers the constrained clustering problems over two graphs. This paper establishes the Cheeger inequality for the proposed algorithm for constrained clustering , which can be a counterpart of Cheeger inequality to the standard spectral clustering over a graph . The proposed algorithm improves spectral clustering , in a scenario that is challenigin for spectral clustering . This paper ..."}
+{"idx": 8, "title": "(PDF) Local Sample- Weighted Multiple Kernel Clustering With...", "date": "", "ddg_snippet": "7) Self - weighted Multiview Clustering with Multiple. Graphs (SwMC) [57] eliminates the undesired hyperbipartite graph for subspace clustering via constrained Laplacian . rank,” IEEE Trans.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/362038239_Local_Sample-Weighted_Multiple_Kernel_Clustering_With_Consensus_Discriminative_Graph", "content": "7) Self - weighted Multiview Clustering with Multiple. Graphs (SwMC) [57] eliminates the undesired hyperbipartite graph for subspace clustering via constrained Laplacian . rank,” IEEE Trans."}
+{"idx": 9, "title": "GitHub - joandre/MCL_spark: An implementation of Markov Clustering ...", "date": "", "ddg_snippet": "Self loops weight management => A percentage of the maximum weight can be applied to added self loops . For example, for a binary graph , 1 is the maximum weight to allocate (see Optimization paragraph for more details).", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/joandre/MCL_spark", "content": "Self loops weight management => A percentage of the maximum weight can be applied to added self loops . For example, for a binary graph , 1 is the maximum weight to allocate (see Optimization paragraph for more details)."}

data/sampled_jsons/Normalizing_Flows_are_Capable_Generative_Models_Equation_6_training_loss.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": "(PDF) Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "PDF | Normalizing Flows (NFs) are likelihood-based models for continuous inputs.We first show a typical training loss curve. together with an online monitoring of the model ’s sample. quality in terms of FID (we use 4096 samples for efficiency).", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/386577118_Normalizing_Flows_are_Capable_Generative_Models", "content": "PDF | Normalizing Flows (NFs) are likelihood-based models for continuous inputs.We first show a typical training loss curve. together with an online monitoring of the model ’s sample. quality in terms of FID (we use 4096 samples for efficiency)."}
+{"idx": 2, "title": "Normalizing Flows are Capable Generative Models | OpenReview", "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": "openreview.net", "link": "https://openreview.net/forum?id=2uheUFcFsM", "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": 3, "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…", "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…"}
+{"idx": 4, "title": "Normalizing Flows (NFs)", "date": "", "ddg_snippet": "Normalizing Flows are Capable Generative Models (2024). 10. Density estimation on low-dimensional manifolds: an inflation-deflation approach (2021).", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/topics/normalizing-flows-nfs", "content": "Normalizing Flows are Capable Generative Models (2024). 10. Density estimation on low-dimensional manifolds: an inflation-deflation approach (2021)."}
+{"idx": 5, "title": "Flow -based Deep Generative Models | Lil'Log", "date": "", "ddg_snippet": "Flow -based deep generative models conquer this hard problem with the help of normalizing flows , a powerful statistics tool for density estimation.Batch normalization is found to help training models with a very deep stack of coupling layers.", "subpage_snippet": "", "source": "lilianweng.github.io", "link": "https://lilianweng.github.io/posts/2018-10-13-flow-models/", "content": "Flow -based deep generative models conquer this hard problem with the help of normalizing flows , a powerful statistics tool for density estimation.Batch normalization is found to help training models with a very deep stack of coupling layers."}
+{"idx": 6, "title": "NeurIPS 2020 : Training Normalizing Flows with the Information...", "date": "", "ddg_snippet": "We pose the question whether the IB can also be used to train generative likelihood models such as normalizing flows . Since normalizing flows use invertible network architectures (INNs), they are information-preserving by construction.", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2020/protected/poster_593906af0d138e69f49d251d3e7cbed0.html", "content": "We pose the question whether the IB can also be used to train generative likelihood models such as normalizing flows . Since normalizing flows use invertible network architectures (INNs), they are information-preserving by construction."}
+{"idx": 7, "title": "Normalizing Flows . I have been learning about Normalizing | Medium", "date": "", "ddg_snippet": "Generative models , as the name suggests, are developed to learn distribution of a given data and then, based on the distribution learnt, can generate new data points that did not exist before, but look quite real. You must be aware of ‘Deepfakes’, which mostly use GANs ( Generative ...", "subpage_snippet": "", "source": "grishmaprs.medium.com", "link": "https://grishmaprs.medium.com/normalizing-flows-5b5a713e45e2", "content": "Generative models , as the name suggests, are developed to learn distribution of a given data and then, based on the distribution learnt, can generate new data points that did not exist before, but look quite real. You must be aware of ‘Deepfakes’, which mostly use GANs ( Generative ..."}
+{"idx": 8, "title": "Normalizing Flows are Capable Generative Models | alphaXiv", "date": "", "ddg_snippet": "Abstract: 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": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/abs/2412.06329", "content": "Abstract: 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": 9, "title": "Kernelized Normalizing Flows -Bohrium", "date": "", "ddg_snippet": "Normalising Flows are generative models characterised by their invertible architecture. Normalising Flows . Generative models . Invertible architecture. Neural network-based transformations.", "subpage_snippet": "", "source": "www.bohrium.com", "link": "https://www.bohrium.com/paper-details/kernelized-normalizing-flows/908055302069289191-108580", "content": "Normalising Flows are generative models characterised by their invertible architecture. Normalising Flows . Generative models . Invertible architecture. Neural network-based transformations."}

data/sampled_jsons/Olah_2020_Circuits_CNN_convolutional_networks_vs_transformers_attention_year_2020.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Stacked Hybrid RNN-CNN Reconstruction of X-ray Influence on", "date": "", "ddg_snippet": "Along with Convolutional Neural Networks ( CNNs ), this architecture combines two substantial forms of recurrent neural networks (RNNs), Long Short ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.05842v1", "content": "Along with Convolutional Neural Networks ( CNNs ), this architecture combines two substantial forms of recurrent neural networks (RNNs), Long Short ..."}
+{"idx": 1, "title": "Studying Cross-cluster Modularity in Neural Networks", "date": "", "ddg_snippet": "We investigate CNNs trained on MNIST and CIFAR, small transformers trained on modular addition, and GPT-2 and Pythia on the Wiki dataset, and Gemma ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.02470v3", "content": "We investigate CNNs trained on MNIST and CIFAR, small transformers trained on modular addition, and GPT-2 and Pythia on the Wiki dataset, and Gemma ..."}
+{"idx": 2, "title": "Interpreting wide-band neural activity using convolutional", "date": "", "ddg_snippet": "To test this proposal, we developed a convolutional network ( LeCun et al., 2015 ) able to take minimally processed, wide-band neural data as input ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/articles/66551", "content": "To test this proposal, we developed a convolutional network ( LeCun et al., 2015 ) able to take minimally processed, wide-band neural data as input ..."}
+{"idx": 3, "title": "Haiku Tech Center", "date": "", "ddg_snippet": "So is it really just all about sparseness not attention ? Or is it really all about mixing as Yannic says? We show that Transformer encoder ...", "subpage_snippet": "", "source": "www.haikutechcenter.com", "link": "https://www.haikutechcenter.com/2021/05/", "content": "So is it really just all about sparseness not attention ? Or is it really all about mixing as Yannic says? We show that Transformer encoder ..."}
+{"idx": 4, "title": "Possible research directions to improve the mechanistic", "date": "", "ddg_snippet": "It turns out that for the Circuits thread the only constraints they applied are transformation robustness to padding, jitter, scaling, and rotation ...", "subpage_snippet": "", "source": "www.lesswrong.com", "link": "https://www.lesswrong.com/posts/HiutLvY2x7zrsTQkx/possible-research-directions-to-improve-the-mechanistic", "content": "It turns out that for the Circuits thread the only constraints they applied are transformation robustness to padding, jitter, scaling, and rotation ..."}
+{"idx": 5, "title": "Tensor Labbet · A blog of deep learnings", "date": "", "ddg_snippet": "Fully Convolutional Networks , 2015] were proposed as a neural network architecture for dense prediction of pixel-wise labels.", "subpage_snippet": "", "source": "tensorlabbet.com", "link": "https://tensorlabbet.com/", "content": "Fully Convolutional Networks , 2015] were proposed as a neural network architecture for dense prediction of pixel-wise labels."}
+{"idx": 6, "title": "Gain, not concomitant changes in spatial receptive field", "date": "", "ddg_snippet": "... behavioral experiments using a categorization task and a convolutional neural network model to test different mechanisms that may support attention ...", "subpage_snippet": "", "source": "elifesciences.org", "link": "https://elifesciences.org/articles/78392", "content": "... behavioral experiments using a categorization task and a convolutional neural network model to test different mechanisms that may support attention ..."}
+{"idx": 7, "title": "The Sensory Moving Image Archive | Boosting Creative Reuse for", "date": "", "ddg_snippet": "... s final stage and share a work that one of these collaborations resulted in – the video work Alien Visions ( 2020 ) by Pablo N.", "subpage_snippet": "", "source": "sensorymovingimagearchive.humanities.uva.nl", "link": "https://sensorymovingimagearchive.humanities.uva.nl/", "content": "... s final stage and share a work that one of these collaborations resulted in – the video work Alien Visions ( 2020 ) by Pablo N."}
+{"idx": 8, "title": "‘AI scaling’ directory · Gwern.net", "date": "", "ddg_snippet": "Over-Tokenized Transformer : Vocabulary Is Generally ... Byte Latent Transformer (BLT): Patches Scale Better Than Tokens ”, Pagnoni et al 2024", "subpage_snippet": "", "source": "gwern.net", "link": "https://gwern.net/doc/ai/scaling/index", "content": "Over-Tokenized Transformer : Vocabulary Is Generally ... Byte Latent Transformer (BLT): Patches Scale Better Than Tokens ”, Pagnoni et al 2024"}
+{"idx": 9, "title": "A Domain-Specific Supercomputer for Training Deep Neural", "date": "", "ddg_snippet": "We also start with a neural network model, which transforms the input into the result through an intensive calculation of weights (also called ...", "subpage_snippet": "", "source": "cacm.acm.org", "link": "https://cacm.acm.org/research/a-domain-specific-supercomputer-for-training-deep-neural-networks/", "content": "We also start with a neural network model, which transforms the input into the result through an intensive calculation of weights (also called ..."}

data/sampled_jsons/Olah_circuit_analysis_attention_matrices_structural_patterns.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "A Mathematical Framework for Transformer Circuits", "date": "", "ddg_snippet": "22 Dec 2021 — Attention heads can be understood as having two largely independent computations: a QK (“query-key”) circuit which computes the attention ...", "subpage_snippet": "", "source": "transformer-circuits.pub", "link": "https://transformer-circuits.pub/2021/framework/index.html", "content": "22 Dec 2021 — Attention heads can be understood as having two largely independent computations: a QK (“query-key”) circuit which computes the attention ..."}
+{"idx": 1, "title": "Analyzing Shared Circuits in Large Language Models", "date": "", "ddg_snippet": "Attention pattern analysis reveals that these heads detect how sequence members (as queries) attend to sequence members (as keys) of the same type, such as ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2311.04131v4", "content": "Attention pattern analysis reveals that these heads detect how sequence members (as queries) attend to sequence members (as keys) of the same type, such as ..."}
+{"idx": 2, "title": "Attention Layers Add Into Low-Dimensional Residual ...", "date": "", "ddg_snippet": "23 Aug 2025 — To investigate whether this low-rank structure originates from the attention heads ( Z Z ), the output projection matrix ( W O W^{O} ), or their ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.16929v1", "content": "23 Aug 2025 — To investigate whether this low-rank structure originates from the attention heads ( Z Z ), the output projection matrix ( W O W^{O} ), or their ..."}
+{"idx": 3, "title": "Circuits Updates - July 2025", "date": "", "ddg_snippet": "Every attention head can be understood in terms of two matrices : the OV circuit (which describes what information the attention head reads and writes) and ...", "subpage_snippet": "", "source": "transformer-circuits.pub", "link": "https://transformer-circuits.pub/2025/july-update/index.html", "content": "Every attention head can be understood in terms of two matrices : the OV circuit (which describes what information the attention head reads and writes) and ..."}
+{"idx": 4, "title": "Analyzing Shared Circuits in Large Language Models", "date": "", "ddg_snippet": "by M Lan · 2024 · Cited by 2 — Attention Pattern Analysis . We analyze the. QK matrix of attention heads to track information movement from keys to queries. We take the mean. 26 pages", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2024.emnlp-main.699.pdf", "content": "by M Lan · 2024 · Cited by 2 — Attention Pattern Analysis . We analyze the. QK matrix of attention heads to track information movement from keys to queries. We take the mean. 26 pages"}
+{"idx": 5, "title": "ICML Poster The underlying structures of self-attention", "date": "", "ddg_snippet": "Poster. The underlying structures of self- attention : symmetry, directionality, and emergent dynamics in Transformer training.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/44452", "content": "Poster. The underlying structures of self- attention : symmetry, directionality, and emergent dynamics in Transformer training."}
+{"idx": 6, "title": "Attention heads of large language models: Patterns", "date": "", "ddg_snippet": "by Z Zheng · 2025 · Cited by 14 — This study focuses on the attention heads of LLMs, offering a systematic review of their behavior and associated experimental methods across various scenarios.", "subpage_snippet": "", "source": "www.cell.com", "link": "https://www.cell.com/patterns/fulltext/S2666-3899(25)00024-8", "content": "by Z Zheng · 2025 · Cited by 14 — This study focuses on the attention heads of LLMs, offering a systematic review of their behavior and associated experimental methods across various scenarios."}
+{"idx": 7, "title": "INTERPRETABILITY IN THE WILD: A CIRCUIT FOR ...", "date": "", "ddg_snippet": "by KR Wang · Cited by 623 — Attention pattern analysis: Using attention patterns to explain behavior is always worrying due to the possibility that information has accumulated on that ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=NpsVSN6o4ul", "content": "by KR Wang · Cited by 623 — Attention pattern analysis: Using attention patterns to explain behavior is always worrying due to the possibility that information has accumulated on that ..."}
+{"idx": 8, "title": "Chris Olah on what the hell is going on inside neural ...", "date": "", "ddg_snippet": "4 Aug 2021 — Chris is a machine learning researcher currently focused on neural network interpretability. Until last December he led OpenAI's interpretability team.", "subpage_snippet": "", "source": "80000hours.org", "link": "https://80000hours.org/podcast/episodes/chris-olah-interpretability-research/", "content": "4 Aug 2021 — Chris is a machine learning researcher currently focused on neural network interpretability. Until last December he led OpenAI's interpretability team."}
+{"idx": 9, "title": "arXiv:2502.10927v1 [cs.LG] 15 Feb 2025", "date": "", "ddg_snippet": "by M Saponati · Cited by 1 — We present a mathematical framework to analyze self- attention matrices by deriving the structures governing their weight updates. Using this ...", "subpage_snippet": "", "source": "stdm.github.io", "link": "https://stdm.github.io/downloads/papers/ArXiv_2025c.pdf", "content": "by M Saponati · Cited by 1 — We present a mathematical framework to analyze self- attention matrices by deriving the structures governing their weight updates. Using this ..."}

data/sampled_jsons/Olsson_et_al_2022_induction_heads_in-context_learning_year_2022.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[2209.11895] In - context Learning and Induction Heads", "date": "", "ddg_snippet": "We find that induction heads develop at precisely the same point as a sudden sharp increase in in - context learning ability, visible as a bump in the training loss.View a PDF of the paper titled In - context Learning and Induction Heads , by Catherine Olsson and 25 other authors.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2209.11895", "content": "We find that induction heads develop at precisely the same point as a sudden sharp increase in in - context learning ability, visible as a bump in the training loss.View a PDF of the paper titled In - context Learning and Induction Heads , by Catherine Olsson and 25 other authors."}
+{"idx": 1, "title": "[PDF] In - context Learning and Induction Heads | Semantic Scholar", "date": "", "ddg_snippet": "Beyond Induction Heads : In - Context Meta Learning Induces Multi-Phase Circuit Emergence.", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/In-context-Learning-and-Induction-Heads-Olsson-Elhage/c90a99eeb57019732a6cc996bb9eaf13faedf00f", "content": "Beyond Induction Heads : In - Context Meta Learning Induces Multi-Phase Circuit Emergence."}
+{"idx": 2, "title": "In - context Learning and Induction Heads", "date": "", "ddg_snippet": "Kaplan et al . is the origin of our approach for using loss at different token indices as a formalism for studying in - context learning . O’Connor & Andreas find that preserving word order in contexts is important, as the induction head hypothesis would suggest.", "subpage_snippet": "", "source": "transformer-circuits.pub", "link": "https://transformer-circuits.pub/2022/in-context-learning-and-induction-heads/index.html?ref=planned-obsolescence.org", "content": "Kaplan et al . is the origin of our approach for using loss at different token indices as a formalism for studying in - context learning . O’Connor & Andreas find that preserving word order in contexts is important, as the induction head hypothesis would suggest."}
+{"idx": 3, "title": "Induction Heads & In - Context Learning", "date": "", "ddg_snippet": "The paper \" In - context Learning and Induction Heads \" by Olsson et al . from Anthropic investigates the mechanisms behind in - context learning in large transformer models, focusing specifically on the role of \" induction heads \" within attention heads .", "subpage_snippet": "", "source": "www.emergentmind.com", "link": "https://www.emergentmind.com/papers/2209.11895", "content": "The paper \" In - context Learning and Induction Heads \" by Olsson et al . from Anthropic investigates the mechanisms behind in - context learning in large transformer models, focusing specifically on the role of \" induction heads \" within attention heads ."}
+{"idx": 4, "title": "What needs to go right for an induction head ? A mechanistic study of...", "date": "", "ddg_snippet": "A mechanistic study of in - context learning circuits and their formation.Then, there’s a phase change in the loss which corresponds to the formation of induction circuits, reproducing the finding of Olsson et al . ( 2022 ). b) Induction head strength for each Layer 2 head plotted over time.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=O8rrXl71D5", "content": "A mechanistic study of in - context learning circuits and their formation.Then, there’s a phase change in the loss which corresponds to the formation of induction circuits, reproducing the finding of Olsson et al . ( 2022 ). b) Induction head strength for each Layer 2 head plotted over time."}
+{"idx": 5, "title": "Induction Heads as an Essential Mechanism for Pattern Matching in", "date": "", "ddg_snippet": "2022 . In - context learning and induction heads . arXiv preprint arXiv:2209.11895. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al . 2019. Language models are unsupervised multitask learners .", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.findings-naacl.283.pdf", "content": "2022 . In - context learning and induction heads . arXiv preprint arXiv:2209.11895. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al . 2019. Language models are unsupervised multitask learners ."}
+{"idx": 6, "title": "In - Context Learning : An Alignment Survey — LessWrong", "date": "", "ddg_snippet": "In - Context Learning is like Supervised Learning Without Weight Updates.Later work supposed that the attention mechanism in transformers is responsible for ICL, specifically, the ‘ induction head ’ ( Olsson et al ., 2022 ).", "subpage_snippet": "", "source": "www.lesswrong.com", "link": "https://www.lesswrong.com/posts/KCbRx4DhR7puBvGkX/in-context-learning-an-alignment-survey", "content": "In - Context Learning is like Supervised Learning Without Weight Updates.Later work supposed that the attention mechanism in transformers is responsible for ICL, specifically, the ‘ induction head ’ ( Olsson et al ., 2022 )."}
+{"idx": 7, "title": "(PDF) N-Gram Induction Heads for In - Context RL: Improving Stability...", "date": "", "ddg_snippet": "in - context reinforcement learning method. (Laskin et al ., 2022 ).which authors denote as a higher order of induction head ( Olsson et al ., 2022 ). They explicitly. implement 1-, 2- and 3-gram attention layers and observe a significant reduction in perplexity.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/385528566_N-Gram_Induction_Heads_for_In-Context_RL_Improving_Stability_and_Reducing_Data_Needs", "content": "in - context reinforcement learning method. (Laskin et al ., 2022 ).which authors denote as a higher order of induction head ( Olsson et al ., 2022 ). They explicitly. implement 1-, 2- and 3-gram attention layers and observe a significant reduction in perplexity."}
+{"idx": 8, "title": "induction heads", "date": "", "ddg_snippet": "induction heads . notes from (Elhage et al ., 2021; Olsson et al ., 2022 ). see also: Transformers. virtual weights. In - context Learning and Induction Heads .", "subpage_snippet": "", "source": "aarnphm.xyz", "link": "https://aarnphm.xyz/thoughts/induction-heads", "content": "induction heads . notes from (Elhage et al ., 2021; Olsson et al ., 2022 ). see also: Transformers. virtual weights. In - context Learning and Induction Heads ."}
+{"idx": 9, "title": "Training Dynamics of In - Context Learning in Linear Attention", "date": "", "ddg_snippet": "In - context learning is critical to the flexibility of large language models, allowing them to solve tasks not explicitly included in their training data.This abrupt learning phase can reflect the formation of an induction head in the in - context learning setting ( Olsson et al .", "subpage_snippet": "", "source": "www.cl.uni-heidelberg.de", "link": "https://www.cl.uni-heidelberg.de/courses/ss25/the_mystery_of_in-context_learning_of_llms/papers/ZhangARXIV2025.pdf", "content": "In - context learning is critical to the flexibility of large language models, allowing them to solve tasks not explicitly included in their training data.This abrupt learning phase can reflect the formation of an induction head in the in - context learning setting ( Olsson et al ."}

data/sampled_jsons/SAFE_Finding_Sparse_and_Flat_Minima_to_Improve_Pruning_CrAM_99.5%_sparsity_validation_accuracy.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Safe: Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "Figure 4: Effects of different choices of penalty parameter schedules (a) on validation accuracy of sparsified/dense network (b-c) and the distance to the target sparsity constraint (d) over the training process of ResNet-20/CIFAR-10 using Safe on 95 % sparsity .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.06866v2", "content": "Figure 4: Effects of different choices of penalty parameter schedules (a) on validation accuracy of sparsified/dense network (b-c) and the distance to the target sparsity constraint (d) over the training process of ResNet-20/CIFAR-10 using Safe on 95 % sparsity ."}
+{"idx": 1, "title": "SAFE: Finding Sparse and Flat Minima to Improve Pruning publications | Dongyeop Lee - GitHub Pages Sparse optimization guided pruning for neural networks Safe: Finding Sparse and Flat Minima to Improve Pruning SAFE: Finding Sparse and Flat Minima to Improve Pruning SAFE: Finding Sparse and Flat Minima to Improve Pruning [2506.06866] SAFE : Finding Sparse and Flat Minima to Improve Pruni… Sparse optimization guided pruning for neural networks Sparse optimization guided pruning for neural networks Sparse optimization guided pruning for neural networks Sparse optimization guided pruning for neural networks Safe: Finding Sparse and Flat Minima to Improve Pruning Safe: Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "Jun 7, 2025 · Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress. Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity ... SAFE : Finding Sparse and Flat Minima to Improve Pruning Dongyeop Lee, Kwanhee Lee, Jinseok Chung, and Namhoon Lee ICML 2025 (spotlight), Jul 2025 Abs arXiv Code Poster Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent ... Mar 14, 2024 · This model enables us to pinpoint sparse weights more precisely than pretrained network. It automatically derives effective pruning criteria, and omits the step of fine-tuning. To implement the two-stage process in practice, we utilize stochastic gradient algorithm for the pretraining and design a threshold algorithm for pruning stage. We first formulate this as a sharpness-aware sparsity -constrained optimization problem: min max f(x + ε), ∥x∥0≤d∥ε∥2≤ρ where goal is to find a sparse solution x⋆ with atmost d non-zero elements that minimizes the objective function in the whole ε-neighborhood, i.e., seek flat minima . Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress.Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time.Specifically, we formulate pruning as a sparsity ... Motivated by recent studies in robust optimiza-tion, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective. Is pruning a sparsity-constrained optimization problem? Specifically, we formulate pruning as a sparsity-constrained optimization problem where flatness is encouraged as an objective. We solve it explicitly via an augmented Lagrange dual approach and extend it further by proposing a generalized projection operation, resulting in novel pruning methods called SAFE and its extension, SAFE +. What is sparse optimization Guided pruning? 3.1. The framework of sparse optimization guided pruning A typical procedure of network pruning mainly contains three stages: pretraining, pruning and fine-tuning . Specifically, an over-parameterized network is firstly prepared in the pretraining stage. Then this network is pruned according to some criteria in the pruning stage. What is a fine-grained sparse regularization? The other is the introduction of a fine-grained sparse regularization T ℓ 1, which not only enables the pruning model to locate the sparse weights more precisely than pretraining, but also helps to derive an effective pruning criterion automatically. How can sparse regularization improve network performance? By minimizing the dissimilarity of the outputs, excessive damages to network performances can be alleviated. The other is fine-grained sparse regularization, which enables the model to judge sparse weights more precisely than the pretraining stage . It also helps to derive an effective pruning criterion instead of setting the criterion manually. What is a pruning criterion? Once the neural network has been fully trained, pruning is performed to compress the network by removing unnecessary parameters. The pruning criterion is closely related to network performance, as it determines which parameters can be safely removed without significantly impacting accuracy . Can safe be applied to neural network pruning? We show that Safe can be applied to neural network pruning , and as a result, it not only obtains the desired flatness as well as high sparsity in the given deep model, but also enhances its generalization performance quite significantly, far better than the compared baselines, as validated across standard benchmarks. Figure 4: Effects of different choices of penalty parameter schedules (a) on validation accuracy of sparsified/dense network (b-c) and the distance to the target sparsity constraint (d) over the training process of ResNet-20/CIFAR-10 using Safe on 95 % sparsity .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2506.06866", "content": "Jun 7, 2025 · Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress. Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity ... SAFE : Finding Sparse and Flat Minima to Improve Pruning Dongyeop Lee, Kwanhee Lee, Jinseok Chung, and Namhoon Lee ICML 2025 (spotlight), Jul 2025 Abs arXiv Code Poster Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent ... Mar 14, 2024 · This model enables us to pinpoint sparse weights more precisely than pretrained network. It automatically derives effective pruning criteria, and omits the step of fine-tuning. To implement the two-stage process in practice, we utilize stochastic gradient algorithm for the pretraining and design a threshold algorithm for pruning stage. We first formulate this as a sharpness-aware sparsity -constrained optimization problem: min max f(x + ε), ∥x∥0≤d∥ε∥2≤ρ where goal is to find a sparse solution x⋆ with atmost d non-zero elements that minimizes the objective function in the whole ε-neighborhood, i.e., seek flat minima . Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress.Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time.Specifically, we formulate pruning as a sparsity ... Motivated by recent studies in robust optimiza-tion, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective. Is pruning a sparsity-constrained optimization problem? Specifically, we formulate pruning as a sparsity-constrained optimization problem where flatness is encouraged as an objective. We solve it explicitly via an augmented Lagrange dual approach and extend it further by proposing a generalized projection operation, resulting in novel pruning methods called SAFE and its extension, SAFE +. What is sparse optimization Guided pruning? 3.1. The framework of sparse optimization guided pruning A typical procedure of network pruning mainly contains three stages: pretraining, pruning and fine-tuning . Specifically, an over-parameterized network is firstly prepared in the pretraining stage. Then this network is pruned according to some criteria in the pruning stage. What is a fine-grained sparse regularization? The other is the introduction of a fine-grained sparse regularization T ℓ 1, which not only enables the pruning model to locate the sparse weights more precisely than pretraining, but also helps to derive an effective pruning criterion automatically. How can sparse regularization improve network performance? By minimizing the dissimilarity of the outputs, excessive damages to network performances can be alleviated. The other is fine-grained sparse regularization, which enables the model to judge sparse weights more precisely than the pretraining stage . It also helps to derive an effective pruning criterion instead of setting the criterion manually. What is a pruning criterion? Once the neural network has been fully trained, pruning is performed to compress the network by removing unnecessary parameters. The pruning criterion is closely related to network performance, as it determines which parameters can be safely removed without significantly impacting accuracy . Can safe be applied to neural network pruning? We show that Safe can be applied to neural network pruning , and as a result, it not only obtains the desired flatness as well as high sparsity in the given deep model, but also enhances its generalization performance quite significantly, far better than the compared baselines, as validated across standard benchmarks. Figure 4: Effects of different choices of penalty parameter schedules (a) on validation accuracy of sparsified/dense network (b-c) and the distance to the target sparsity constraint (d) over the training process of ResNet-20/CIFAR-10 using Safe on 95 % sparsity ."}
+{"idx": 2, "title": "Sparse optimization guided pruning for neural networks", "date": "", "ddg_snippet": "Mar 14, 2024 · This model enables us to pinpoint sparse weights more precisely than pretrained network. It automatically derives effective pruning criteria, and omits the step of fine-tuning. To implement the two-stage process in practice, we utilize stochastic gradient algorithm for the pretraining and design a threshold algorithm for pruning stage.", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0925231224000511", "content": "Mar 14, 2024 · This model enables us to pinpoint sparse weights more precisely than pretrained network. It automatically derives effective pruning criteria, and omits the step of fine-tuning. To implement the two-stage process in practice, we utilize stochastic gradient algorithm for the pretraining and design a threshold algorithm for pruning stage."}
+{"idx": 3, "title": "Safe: Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "We first formulate this as a sharpness-aware sparsity -constrained optimization problem: min max f(x + ε), ∥x∥0≤d∥ε∥2≤ρ where goal is to find a sparse solution x⋆ with atmost d non-zero elements that minimizes the objective function in the whole ε-neighborhood, i.e., seek flat minima .", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/media/icml-2025/Slides/46658.pdf", "content": "We first formulate this as a sharpness-aware sparsity -constrained optimization problem: min max f(x + ε), ∥x∥0≤d∥ε∥2≤ρ where goal is to find a sparse solution x⋆ with atmost d non-zero elements that minimizes the objective function in the whole ε-neighborhood, i.e., seek flat minima ."}
+{"idx": 4, "title": "SAFE: Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress.Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time.Specifically, we formulate pruning as a sparsity ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/LOG-postech/safe-torch", "content": "Sparsifying neural networks often suffers from seemingly inevitable performance degradation, and it remains challenging to restore the original performance despite much recent progress.Motivated by recent studies in robust optimization, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time.Specifically, we formulate pruning as a sparsity ..."}
+{"idx": 5, "title": "SAFE: Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "Motivated by recent studies in robust optimiza-tion, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2506.06866", "content": "Motivated by recent studies in robust optimiza-tion, we aim to tackle this problem by finding subnetworks that are both sparse and flat at the same time. Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective."}
+{"idx": 6, "title": "(PDF) SAFE : Finding Sparse and Flat Minima to Improve Pruning", "date": "", "ddg_snippet": "Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/392531034_SAFE_Finding_Sparse_and_Flat_Minima_to_Improve_Pruning", "content": "Specifically, we formulate pruning as a sparsity -constrained optimization problem where flatness is encouraged as an objective."}
+{"idx": 7, "title": "ICML Poster SAFE : Finding Sparse and Flat Minima to Improve ...", "date": "", "ddg_snippet": "further present that SAFE exhibits practically desirable properties such as robustness to noisy data and insensitivity to target sparsity , rendering its advantages for efficient and general use. Live content is unavailable.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46658", "content": "further present that SAFE exhibits practically desirable properties such as robustness to noisy data and insensitivity to target sparsity , rendering its advantages for efficient and general use. Live content is unavailable."}
+{"idx": 8, "title": "python - how to stop training if the sparse _ validation _ accuracy is...", "date": "", "ddg_snippet": "if logs[' sparse _categorical_ accuracy '] > 0.95: self.model.stop_training = True. I haven't practiced TF for quite a while, so this code might not work out of the box, but it's something to start with. More info can be found in official docs on writing custom callback and Callback class reference.", "subpage_snippet": "", "source": "stackoverflow.com", "link": "https://stackoverflow.com/questions/69472449/how-to-stop-training-if-the-sparse-validation-accuracy-is-higher-than-a-baseline", "content": "if logs[' sparse _categorical_ accuracy '] > 0.95: self.model.stop_training = True. I haven't practiced TF for quite a while, so this code might not work out of the box, but it's something to start with. More info can be found in official docs on writing custom callback and Callback class reference."}
+{"idx": 9, "title": "Sparsity in Deep Learning: Pruning and growth for efficient inference...", "date": "", "ddg_snippet": "70% medium sparsity . 90% moderate sparsity . 99.9% high sparsity .Dynamic sparsity combines pruning and regrowth of elements during the training process, while static sparsity prunes once before the training starts and does not update the model structure during training.", "subpage_snippet": "", "source": "htor.inf.ethz.ch", "link": "https://htor.inf.ethz.ch/publications/img/hoefler-sparsity-in-deep-learning.pdf", "content": "70% medium sparsity . 90% moderate sparsity . 99.9% high sparsity .Dynamic sparsity combines pruning and regrowth of elements during the training process, while static sparsity prunes once before the training starts and does not update the model structure during training."}

data/sampled_jsons/SF2M_neglects_growth_factors_balanced_diffusion_Schrödinger_bridge_year_2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Simulation-free Schrödinger bridges via score and flow matching", "date": "", "ddg_snippet": "Our method generalizes both the score-matching loss used in the training of diffusion models and the recently proposed flow matching loss used in the training of continuous normalizing flows. [SF] 2 M interprets continuous-time stochastic generative modeling as a Schrödinger bridge problem.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2307.03672", "content": "Our method generalizes both the score-matching loss used in the training of diffusion models and the recently proposed flow matching loss used in the training of continuous normalizing flows. [SF] 2 M interprets continuous-time stochastic generative modeling as a Schrödinger bridge problem."}
+{"idx": 1, "title": "GitHub - johndpope/Simplified-Diffusion-Schrodinger-Bridge", "date": "", "ddg_snippet": "This is the official implementation of the paper Simplified Diffusion Schrödinger Bridge . Abstract This paper introduces a novel theoretical simplification of the Diffusion Schrödinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance ...", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/johndpope/Simplified-Diffusion-Schrodinger-Bridge", "content": "This is the official implementation of the paper Simplified Diffusion Schrödinger Bridge . Abstract This paper introduces a novel theoretical simplification of the Diffusion Schrödinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance ..."}
+{"idx": 2, "title": "Simulation-free Schr\\\"odinger bridges via score and flow matching", "date": "", "ddg_snippet": "Our method generalizes both the score-matching loss used in the training of diffusion models and the recently proposed flow matching loss used in the training of continuous normalizing flows.", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/372234530_Simulation-free_Schrodinger_bridges_via_score_and_flow_matching", "content": "Our method generalizes both the score-matching loss used in the training of diffusion models and the recently proposed flow matching loss used in the training of continuous normalizing flows."}
+{"idx": 3, "title": "PDF Diffusion Schrödinger Bridge Matching - NeurIPS", "date": "", "ddg_snippet": "In contrast, the Schrödinger Bridge (SB) problem is a dynamic version of entropy-regularized OT (EOT) (Föllmer,1988;Léonard,2014b). The SB is the finite-time diffusion which admits as initial and terminal distributions the two distributions of interest and is the closest in Kullback-Leibler divergence to a reference diffusion .", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/c428adf74782c2092d254329b6b02482-Paper-Conference.pdf", "content": "In contrast, the Schrödinger Bridge (SB) problem is a dynamic version of entropy-regularized OT (EOT) (Föllmer,1988;Léonard,2014b). The SB is the finite-time diffusion which admits as initial and terminal distributions the two distributions of interest and is the closest in Kullback-Leibler divergence to a reference diffusion ."}
+{"idx": 4, "title": "PDF Generative modeling via Schrödinger bridge (basics on ... - vdb", "date": "", "ddg_snippet": "We present generative modeling via Schrödinger Bridge (SB). We introduce dynamic and static SB. We draw links with regularized Optimal Transport (OT).", "subpage_snippet": "", "source": "vdeborto.github.io", "link": "https://vdeborto.github.io/project/generative_modeling/session_5.pdf", "content": "We present generative modeling via Schrödinger Bridge (SB). We introduce dynamic and static SB. We draw links with regularized Optimal Transport (OT)."}
+{"idx": 5, "title": "[2303.16852] Diffusion Schrödinger Bridge Matching - arXiv.org", "date": "", "ddg_snippet": "Solving transport problems, i.e. finding a map transporting one given distribution to another, has numerous applications in machine learning. Novel mass transport methods motivated by generative modeling have recently been proposed, e.g. Denoising Diffusion Models (DDMs) and Flow Matching Models (FMMs) implement such a transport through a Stochastic Differential Equation (SDE) or an Ordinary ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2303.16852", "content": "Solving transport problems, i.e. finding a map transporting one given distribution to another, has numerous applications in machine learning. Novel mass transport methods motivated by generative modeling have recently been proposed, e.g. Denoising Diffusion Models (DDMs) and Flow Matching Models (FMMs) implement such a transport through a Stochastic Differential Equation (SDE) or an Ordinary ..."}
+{"idx": 6, "title": "Simulation-Free Schrödinger Bridges via Score and Flow Matching", "date": "", "ddg_snippet": "Simulation-Free Schrödinger Bridges via Score and Flow MatchingAlexander Y. Tong, Nikolay Malkin, Kilian Fatras, Lazar Atanackovic, Yanle...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v238/tong24a.html", "content": "Simulation-Free Schrödinger Bridges via Score and Flow MatchingAlexander Y. Tong, Nikolay Malkin, Kilian Fatras, Lazar Atanackovic, Yanle..."}
+{"idx": 7, "title": "PDF Diffusion Schrödinger Bridge with Applications to Score-Based ... - BIMSA", "date": "", "ddg_snippet": "3.3 Diffusion Schrödinger Bridge as Iterative Mean-Matching Proportional Fitting To approximate the IPF recursion defined in Proposition 2, we use similar approximations to Sec-tion 2.1.", "subpage_snippet": "", "source": "bimsa.net", "link": "https://bimsa.net/doc/notes/31316.pdf", "content": "3.3 Diffusion Schrödinger Bridge as Iterative Mean-Matching Proportional Fitting To approximate the IPF recursion defined in Proposition 2, we use similar approximations to Sec-tion 2.1."}
+{"idx": 8, "title": "Diffusion Schrödinger Bridge Matching - GitHub", "date": "", "ddg_snippet": "The goal of learning Schrödinger Bridges is to build a bridge between two distributions π 0 and π T such that the bridge is optimal in some sense. This transport setting covers many applications:", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/yuyang-shi/dsbm-pytorch", "content": "The goal of learning Schrödinger Bridges is to build a bridge between two distributions π 0 and π T such that the bridge is optimal in some sense. This transport setting covers many applications:"}
+{"idx": 9, "title": "A Unified Framework for Diffusion Bridge Problems: Flow Matching and ...", "date": "", "ddg_snippet": "The bridge problem is to find an SDE (or sometimes an ODE) that bridges two given distributions. The application areas of the bridge problem are enormous, among which the recent generative modeling (e.g., conditional or unconditional image generation) is the most popular. Also the famous Schrödinger bridge problem, a widely known problem for a century, is a special instance of the bridge ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2503.21756", "content": "The bridge problem is to find an SDE (or sometimes an ODE) that bridges two given distributions. The application areas of the bridge problem are enormous, among which the recent generative modeling (e.g., conditional or unconditional image generation) is the most popular. Also the famous Schrödinger bridge problem, a widely known problem for a century, is a special instance of the bridge ..."}

data/sampled_jsons/SLOPER4D_Dai_et_al.,_2023_abstract_year_2023.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "SLOPER4D: A Scene-Aware Dataset for Global 4D Human ...", "date": "", "ddg_snippet": "by Y Dai · 2023 · Cited by 51 — We present SLOPER4D , a novel scene-aware dataset collected in large urban environments to facilitate the re- search of global human pose estimation (GHPE) ...", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2023/papers/Dai_SLOPER4D_A_Scene-Aware_Dataset_for_Global_4D_Human_Pose_Estimation_CVPR_2023_paper.pdf", "content": "by Y Dai · 2023 · Cited by 51 — We present SLOPER4D , a novel scene-aware dataset collected in large urban environments to facilitate the re- search of global human pose estimation (GHPE) ..."}
+{"idx": 1, "title": "arXiv:2303.09095v1 [cs.CV] 16 Mar 2023", "date": "", "ddg_snippet": "by Y Dai · 2023 · Cited by 50 — In this paper, we propose a huge scene-aware dataset for sequential human pose estimation in urban environments, named SLOPER4D . To our ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2303.09095", "content": "by Y Dai · 2023 · Cited by 50 — In this paper, we propose a huge scene-aware dataset for sequential human pose estimation in urban environments, named SLOPER4D . To our ..."}
+{"idx": 2, "title": "OWP-LIO: A method and experimental study for pose ...", "date": "", "ddg_snippet": "by Y Deng · 2025 · Cited by 1 — Dai ( Dai et al ., 2023 ) et al . developed the SLOPER4D system for the estimation of human poses at the centimeter level in urban environments ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0029801824034450", "content": "by Y Deng · 2025 · Cited by 1 — Dai ( Dai et al ., 2023 ) et al . developed the SLOPER4D system for the estimation of human poses at the centimeter level in urban environments ..."}
+{"idx": 3, "title": "A 4D Radar Tensor-Based 3D Human Pose Estimation and ...", "date": "", "ddg_snippet": "by YH Ho · 2024 · Cited by 7 — Dai , Y., et al .: Sloper4d : a scene-aware dataset for global 4d human pose estimation in urban environments. In: Proceedings of the IEEE/CVF ...", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.1007/978-3-031-73036-8_7", "content": "by YH Ho · 2024 · Cited by 7 — Dai , Y., et al .: Sloper4d : a scene-aware dataset for global 4d human pose estimation in urban environments. In: Proceedings of the IEEE/CVF ..."}
+{"idx": 4, "title": "Pre-training a Density-Aware Pose Transformer for Robust ...", "date": "", "ddg_snippet": "18 Dec 2024 — SLOPER4D ( Dai et al . 2023 ) An IMU annotated dataset captured within a more realistic environment. A mobile LiDAR sensor is utilized to track a ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.13454v1", "content": "18 Dec 2024 — SLOPER4D ( Dai et al . 2023 ) An IMU annotated dataset captured within a more realistic environment. A mobile LiDAR sensor is utilized to track a ..."}
+{"idx": 5, "title": "A survey on deep 3D human pose estimation", "date": "", "ddg_snippet": "by RB Neupane · 2024 · Cited by 12 — For example, SLOPER4D ( Dai et al . 2023 ) offers a scene-aware dataset situated in an urban environment, providing contextually relevant data.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s10462-024-11019-3", "content": "by RB Neupane · 2024 · Cited by 12 — For example, SLOPER4D ( Dai et al . 2023 ) offers a scene-aware dataset situated in an urban environment, providing contextually relevant data."}
+{"idx": 6, "title": "FTP: A Human Pose Estimation Method Integrating Temporal ...", "date": "", "ddg_snippet": "Abstract . Human pose estimation is a significant research direction in the field of computer vision, with critical applications in human motion ... 15 pages", "subpage_snippet": "", "source": "raw.githubusercontent.com", "link": "https://raw.githubusercontent.com/mlresearch/v260/main/assets/cai25a/cai25a.pdf", "content": "Abstract . Human pose estimation is a significant research direction in the field of computer vision, with critical applications in human motion ... 15 pages"}
+{"idx": 7, "title": "HmPEAR: A Dataset for Human Pose Estimation and ...", "date": "", "ddg_snippet": "ABSTRACT . We introduce HmPEAR, a novel dataset crafted for advancing re- search in 3D Human Pose Estimation (3D HPE) and Human Action.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf/8bcbd5f375932ca306a5b19acd66d7f6ed085ef2.pdf", "content": "ABSTRACT . We introduce HmPEAR, a novel dataset crafted for advancing re- search in 3D Human Pose Estimation (3D HPE) and Human Action."}
+{"idx": 8, "title": "MM-Fi: multi-modal non-intrusive 4D human dataset for ...", "date": "", "ddg_snippet": "by J Yang · 2023 · Cited by 101 — In this paper, we propose MM-Fi, the first multi-modal non-intrusive 4D human dataset with 27 daily or rehabilitation action categories.", "subpage_snippet": "", "source": "dl.acm.org", "link": "https://dl.acm.org/doi/10.5555/3666122.3666944", "content": "by J Yang · 2023 · Cited by 101 — In this paper, we propose MM-Fi, the first multi-modal non-intrusive 4D human dataset with 27 daily or rehabilitation action categories."}
+{"idx": 9, "title": "MM-Fi: Multi-Modal Non-Intrusive 4D Human Dataset for ...", "date": "", "ddg_snippet": "2023 . ·. NeurIPS. Paper. Abstract . 4D human perception plays an essential role in a myriad of applications, such as home automation and metaverse avatar ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2023/poster/73714", "content": "2023 . ·. NeurIPS. Paper. Abstract . 4D human perception plays an essential role in a myriad of applications, such as home automation and metaverse avatar ..."}

data/sampled_jsons/Schulman_2017_PPO_Atari_A2C_performance_table_year_2017.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "DQN VS PPO VS A2C", "date": "", "ddg_snippet": "19 Jul 2024 — This study conducts a comparative analysis of three advanced Deep Reinforcement Learning models – Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2407.14151v1", "content": "19 Jul 2024 — This study conducts a comparative analysis of three advanced Deep Reinforcement Learning models – Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), ..."}
+{"idx": 1, "title": "PPO Explained: The RL Algorithm That Took the World by ...", "date": "", "ddg_snippet": "Proximal Policy Optimization offers a sweet spot: robust performance , relative ease of coding, and strong compatibility with various neural architectures.", "subpage_snippet": "", "source": "medium.com", "link": "https://medium.com/@vivek_tiwari_vt/ppo-explained-the-rl-algorithm-that-took-the-world-by-storm-8a245910b8ef", "content": "Proximal Policy Optimization offers a sweet spot: robust performance , relative ease of coding, and strong compatibility with various neural architectures."}
+{"idx": 2, "title": "Table 6 from Proximal Policy Optimization Algorithms", "date": "", "ddg_snippet": "Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). - \"Proximal Policy Optimization ...", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/Proximal-Policy-Optimization-Algorithms-Schulman-Wolski/dce6f9d4017b1785979e7520fd0834ef8cf02f4b/figure/11", "content": "Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). - \"Proximal Policy Optimization ..."}
+{"idx": 3, "title": "ECE 276 final report Advantage Actor Critic (A2C) with ...", "date": "", "ddg_snippet": "TABLE I REWARD PLOTS FOR DIFFERENT GAMES - \"ECE 276 final report Advantage Actor Critic ( A2C ) with Experience Replay\"", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/ECE-276-final-report-Advantage-Actor-Critic-(A2C)-Hu-Lai/86b974303d900c2ca6db92a7287cd7a29acd6692/figure/0", "content": "TABLE I REWARD PLOTS FOR DIFFERENT GAMES - \"ECE 276 final report Advantage Actor Critic ( A2C ) with Experience Replay\""}
+{"idx": 4, "title": "Proximal Policy Optimization Algorithms", "date": "", "ddg_snippet": "by J Schulman · 2017 · Cited by 30034 — Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). 12.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1707.06347", "content": "by J Schulman · 2017 · Cited by 30034 — Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). 12."}
+{"idx": 5, "title": "POLICY OPTIMIZATION BY GENETIC DISTILLATION", "date": "", "ddg_snippet": "by T Gangwani · 2018 · Cited by 47 — PPO performs 10 steps of full-batch gradient descent on the policy parameters using the same collected batch of simulation data, while A2C does a single descent ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=ByOnmlWC-", "content": "by T Gangwani · 2018 · Cited by 47 — PPO performs 10 steps of full-batch gradient descent on the policy parameters using the same collected batch of simulation data, while A2C does a single descent ..."}
+{"idx": 6, "title": "The 37 Implementation Details of Proximal Policy Optimization", "date": "", "ddg_snippet": "25 Mar 2022 — The following table collects the best-reported performance of PPO in popular RL libraries in Atari and MuJoCo environments. RL Library ...", "subpage_snippet": "", "source": "iclr-blog-track.github.io", "link": "https://iclr-blog-track.github.io/2022/03/25/ppo-implementation-details/", "content": "25 Mar 2022 — The following table collects the best-reported performance of PPO in popular RL libraries in Atari and MuJoCo environments. RL Library ..."}
+{"idx": 7, "title": "Cliff Diving: Exploring Reward Surfaces in Reinforcement ...", "date": "", "ddg_snippet": "the performance of PPO ( Schulman et al., 2017 ) and A2C . (Mnih et al ... Table of A2C and PPO's average percent change in reward after taking a few ... 33 pages", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v162/sullivan22a/sullivan22a.pdf", "content": "the performance of PPO ( Schulman et al., 2017 ) and A2C . (Mnih et al ... Table of A2C and PPO's average percent change in reward after taking a few ... 33 pages"}
+{"idx": 8, "title": "Potential Driven Reinforcement Learning for Hard ...", "date": "", "ddg_snippet": "by E Zhao · Cited by 8 — Actor-critic RL algorithms such as. A2C /A3C [Mnih et al., 2016] and PPO [ Schulman et al.,. 2017 ] are known for their sampling inefficiency. Prioritized.", "subpage_snippet": "", "source": "www.ijcai.org", "link": "https://www.ijcai.org/proceedings/2020/0290.pdf", "content": "by E Zhao · Cited by 8 — Actor-critic RL algorithms such as. A2C /A3C [Mnih et al., 2016] and PPO [ Schulman et al.,. 2017 ] are known for their sampling inefficiency. Prioritized."}
+{"idx": 9, "title": "(PDF) Proximal Policy Optimization Algorithms", "date": "", "ddg_snippet": "Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). arrow_forward_ios. Related papers.", "subpage_snippet": "", "source": "www.academia.edu", "link": "https://www.academia.edu/72572628/Proximal_Policy_Optimization_Algorithms", "content": "Table 6: Mean final scores (last 100 episodes) of PPO and A2C on Atari games after 40M game frames (10M timesteps). arrow_forward_ios. Related papers."}

data/sampled_jsons/Section_6_Scaling_Archetypal-SAE_norm_constraint_relaxation_term_year_2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Archetypal SAE: Adaptive and Stable Dictionary Learning for ...", "date": "", "ddg_snippet": "This implementation ensures that Wremains row-stochastic and that the deviation term Λ stays within the prescribed norm constraint . As shown in Figure4, RA-SAE achieves reconstruction performance on par with conventional Top-K SAEs while maintaining the stability benefits of the archetypal constraint . classArchetypalDictionary(nn.Module):", "subpage_snippet": "", "source": "konklab.fas.harvard.edu", "link": "https://konklab.fas.harvard.edu/Papers/Fel_2025_ICML.pdf", "content": "This implementation ensures that Wremains row-stochastic and that the deviation term Λ stays within the prescribed norm constraint . As shown in Figure4, RA-SAE achieves reconstruction performance on par with conventional Top-K SAEs while maintaining the stability benefits of the archetypal constraint . classArchetypalDictionary(nn.Module):"}
+{"idx": 1, "title": "Archetypal SAE: Adaptive and Stable Dictionary Learning ...", "date": "", "ddg_snippet": "18 Feb 2025 — RA- SAE further extend this framework by incorporating a small relaxation term ... scaling its norm to remain within a relaxation factor. Report ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2502.12892v1", "content": "18 Feb 2025 — RA- SAE further extend this framework by incorporating a small relaxation term ... scaling its norm to remain within a relaxation factor. Report ..."}
+{"idx": 2, "title": "Archetypal SAE: Adaptive and Stable Dictionary Learning ...", "date": "", "ddg_snippet": "18 Jun 2025 — Our work introduces a new method, Archetypal Sparse Autoencoders, that builds more reliable and interpretable concept dictionaries by geometrically anchoring ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=9v1eW8HgMU&noteId=jupnkmN3Zt", "content": "18 Jun 2025 — Our work introduces a new method, Archetypal Sparse Autoencoders, that builds more reliable and interpretable concept dictionaries by geometrically anchoring ..."}
+{"idx": 3, "title": "Archetypal SAE: Adaptive and Stable Dictionary Learning for ...", "date": "", "ddg_snippet": "A relaxed variant (RA-SAE) allows mild relaxation, matching standard SAEs in reconstruction while maintaining stabil- ity. Both integrate with any SAE variant ...", "subpage_snippet": "", "source": "konklab.sites.fas.harvard.edu", "link": "https://konklab.sites.fas.harvard.edu/Papers/Fel_2025_ICML.pdf", "content": "A relaxed variant (RA-SAE) allows mild relaxation, matching standard SAEs in reconstruction while maintaining stabil- ity. Both integrate with any SAE variant ..."}
+{"idx": 4, "title": "Archetypal SAE: Adaptive and Stable Dictionary Learning ...", "date": "", "ddg_snippet": "A relaxed variant (RA- SAE ) allows mild relaxation , matching standard SAEs in reconstruction while maintaining stability. Both integrate with any SAE variant ( ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46195", "content": "A relaxed variant (RA- SAE ) allows mild relaxation , matching standard SAEs in reconstruction while maintaining stability. Both integrate with any SAE variant ( ..."}
+{"idx": 5, "title": "Archetypal SAEs: Adaptive and Stable Dictionary Learning for ...", "date": "", "ddg_snippet": "20 Mar 2025 — A relaxed variant (RA- SAE ) allows mild relaxation , matching standard SAEs in reconstruction while maintaining stability. Both integrate with ...", "subpage_snippet": "", "source": "kempnerinstitute.harvard.edu", "link": "https://kempnerinstitute.harvard.edu/research/deeper-learning/archetypal-saes-adaptive-and-stable-dictionary-learning-for-concept-extraction-in-large-vision-models/", "content": "20 Mar 2025 — A relaxed variant (RA- SAE ) allows mild relaxation , matching standard SAEs in reconstruction while maintaining stability. Both integrate with ..."}
+{"idx": 6, "title": "Understanding sparse autoencoder scaling in the ...", "date": "", "ddg_snippet": "by EJ Michaud · 2025 — In this work, we develop a formal analysis of SAE scaling behavior and scaling laws.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.02565", "content": "by EJ Michaud · 2025 — In this work, we develop a formal analysis of SAE scaling behavior and scaling laws."}
+{"idx": 7, "title": "Regularization in Data-driven Predictive Control: A Convex ...", "date": "", "ddg_snippet": "Sep 12, 2025 · 1- norm penalties, projection-based regularizers, and a newly introduced causality-based regularizer, can be viewed as convex relaxations of their respective bi-level problems. This perspective clarifies the conceptual links between direct and indirect data-driven control and highlights how regularization implicitly enforces system identification.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.09027", "content": "Sep 12, 2025 · 1- norm penalties, projection-based regularizers, and a newly introduced causality-based regularizer, can be viewed as convex relaxations of their respective bi-level problems. This perspective clarifies the conceptual links between direct and indirect data-driven control and highlights how regularization implicitly enforces system identification."}
+{"idx": 8, "title": "Revisiting Lossless Convexification: Theoretical guarantees ...", "date": "", "ddg_snippet": "LCvx is a technique designed to efficiently solve a class of nonconvex continuous-time optimal control problems, where nonconvexity primarily arises from control constraints . The central idea of LCvx is to introduce slack variables to relax these nonconvex constraints , thereby converting the nonconvex feasible set into a convex one. Importantly, under mild conditions, Pontryagin’s maximum ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0005109825004327", "content": "LCvx is a technique designed to efficiently solve a class of nonconvex continuous-time optimal control problems, where nonconvexity primarily arises from control constraints . The central idea of LCvx is to introduce slack variables to relax these nonconvex constraints , thereby converting the nonconvex feasible set into a convex one. Importantly, under mild conditions, Pontryagin’s maximum ..."}
+{"idx": 9, "title": "Specialized Sparse Autoencoders for Interpreting Rare ...", "date": "", "ddg_snippet": "by A Muhamed · 2025 · Cited by 3 — TERM en- courages the SAE to learn compositional features leading to more interpretable representations (see Appendix L for a formal argument). ...", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/2025.findings-naacl.87.pdf", "content": "by A Muhamed · 2025 · Cited by 3 — TERM en- courages the SAE to learn compositional features leading to more interpretable representations (see Appendix L for a formal argument). ..."}

data/sampled_jsons/SimXRD-4M_Table_2_Bidirectional-GRU_baseline_F1_score_year_2023-2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "F- score - Wikipedia", "date": "", "ddg_snippet": "In statistical analysis of binary classification and information retrieval systems, the F- score or F-measure is a measure of predictive performance. It is calculated from the precision and recall of the test...", "subpage_snippet": "", "source": "en.wikipedia.org", "link": "https://en.wikipedia.org/wiki/F-score", "content": "In statistical analysis of binary classification and information retrieval systems, the F- score or F-measure is a measure of predictive performance. It is calculated from the precision and recall of the test..."}
+{"idx": 1, "title": "F-мера в машинном обучении: что это и как применяется на практике", "date": "", "ddg_snippet": "F-мера (F- score , F 1 - score ) представляет собой гармоническое среднее между точностью (precision) и полнотой (recall), что делает её сбалансированной метрикой для оценки качества бинарной классификации.", "subpage_snippet": "", "source": "sky.pro", "link": "https://sky.pro/wiki/analytics/f-mera-v-mashinnom-obuchenii-chto-eto-i-kak-primenyaetsya-na-praktike/", "content": "F-мера (F- score , F 1 - score ) представляет собой гармоническое среднее между точностью (precision) и полнотой (recall), что делает её сбалансированной метрикой для оценки качества бинарной классификации."}
+{"idx": 2, "title": "Метрики оценки моделей нейронных сетей для чайников / Хабр", "date": "", "ddg_snippet": "Micro‑averaging — суммирование TP, FP и FN по всем классам перед расчетом Precision, Recall и F 1 - score . Macro‑averaging — вычисление Precision, Recall и F 1 - score отдельно для каждого класса и усреднение полученных значений.", "subpage_snippet": "", "source": "habr.com", "link": "https://habr.com/ru/companies/slsoft/articles/893694/", "content": "Micro‑averaging — суммирование TP, FP и FN по всем классам перед расчетом Precision, Recall и F 1 - score . Macro‑averaging — вычисление Precision, Recall и F 1 - score отдельно для каждого класса и усреднение полученных значений."}
+{"idx": 3, "title": "3.4. Метрики и оценка: количественная оценка качества прогнозов", "date": "", "ddg_snippet": "Вы можете получить имена всех доступных моделей, вызвав get_ scorer _names. 3.4. 1 . 2 . Определение стратегии оценки с помощью метрических функций#.любые дополнительные параметры, такие как beta или labels в f 1 _ score .", "subpage_snippet": "", "source": "scikit-learn.ru", "link": "https://scikit-learn.ru/stable/modules/model_evaluation.html", "content": "Вы можете получить имена всех доступных моделей, вызвав get_ scorer _names. 3.4. 1 . 2 . Определение стратегии оценки с помощью метрических функций#.любые дополнительные параметры, такие как beta или labels в f 1 _ score ."}
+{"idx": 4, "title": "Feet to Meters (ft to m) Converter", "date": "", "ddg_snippet": "RAPID TABLES . Recommend Site. Send Feedback.", "subpage_snippet": "", "source": "www.rapidtables.com", "link": "https://www.rapidtables.com/convert/length/feet-to-meter.html", "content": "RAPID TABLES . Recommend Site. Send Feedback."}
+{"idx": 5, "title": "Spatial Audio Motion Understanding and Reasoning", "date": "", "ddg_snippet": "The Table 2 showcases the performance of our spatial audio encoder for seen and unseen audio classes. The DSAST model performs on par with the DCASE 2025 Task 3 baseline in F- score and relative distance error without using any synthetic training data.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2509.14666", "content": "The Table 2 showcases the performance of our spatial audio encoder for seen and unseen audio classes. The DSAST model performs on par with the DCASE 2025 Task 3 baseline in F- score and relative distance error without using any synthetic training data."}
+{"idx": 6, "title": "Desmos | Графический калькулятор", "date": "", "ddg_snippet": "\"a\" squared. a 2 . \"a\" Superscript, \"b\" , Baseline .", "subpage_snippet": "", "source": "www.desmos.com", "link": "https://www.desmos.com/calculator", "content": "\"a\" squared. a 2 . \"a\" Superscript, \"b\" , Baseline ."}
+{"idx": 7, "title": "2025 F 1 Drivers Standings - ESPN", "date": "", "ddg_snippet": "Visit ESPN for the complete 2025 F 1 Drivers standings. Includes winning percentage, home and away record, and current streak.", "subpage_snippet": "", "source": "www.espn.com", "link": "https://www.espn.com/f1/standings", "content": "Visit ESPN for the complete 2025 F 1 Drivers standings. Includes winning percentage, home and away record, and current streak."}
+{"idx": 8, "title": "[Roblox] How to get the “?” Badge in F 4 M 1 LY... - YouTube", "date": "", "ddg_snippet": "This is my first Roblox tutorial :') and please credit this since it’s originalGame link: https://web.roblox.com/games/6485013482/F 4 M 1 LY-D 1 NN3R-M33TUP-Weirdc...", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=0dsBnRacaSY", "content": "This is my first Roblox tutorial :') and please credit this since it’s originalGame link: https://web.roblox.com/games/6485013482/F 4 M 1 LY-D 1 NN3R-M33TUP-Weirdc..."}
+{"idx": 9, "title": "Упражнения на времена Present с ответами", "date": "", "ddg_snippet": "5 упражнений на тренировку времен группы презент в английском языке с ответами для продолжающих. Отработка Present Simple, Present Continuous, Present Perfect...", "subpage_snippet": "", "source": "EnglishWeb.ru", "link": "https://EnglishWeb.ru/grammar/present-tenses-exercises.html", "content": "5 упражнений на тренировку времен группы презент в английском языке с ответами для продолжающих. Отработка Present Simple, Present Continuous, Present Perfect..."}

data/sampled_jsons/Sinkhorn_algorithm_dual_coordinate_descent_entropy_optimization.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "(PDF) Accelerated Primal- Dual Coordinate Descent for...", "date": "", "ddg_snippet": "accelerated primal- dual randomized coordinate descent (APDRCD) algorithm . entropy minimization via the RAS algorithm . Mathematical Programming, 112(2):371–401", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/333324650_Accelerated_Primal-Dual_Coordinate_Descent_for_Computational_Optimal_Transport", "content": "accelerated primal- dual randomized coordinate descent (APDRCD) algorithm . entropy minimization via the RAS algorithm . Mathematical Programming, 112(2):371–401"}
+{"idx": 1, "title": "On the Acceleration of the Sinkhorn and Greenkhorn Algorithms for...", "date": "", "ddg_snippet": "Accelerated Primal- Dual Coordinate Descent for Computational Optimal Transport. We propose and analyze a novel accelerated primal- dual coordinate descen...Computational Optimal Transport: Complexity by Accelerated Gradient Descent Is Better Than by Sinkhorn 's Algorithm .", "subpage_snippet": "", "source": "deepai.org", "link": "https://deepai.org/publication/on-the-acceleration-of-the-sinkhorn-and-greenkhorn-algorithms-for-optimal-transport", "content": "Accelerated Primal- Dual Coordinate Descent for Computational Optimal Transport. We propose and analyze a novel accelerated primal- dual coordinate descen...Computational Optimal Transport: Complexity by Accelerated Gradient Descent Is Better Than by Sinkhorn 's Algorithm ."}
+{"idx": 2, "title": "Dual coordinate descent methods for logistic regression", "date": "", "ddg_snippet": "optimization algorithms for each classier. Coordinate descent methods, a classic optimization approach, have been very success-. fully applied to solve the dual form of large linear SVM (Hsieh et al. 2008). Motivated by.", "subpage_snippet": "", "source": "www.khoury.northeastern.edu", "link": "https://www.khoury.northeastern.edu/home/vip/teach/MLcourse/6_SVM_kernels/materials/Dual+coordinate+descent+methods+for+logistic+regression+and+maximum+entropy+models+-+Yu,+Huang,+Lin+-+2010.pdf", "content": "optimization algorithms for each classier. Coordinate descent methods, a classic optimization approach, have been very success-. fully applied to solve the dual form of large linear SVM (Hsieh et al. 2008). Motivated by."}
+{"idx": 3, "title": "Wasserstein distance via entropy regularization ( Sinkhorn algorithm )", "date": "", "ddg_snippet": "The Sinkhorn algorithm provides a computationally efficient method for approximating the Wasserstein distance, making it a practical choice for many applications, especially for large datasets.", "subpage_snippet": "", "source": "www.fabriziomusacchio.com", "link": "https://www.fabriziomusacchio.com/blog/2023-07-23-wasserstein_distance_sinkhorn/", "content": "The Sinkhorn algorithm provides a computationally efficient method for approximating the Wasserstein distance, making it a practical choice for many applications, especially for large datasets."}
+{"idx": 4, "title": "On the geometry and dynamical formulation of the Sinkhorn algorithm ...", "date": "", "ddg_snippet": "The Sinkhorn algorithm is a numerical method for the solution of optimal transport problems. Here, I give a brief survey of this algorithm , with a strong emphasis on its geometric origin: it is natural to view it as a discretization, by standard methods, of a non-linear integral equation.", "subpage_snippet": "", "source": "www.aimsciences.org", "link": "https://www.aimsciences.org/article/doi/10.3934/jcd.2024006", "content": "The Sinkhorn algorithm is a numerical method for the solution of optimal transport problems. Here, I give a brief survey of this algorithm , with a strong emphasis on its geometric origin: it is natural to view it as a discretization, by standard methods, of a non-linear integral equation."}
+{"idx": 5, "title": "Near-linear time approximation algorithms for optimal transport via...", "date": "", "ddg_snippet": "This paper demonstrates that this ambitious goal is in fact achieved by Cuturi's Sinkhorn Distances. This result relies on a new analysis of Sinkhorn iterations, which also directly suggests a new greedy coordinate descent algorithm Greenkhorn with the same theoretical guarantees.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper/2017/hash/491442df5f88c6aa018e86dac21d3606-Abstract.html", "content": "This paper demonstrates that this ambitious goal is in fact achieved by Cuturi's Sinkhorn Distances. This result relies on a new analysis of Sinkhorn iterations, which also directly suggests a new greedy coordinate descent algorithm Greenkhorn with the same theoretical guarantees."}
+{"idx": 6, "title": "Algorithms for Euclidean-regularised Optimal", "date": "", "ddg_snippet": "Keywords: Optimal transport · Euclidean regularisation · Sinkhorn algorithm · Primal- dual algorithm · Alternating optimisation .For instance, the Sinkhorn –Knopp algorithm for entropy -regularised OT requires computing the exponent with γ in the denominator.", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/pdf/algorithms-for-euclidean-regularised-optimal-transport-1ps1ygyx.pdf", "content": "Keywords: Optimal transport · Euclidean regularisation · Sinkhorn algorithm · Primal- dual algorithm · Alternating optimisation .For instance, the Sinkhorn –Knopp algorithm for entropy -regularised OT requires computing the exponent with γ in the denominator."}
+{"idx": 7, "title": "On a Combination of Alternating Minimization and Nesterov's Momentum", "date": "", "ddg_snippet": "The ubiquitous Sinkhorn ’s algorithm can be seen as an alternating minimization algorithm for the dual to the entropy -regularized optimal transport problem.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/1906.03622", "content": "The ubiquitous Sinkhorn ’s algorithm can be seen as an alternating minimization algorithm for the dual to the entropy -regularized optimal transport problem."}
+{"idx": 8, "title": "On accelerated alternating minimization", "date": "", "ddg_snippet": "The ubiquitous Sinkhorn ’s algorithm can be seen as an alternating minimization algorithm for the dual to the entropy -regularized optimal transport problem.Efciency of coordinate descent methods on huge-scale optimization problems.", "subpage_snippet": "", "source": "www.wias-berlin.de", "link": "https://www.wias-berlin.de/preprint/2695/wias_preprints_2695.pdf", "content": "The ubiquitous Sinkhorn ’s algorithm can be seen as an alternating minimization algorithm for the dual to the entropy -regularized optimal transport problem.Efciency of coordinate descent methods on huge-scale optimization problems."}
+{"idx": 9, "title": "Kantorovich-Initiative-multimarginals.dvi", "date": "", "ddg_snippet": "Note that Sinkhorn is also block coordinate descent in the dual . (4), indeed xing ϕ and maximizing the dual functional in ψ.and if ϕ solves (21), the measure Qϕ = Ke⊕ϕi m solves the multi-marginal entropy minimization", "subpage_snippet": "", "source": "www.mathtube.org", "link": "https://www.mathtube.org/sites/default/files/lecture-extra-files/Kantorovich-Initiative-multimarginals.pdf", "content": "Note that Sinkhorn is also block coordinate descent in the dual . (4), indeed xing ϕ and maximizing the dual functional in ψ.and if ϕ solves (21), the measure Qϕ = Ke⊕ϕi m solves the multi-marginal entropy minimization"}

data/sampled_jsons/Sync-Point_Drop_for_Efficient_Tensor_Parallelism_of_Large_Language_Models_Attention_Head_Grouping_Ex.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "TPLA: Tensor Parallel Latent Attention for Efficient", "date": "", "ddg_snippet": "... tensor parallelism limitations of MLA by dividing the attention heads and latent representations into g g groups (typically g g = 2), such that each ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.15881v1", "content": "... tensor parallelism limitations of MLA by dividing the attention heads and latent representations into g g groups (typically g g = 2), such that each ..."}
+{"idx": 1, "title": "FFN Fusion: Rethinking Sequential Computation in Large Language", "date": "", "ddg_snippet": "... technique that reduces sequential computation in large language models by identifying and exploiting natural opportunities for parallelization.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2503.18908v1", "content": "... technique that reduces sequential computation in large language models by identifying and exploiting natural opportunities for parallelization."}
+{"idx": 2, "title": "Optimizing LLM Performance with LM Cache: Architectures,", "date": "", "ddg_snippet": "... offers an in-depth technical research-minded view of LM Cache operates and how the caching machinery improves the efficiency, scalability, and cost ...", "subpage_snippet": "", "source": "hackernoon.com", "link": "https://hackernoon.com/optimizing-llm-performance-with-lm-cache-architectures-strategies-and-real-world-applications", "content": "... offers an in-depth technical research-minded view of LM Cache operates and how the caching machinery improves the efficiency, scalability, and cost ..."}
+{"idx": 3, "title": "Contents", "date": "", "ddg_snippet": "... Falcon-H1—an innovative series of large language models that feature a novel parallel hybrid architecture integrating Transformer-style attention ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.22448v1", "content": "... Falcon-H1—an innovative series of large language models that feature a novel parallel hybrid architecture integrating Transformer-style attention ..."}
+{"idx": 4, "title": "ICLR 2023 Papers", "date": "", "ddg_snippet": "Forward Super-Resolution: How Can GANs Learn Hierarchical Generative Models for Real-World Distributions ... of Data Selection for Real-world ...", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2023/papers.html", "content": "Forward Super-Resolution: How Can GANs Learn Hierarchical Generative Models for Real-World Distributions ... of Data Selection for Real-world ..."}
+{"idx": 5, "title": "LLM Inference - Hw-Sw Optimizations", "date": "", "ddg_snippet": "To train the model , the tokens are broken into the array of size batch_size (B) x sequence length , and these batches are fed to large neural ...", "subpage_snippet": "", "source": "community.juniper.net", "link": "https://community.juniper.net/blogs/sharada-yeluri/2024/02/20/llm-inference-hw-sw-optimizations", "content": "To train the model , the tokens are broken into the array of size batch_size (B) x sequence length , and these batches are fed to large neural ..."}
+{"idx": 6, "title": "ECCV 2024 Papers", "date": "", "ddg_snippet": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ...", "subpage_snippet": "", "source": "eccv2024.ecva.net", "link": "https://eccv2024.ecva.net/virtual/2024/papers.html?filter=titles", "content": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ..."}
+{"idx": 7, "title": "ECCV 2024 Papers", "date": "", "ddg_snippet": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ...", "subpage_snippet": "", "source": "eccv2024.ecva.net", "link": "https://eccv2024.ecva.net/virtual/2024/papers.html", "content": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ..."}
+{"idx": 8, "title": "ECCV 2024 Papers", "date": "", "ddg_snippet": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ...", "subpage_snippet": "", "source": "eccv.ecva.net", "link": "https://eccv.ecva.net/virtual/2024/papers.html", "content": "Optimizing Factorized Encoder Models : Time and Memory Reduction for Scalable and Efficient Action Recognition ... for Grounding Multimodal Large ..."}
+{"idx": 9, "title": "Artificial Intelligence | 0 articles | Tech News, Tutorials", "date": "", "ddg_snippet": "This book enables you to develop highly sought-after skills as corporate investment in generative AI soars.IntroductionAs the adoption of Retrieval ...", "subpage_snippet": "", "source": "www.packtpub.com", "link": "https://www.packtpub.com/en-co/learning/how-to-tutorials/tag/artificial-intelligence", "content": "This book enables you to develop highly sought-after skills as corporate investment in generative AI soars.IntroductionAs the adoption of Retrieval ..."}

data/sampled_jsons/The_Randomized_Midpoint_Method_for_Log-Concave_Sampling_Shen_Lee_abstract_method.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "[1909.05503] The Randomized Midpoint Method for Log-Concave", "date": "", "ddg_snippet": "View a PDF of the paper titled The Randomized Midpoint Method for Log - Concave Sampling , by Ruoqi Shen and 1 other authors", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/1909.05503", "content": "View a PDF of the paper titled The Randomized Midpoint Method for Log - Concave Sampling , by Ruoqi Shen and 1 other authors"}
+{"idx": 1, "title": "Beyond Log-Concavity and Score Regularity: Improved Convergence", "date": "", "ddg_snippet": "The generation process begins with random noise and iteratively refines it into meaningful samples , such as images or sounds, through a denoising ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2501.02298v4", "content": "The generation process begins with random noise and iteratively refines it into meaningful samples , such as images or sounds, through a denoising ..."}
+{"idx": 2, "title": "[2406.00924] Faster Diffusion Sampling with Randomized", "date": "", "ddg_snippet": "In this work, we propose a new scheme inspired by Shen and Lee s randomized midpoint method for log - concave sampling ~\\cite{ShenL19}.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2406.00924", "content": "In this work, we propose a new scheme inspired by Shen and Lee s randomized midpoint method for log - concave sampling ~\\cite{ShenL19}."}
+{"idx": 3, "title": "HMC and Underdamped Langevin United in the Unadjusted Convex", "date": "", "ddg_snippet": "Then, a stochastic gradient version of the samplers is considered, for which dimension-free convergence rates are established for log - concave smooth ...", "subpage_snippet": "", "source": "epubs.siam.org", "link": "https://epubs.siam.org/doi/10.1137/23M1608963?cookieSet=1", "content": "Then, a stochastic gradient version of the samplers is considered, for which dimension-free convergence rates are established for log - concave smooth ..."}
+{"idx": 4, "title": "Optimal Underdamped Langevin MCMC Method - NIPS - PDF Document", "date": "", "ddg_snippet": "In particular, we apply our method to sample the strongly- log - concave distribution and obtain gradient complexity better than all existing gradient ...", "subpage_snippet": "", "source": "zbook.org", "link": "https://zbook.org/read/c9bb8e_optimal-underdamped-langevin-mcmc-method-nips.html", "content": "In particular, we apply our method to sample the strongly- log - concave distribution and obtain gradient complexity better than all existing gradient ..."}
+{"idx": 5, "title": "1 Introduction", "date": "", "ddg_snippet": "One of the most widely used Markov Chain Monte Carlo methods for sampling in statistics are Langevin algorithms , that allows one to sample from a ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.00941v1", "content": "One of the most widely used Markov Chain Monte Carlo methods for sampling in statistics are Langevin algorithms , that allows one to sample from a ..."}
+{"idx": 6, "title": "Convergence of Unadjusted Langevin in High Dimensions:", "date": "", "ddg_snippet": "For strongly log - concave distributions, ρ k h \\rho_{kh} converges in the W 2 W_{2} metric to the stationary distribution of ( 1.2 ), π h \\pi_{h ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2408.13115v2", "content": "For strongly log - concave distributions, ρ k h \\rho_{kh} converges in the W 2 W_{2} metric to the stationary distribution of ( 1.2 ), π h \\pi_{h ..."}
+{"idx": 7, "title": "Contraction rate estimates of stochastic gradient kinetic", "date": "", "ddg_snippet": "Dalalyan, Theoretical guarantees for approximate sampling from smooth and log - concave densities. ... Log - concave sampling : Metropolis-hastings ...", "subpage_snippet": "", "source": "www.esaim-m2an.org", "link": "https://www.esaim-m2an.org/articles/m2an/ref/2024/06/m2an230147/m2an230147.html", "content": "Dalalyan, Theoretical guarantees for approximate sampling from smooth and log - concave densities. ... Log - concave sampling : Metropolis-hastings ..."}
+{"idx": 8, "title": "Linda Cai", "date": "", "ddg_snippet": "... this work, we propose a new discretization scheme for diffusion models inspired by Shen and Lee s randomized midpoint method for log - concave sampling ...", "subpage_snippet": "", "source": "www.catalyzex.com", "link": "https://www.catalyzex.com/author/Linda+Cai", "content": "... this work, we propose a new discretization scheme for diffusion models inspired by Shen and Lee s randomized midpoint method for log - concave sampling ..."}
+{"idx": 9, "title": "Publications", "date": "", "ddg_snippet": "... Lee and Lorenzo Orecchia and Aaron Sidford}, title = {An Almost-Linear-Time Algorithm for Approximate Max Flow in Undirected Graphs, and its ...", "subpage_snippet": "", "source": "yintat.com", "link": "https://yintat.com/publications.html", "content": "... Lee and Lorenzo Orecchia and Aaron Sidford}, title = {An Almost-Linear-Time Algorithm for Approximate Max Flow in Undirected Graphs, and its ..."}

data/sampled_jsons/Theorem_3.1_lower_bound_Catoni_contextual_bandits_heavy-tailed.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Catoni Contextual Bandits are Robust to Heavy-tailed ...", "date": "", "ddg_snippet": "Q5: In Theorem 3.1 , the regret lower bound depends on the used policy. A5: The intuition is as follows. Consider two bandit problems, each with two arms.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=5IpVe9PH14&noteId=J3K6uYfoM5", "content": "Q5: In Theorem 3.1 , the regret lower bound depends on the used policy. A5: The intuition is as follows. Consider two bandit problems, each with two arms."}
+{"idx": 1, "title": "Catoni Contextual Bandits are Robust to Heavy-tailed ...", "date": "", "ddg_snippet": "The variance dependence in our theorem matches the lower bound in Theorem 3.1 . Specifically, for the deterministic case where σt = 0 for all t ∈ [T] , the ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46438", "content": "The variance dependence in our theorem matches the lower bound in Theorem 3.1 . Specifically, for the deterministic case where σt = 0 for all t ∈ [T] , the ..."}
+{"idx": 2, "title": "Catoni Contextual Bandits are Robust to Heavy-tailed ...", "date": "", "ddg_snippet": "by C Ye · Cited by 1 — Moreover, we demonstrate the optimal- ity of the leading-order term in our regret bound through a matching lower bound . ... Theorem 3.1 . For any integer T > 0, ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=5IpVe9PH14", "content": "by C Ye · Cited by 1 — Moreover, we demonstrate the optimal- ity of the leading-order term in our regret bound through a matching lower bound . ... Theorem 3.1 . For any integer T > 0, ..."}
+{"idx": 3, "title": "Differential Private Stochastic Optimization with Heavy- ...", "date": "", "ddg_snippet": "by P Zhao · 2024 · Cited by 5 — To be best of our knowledge, currently, it is unknown whether the minimax lower bound shown in [1] is achievable. ... ([1], Theorem 3.1 ) Define.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2408.09891", "content": "by P Zhao · 2024 · Cited by 5 — To be best of our knowledge, currently, it is unknown whether the minimax lower bound shown in [1] is achievable. ... ([1], Theorem 3.1 ) Define."}
+{"idx": 4, "title": "Catoni-style confidence sequences for heavy-tailed mean ...", "date": "", "ddg_snippet": "by H Wang · 2023 · Cited by 55 — [11, Theorem 3.1 ]. If we let { v t } grow, say in a rate of v t = Θ ̃ ... tail -symmetric CI, which clearly implies the minimax lower bound . □. Proof ...", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/pii/S0304414923001084", "content": "by H Wang · 2023 · Cited by 55 — [11, Theorem 3.1 ]. If we let { v t } grow, say in a rate of v t = Θ ̃ ... tail -symmetric CI, which clearly implies the minimax lower bound . □. Proof ..."}
+{"idx": 5, "title": "Nearly Optimal Catoni's M-estimator for Infinite Variance", "date": "", "ddg_snippet": "by S Bhatt · 2022 · Cited by 19 — (2016, Theorem 3.1 ). See (Lugosi and Mendelson, 2019) for an excellent sum ... A lower bound can be derived using symmetric arguments. Note that an ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v162/bhatt22b/bhatt22b.pdf", "content": "by S Bhatt · 2022 · Cited by 19 — (2016, Theorem 3.1 ). See (Lugosi and Mendelson, 2019) for an excellent sum ... A lower bound can be derived using symmetric arguments. Note that an ..."}
+{"idx": 6, "title": "Nearly Optimal Algorithms for Level Set Estimation", "date": "", "ddg_snippet": "linear lower bound in Theorem 3.1 . The added factor of log(|X |) stems from a union bound, while the depen- dence on log(∆−1 min) is an additional ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v151/mason22a/mason22a.pdf", "content": "linear lower bound in Theorem 3.1 . The added factor of log(|X |) stems from a union bound, while the depen- dence on log(∆−1 min) is an additional ..."}
+{"idx": 7, "title": "Learning Beyond the Standard Model (of Data)", "date": "", "ddg_snippet": "by N Tripuraneni · 2022 — ... lower bound on the difficulty of estimating. B. We instantiate the f ... Theorem 3.1 . Let ˆh be an empirical risk minimizer of ˆRtrain(·, ·) in (3.2) ...", "subpage_snippet": "", "source": "escholarship.org", "link": "https://escholarship.org/uc/item/5vn1z529", "content": "by N Tripuraneni · 2022 — ... lower bound on the difficulty of estimating. B. We instantiate the f ... Theorem 3.1 . Let ˆh be an empirical risk minimizer of ˆRtrain(·, ·) in (3.2) ..."}
+{"idx": 8, "title": "Robust Gradient Descent for Phase Retrieval - GitHub", "date": "", "ddg_snippet": "by A Buna — Next, we lower-bound λmin ∇2r(x) in the region ∥h∥ ≤ R ∥x∗∥. For this ... From this point, the proof is identical to the Proof of Theorem 3.1 .", "subpage_snippet": "", "source": "raw.githubusercontent.com", "link": "https://raw.githubusercontent.com/mlresearch/v258/main/assets/buna25a/buna25a.pdf", "content": "by A Buna — Next, we lower-bound λmin ∇2r(x) in the region ∥h∥ ≤ R ∥x∗∥. For this ... From this point, the proof is identical to the Proof of Theorem 3.1 ."}
+{"idx": 9, "title": "variance-aware-robust-reinforcement-learning-with-linear- ...", "date": "", "ddg_snippet": "by X Li · 2023 · Cited by 2 — lower bound [Dani et al., 2008]. Hence, our ... The following two lemmas are the counterpart lemmas of Theorem 3.1 under light- tail as-.", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/pdf/variance-aware-robust-reinforcement-learning-with-linear-1vx7dbts.pdf", "content": "by X Li · 2023 · Cited by 2 — lower bound [Dani et al., 2008]. Hence, our ... The following two lemmas are the counterpart lemmas of Theorem 3.1 under light- tail as-."}

data/sampled_jsons/Unbiased_Recommender_Learning_from_Implicit_Feedback_via_Weakly_Supervised_Learning_Equation_6_empir.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Efficient computation of feedback arc set at web-scale |", "date": "", "ddg_snippet": "The minimum feedback arc set problem is an NP-hard problem on graphs that seeks a minimum set of arcs which, when removed from the graph, leave it ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/312874966_Efficient_computation_of_feedback_arc_set_at_web-scale", "content": "The minimum feedback arc set problem is an NP-hard problem on graphs that seeks a minimum set of arcs which, when removed from the graph, leave it ..."}
+{"idx": 1, "title": "ICLR 2024 Schedule", "date": "", "ddg_snippet": "Understanding Augmentation-based Self- Supervised Representation Learning via RKHS Approximation and Regression ... Contrastive Learning : Implicitly ...", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2024/calendar", "content": "Understanding Augmentation-based Self- Supervised Representation Learning via RKHS Approximation and Regression ... Contrastive Learning : Implicitly ..."}
+{"idx": 2, "title": "ICLR 2021 Schedule", "date": "", "ddg_snippet": "MELR: Meta- Learning via Modeling Episode-Level Relationships for Few-Shot Learning ... What Can You Learn From Your Muscles? Learning Visual ...", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2021/calendar", "content": "MELR: Meta- Learning via Modeling Episode-Level Relationships for Few-Shot Learning ... What Can You Learn From Your Muscles? Learning Visual ..."}
+{"idx": 3, "title": "ICLR 2021 Papers", "date": "", "ddg_snippet": "FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/virtual/2021/papers.html", "content": "FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization"}
+{"idx": 4, "title": "Downloads", "date": "", "ddg_snippet": "Benefit of deep learning with non-convex noisy gradient descent: Provable excess risk bound and superiority to kernel methods", "subpage_snippet": "", "source": "iclr.cc", "link": "https://iclr.cc/Downloads/2021", "content": "Benefit of deep learning with non-convex noisy gradient descent: Provable excess risk bound and superiority to kernel methods"}
+{"idx": 5, "title": "ICML 2024 Schedule", "date": "", "ddg_snippet": "Weakly - Supervised Residual Evidential ... FedSC: Provable Federated Self- supervised Learning with Spectral Contrastive Objective over Non-i.i.d.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2024/calendar", "content": "Weakly - Supervised Residual Evidential ... FedSC: Provable Federated Self- supervised Learning with Spectral Contrastive Objective over Non-i.i.d."}
+{"idx": 6, "title": "ICML 2020 Schedule", "date": "", "ddg_snippet": "How We Leverage Machine Learning and AI to Develop Life-Changing Medicines - A Case Study with COVID-19. ... Learning for Financial Portfolio ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2020/calendar", "content": "How We Leverage Machine Learning and AI to Develop Life-Changing Medicines - A Case Study with COVID-19. ... Learning for Financial Portfolio ..."}
+{"idx": 7, "title": "Papiers les plus cités de Nips 2018 – Ludovic Arnold", "date": "", "ddg_snippet": "Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agents", "subpage_snippet": "", "source": "ludovicarnold.com", "link": "https://ludovicarnold.com/fr/papiers-les-plus-cites-de-nips-2018/", "content": "Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agents"}
+{"idx": 8, "title": "Frontiers | Constructive Preference Elicitation", "date": "", "ddg_snippet": "... learning the DM s preferences, which are unobserved, and using them to suggest progressively better recommendations until a satisfactory one is found.", "subpage_snippet": "", "source": "www.frontiersin.org", "link": "https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2017.00071/full", "content": "... learning the DM s preferences, which are unobserved, and using them to suggest progressively better recommendations until a satisfactory one is found."}
+{"idx": 9, "title": "Track: Poster Session 1", "date": "", "ddg_snippet": "... unbiased gradient estimates with stochastic rounding (SR), resulting in more accurate model updates.However, directly applying SR to MXFP4 can result ...", "subpage_snippet": "", "source": "virtual.aistats.org", "link": "https://virtual.aistats.org/virtual/2025/session/8798", "content": "... unbiased gradient estimates with stochastic rounding (SR), resulting in more accurate model updates.However, directly applying SR to MXFP4 can result ..."}

data/sampled_jsons/Video-ColBERT_Contextualized_Late_Interaction_for_Text-to-Video_Retrieval_references.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video ...", "date": "", "ddg_snippet": "Effect of Query Pad Token Choice. Visualization. Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval .gle vectors to represent the text query and the video (e.g. through mean pooling), then use a single dot product for similarity calculation at retrieval time.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2503.19009", "content": "Effect of Query Pad Token Choice. Visualization. Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval .gle vectors to represent the text query and the video (e.g. through mean pooling), then use a single dot product for similarity calculation at retrieval time."}
+{"idx": 1, "title": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video ...", "date": "", "ddg_snippet": "Video - ColBERT is built upon 3 main components: a fine-grained spatial and temporal token-wise interaction , query and visual expansions, and a dual sigmoid loss during training. We find that this interaction and training paradigm leads to strong individual, yet compatible...", "subpage_snippet": "", "source": "www.alphaxiv.org", "link": "https://www.alphaxiv.org/abs/2503.19009", "content": "Video - ColBERT is built upon 3 main components: a fine-grained spatial and temporal token-wise interaction , query and visual expansions, and a dual sigmoid loss during training. We find that this interaction and training paradigm leads to strong individual, yet compatible..."}
+{"idx": 2, "title": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video ...", "date": "", "ddg_snippet": "Reddy_ Video - ColBERT _ Contextualized _ Late _ Interaction _ for _ Text - to - Video _ Retrieval @CVPR2025@CVF.These representations lead to increases in performance on common text - to - video retrieval benchmarks compared to other bi-encoder methods.", "subpage_snippet": "", "source": "papers.cool", "link": "https://papers.cool/venue/Reddy_Video-ColBERT_Contextualized_Late_Interaction_for_Text-to-Video_Retrieval@CVPR2025@CVF", "content": "Reddy_ Video - ColBERT _ Contextualized _ Late _ Interaction _ for _ Text - to - Video _ Retrieval @CVPR2025@CVF.These representations lead to increases in performance on common text - to - video retrieval benchmarks compared to other bi-encoder methods."}
+{"idx": 3, "title": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video ...", "date": "", "ddg_snippet": "TokenBinder: Text - Video Retrieval with One-to-Many Alignment Paradigm. Video - ColBERT is built upon 3 main components: a fine-grained spatial and temporal token-wise interaction , query and visual expansions, and a dual sigmoid loss during training.", "subpage_snippet": "", "source": "www.connectedpapers.com", "link": "https://www.connectedpapers.com/main/bff2f91c763830a2d14dbbbeca150e92ede02323/Video+ColBERT:-Contextualized-Late-Interaction-for-Text+to+Video-Retrieval/graph", "content": "TokenBinder: Text - Video Retrieval with One-to-Many Alignment Paradigm. Video - ColBERT is built upon 3 main components: a fine-grained spatial and temporal token-wise interaction , query and visual expansions, and a dual sigmoid loss during training."}
+{"idx": 4, "title": "ContextIQ: A Multimodal Expert-Based Video Retrieval System for...", "date": "", "ddg_snippet": "[7] Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval . In this work, we tackle the problem of text - to - video retrieval (T2VR). AArun ReddyAAlexander Martin.", "subpage_snippet": "", "source": "www.bohrium.com", "link": "https://www.bohrium.com/paper-details/contextiq-a-multimodal-expert-based-video-retrieval-system-for-contextual-advertising/1058628209777573922-108597", "content": "[7] Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval . In this work, we tackle the problem of text - to - video retrieval (T2VR). AArun ReddyAAlexander Martin."}
+{"idx": 5, "title": "[PDF] VideoCLIP: Contrastive Pre-training for Zero-shot Video - Text ...", "date": "", "ddg_snippet": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval . Arun ReddyAlexander Martin. +7 authors.", "subpage_snippet": "", "source": "www.semanticscholar.org", "link": "https://www.semanticscholar.org/paper/VideoCLIP:-Contrastive-Pre-training-for-Zero-shot-Xu-Ghosh/821ad6c9f0fecb5fabb486a5a87a93b7ea65bcc0", "content": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval . Arun ReddyAlexander Martin. +7 authors."}
+{"idx": 6, "title": "[CVPR 2025] Video - ColBERT : Contextualized Late Interaction for ...", "date": "", "ddg_snippet": "О сервисе Прессе Авторские права Связаться с нами Авторам Рекламодателям...", "subpage_snippet": "", "source": "www.youtube.com", "link": "https://www.youtube.com/watch?v=TtcPgsih2TQ", "content": "О сервисе Прессе Авторские права Связаться с нами Авторам Рекламодателям..."}
+{"idx": 7, "title": "Alexander Martin - Google Scholar", "date": "", "ddg_snippet": "MultiVENT 2.0: A Massive Multilingual Benchmark for Event-Centric Video Retrieval . Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval .", "subpage_snippet": "", "source": "scholar.google.com", "link": "https://scholar.google.com/citations?user=6_4ethMAAAAJ&hl=en", "content": "MultiVENT 2.0: A Massive Multilingual Benchmark for Event-Centric Video Retrieval . Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval ."}
+{"idx": 8, "title": "Reno KRIZ | Doctor of Philosophy | University of Pennsylvania...", "date": "", "ddg_snippet": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval .How are we able to learn about complex current events just from short snippets of video?", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/profile/Reno-Kriz", "content": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval .How are we able to learn about complex current events just from short snippets of video?"}
+{"idx": 9, "title": "Johns Hopkins Data Science and AI Institute (@HopkinsDSAI) | Aguea", "date": "", "ddg_snippet": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval Arun Reddy, Alexander Martin, Eugene Yang, Andrew Yates, Kate Sanders, Kenton Murray, Reno Kriz, Celso M de Melo, Benjamin Van Durme, and Rama Chellappa arxiv.org/abs/2503.19009.", "subpage_snippet": "", "source": "aguea.net", "link": "https://aguea.net/HopkinsDSAI", "content": "Video - ColBERT : Contextualized Late Interaction for Text - to - Video Retrieval Arun Reddy, Alexander Martin, Eugene Yang, Andrew Yates, Kate Sanders, Kenton Murray, Reno Kriz, Celso M de Melo, Benjamin Van Durme, and Rama Chellappa arxiv.org/abs/2503.19009."}

data/sampled_jsons/concept_bottleneck_model_reduce_dependency_segmentation_foundation_model_2024_2025_year_2024.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "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 .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2412.11576", "content": "Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts ."}
+{"idx": 1, "title": "Language Guided Concept Bottleneck Models for Interpretable...", "date": "", "ddg_snippet": "Among these interpretable models , Concept Bottleneck Models [22, 48, 58, 61, 65] provide explanations of the model ’s decision-making process in a straightforward man-ner. CBMs are designed to be interpretable, incorporating an intermediate Concept Bottleneck Layer (CBL), where.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/Yu_Language_Guided_Concept_Bottleneck_Models_for_Interpretable_Continual_Learning_CVPR_2025_paper.pdf", "content": "Among these interpretable models , Concept Bottleneck Models [22, 48, 58, 61, 65] provide explanations of the model ’s decision-making process in a straightforward man-ner. CBMs are designed to be interpretable, incorporating an intermediate Concept Bottleneck Layer (CBL), where."}
+{"idx": 2, "title": "GitHub - yewsiang/ConceptBottleneck: Concept Bottleneck Models ...", "date": "", "ddg_snippet": "Concept Bottleneck Models , ICML 2020. Contribute to yewsiang/ConceptBottleneck development by creating an account on GitHub. Concept Bottleneck Models . Pang Wei Koh*, Thao Nguyen*, Yew Siang Tang*, Stephen Mussmann, Emma Pierson, Been Kim, and Percy Liang. ICML 2020.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/yewsiang/ConceptBottleneck", "content": "Concept Bottleneck Models , ICML 2020. Contribute to yewsiang/ConceptBottleneck development by creating an account on GitHub. Concept Bottleneck Models . Pang Wei Koh*, Thao Nguyen*, Yew Siang Tang*, Stephen Mussmann, Emma Pierson, Been Kim, and Percy Liang. ICML 2020."}
+{"idx": 3, "title": "Label-free Concept Bottleneck Models | OpenReview", "date": "", "ddg_snippet": "Label-free Concept Bottleneck Models Download PDF. Tuomas Oikarinen, Subhro Das, Lam M. Nguyen, Tsui-Wei Weng. Published: 01 Feb 2023, Last Modified: 04 Aug 2025 ICLR 2023 posterReaders: Everyone.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=FlCg47MNvBA", "content": "Label-free Concept Bottleneck Models Download PDF. Tuomas Oikarinen, Subhro Das, Lam M. Nguyen, Tsui-Wei Weng. Published: 01 Feb 2023, Last Modified: 04 Aug 2025 ICLR 2023 posterReaders: Everyone."}
+{"idx": 4, "title": "NeurIPS 2024 Spotlight Posters", "date": "", "ddg_snippet": "2025 .Then, to model high-order dependencies among these components, we propose a hypergraph-based relational reasoning module that models the intricate relations of nodes (slots) with structural constraints.", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/virtual/2024/events/spotlight-posters-2024", "content": "2025 .Then, to model high-order dependencies among these components, we propose a hypergraph-based relational reasoning module that models the intricate relations of nodes (slots) with structural constraints."}
+{"idx": 5, "title": "Latest 15 Papers - December 23, 2024 - Githubissues", "date": "", "ddg_snippet": "Semantics Foundation of Reductive Reasoning. 2024 -12-19.Adaptive Concept Bottleneck for Foundation Models Under Distribution Shifts.", "subpage_snippet": "", "source": "githubissues.com", "link": "https://githubissues.com/tangwen-qian/DailyArXiv/182", "content": "Semantics Foundation of Reductive Reasoning. 2024 -12-19.Adaptive Concept Bottleneck for Foundation Models Under Distribution Shifts."}
+{"idx": 6, "title": "Bottleneck Calculator | CPU GPU Performance Fix", "date": "", "ddg_snippet": "Find and fix performance bottlenecks in your gaming PC. Our free tool helps you identify if your CPU is bottlenecking your GPU or vice versa, with personalized upgrade recommendations.", "subpage_snippet": "", "source": "bottleneckcalculator.to", "link": "https://bottleneckcalculator.to/", "content": "Find and fix performance bottlenecks in your gaming PC. Our free tool helps you identify if your CPU is bottlenecking your GPU or vice versa, with personalized upgrade recommendations."}
+{"idx": 7, "title": "Cover Pages, 2025", "date": "", "ddg_snippet": "Fast Adaptation with Behavioral Foundation Models . Harshit Sikchi, Andrea Tirinzoni, Ahmed Touati, Yingchen Xu, Anssi Kanervisto, Scott Niekum, Amy Zhang, Alessandro LazaricKeywords: Off Policy Evalutaion, Interpretability, Concept Bottleneck Models , Reliable OPE. Summary.", "subpage_snippet": "", "source": "rlj.cs.umass.edu", "link": "https://rlj.cs.umass.edu/2025/RLJ_2025_CoverPages.pdf", "content": "Fast Adaptation with Behavioral Foundation Models . Harshit Sikchi, Andrea Tirinzoni, Ahmed Touati, Yingchen Xu, Anssi Kanervisto, Scott Niekum, Amy Zhang, Alessandro LazaricKeywords: Off Policy Evalutaion, Interpretability, Concept Bottleneck Models , Reliable OPE. Summary."}
+{"idx": 8, "title": "Addressing Concept Mislabeling in Concept Bottleneck Models ...", "date": "", "ddg_snippet": "Abstract. Concept Bottleneck Models (CBMs) propose to enhance the trustworthiness of AIsystems by constraining their decisions on a set of human-understandableconcepts.", "subpage_snippet": "", "source": "deeplearn.org", "link": "https://deeplearn.org/arxiv/635257/addressing-concept-mislabeling-in-concept-bottleneck-models-through-preference-optimization", "content": "Abstract. Concept Bottleneck Models (CBMs) propose to enhance the trustworthiness of AIsystems by constraining their decisions on a set of human-understandableconcepts."}
+{"idx": 9, "title": "Sascha Marton - Google Scholar", "date": "", "ddg_snippet": "2024 . Beyond Pixels: Enhancing LIME with Hierarchical Features and Segmentation Foundation Models . P Knab, S Marton, C Bartelt. ICLR 2025 Workshop on Foundation Models in the Wild, 2025 .", "subpage_snippet": "", "source": "scholar.google.com", "link": "https://scholar.google.com/citations?user=5PQJ3sEAAAAJ&hl=en", "content": "2024 . Beyond Pixels: Enhancing LIME with Hierarchical Features and Segmentation Foundation Models . P Knab, S Marton, C Bartelt. ICLR 2025 Workshop on Foundation Models in the Wild, 2025 ."}

data/sampled_jsons/cost_function_for_hierarchical_overlapping_clustering_year_2023.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "(PDF) Hierarchical Clustering", "date": "", "ddg_snippet": "To implement a hierarchical clustering algorithm, one has to choose a linkage function (single linkage, average linkage, complete linkage, Ward ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/314700681_Hierarchical_Clustering", "content": "To implement a hierarchical clustering algorithm, one has to choose a linkage function (single linkage, average linkage, complete linkage, Ward ..."}
+{"idx": 1, "title": "(PDF) Cognitive Manager for Hierarchical Cluster Networks Based", "date": "", "ddg_snippet": "Cognitive Manager for Hierarchical Cluster Networks Based on Multi-Stage Machine Method ... for data broadcasting in the control channel in order to ...", "subpage_snippet": "", "source": "www.researchgate.net", "link": "https://www.researchgate.net/publication/271715914_Cognitive_Manager_for_Hierarchical_Cluster_Networks_Based_on_Multi-Stage_Machine_Method", "content": "Cognitive Manager for Hierarchical Cluster Networks Based on Multi-Stage Machine Method ... for data broadcasting in the control channel in order to ..."}
+{"idx": 2, "title": "graphs - Choice of algorithm for hierarchical clustering for", "date": "", "ddg_snippet": "Choice of algorithm for hierarchical clustering for minimizing network communication costs ... cost of making it a little harder to understand why it ...", "subpage_snippet": "", "source": "cs.stackexchange.com", "link": "https://cs.stackexchange.com/questions/28821/choice-of-algorithm-for-hierarchical-clustering-for-minimizing-network-communica", "content": "Choice of algorithm for hierarchical clustering for minimizing network communication costs ... cost of making it a little harder to understand why it ..."}
+{"idx": 3, "title": "On hierarchical clustering-based approach for RDDBS design |", "date": "", "ddg_snippet": "In fact, this is the main justification for using the hierarchical clustering method (HC), which performs better on tuples than attributes.", "subpage_snippet": "", "source": "journalofbigdata.springeropen.com", "link": "https://journalofbigdata.springeropen.com/articles/10.1186/s40537-023-00849-7", "content": "In fact, this is the main justification for using the hierarchical clustering method (HC), which performs better on tuples than attributes."}
+{"idx": 4, "title": "Hierarchical Level-Wise News Article Clustering via", "date": "", "ddg_snippet": "... in this work, we introduce a novel adaptation of multilingual hierarchical Matryoshka embeddings and a hierarchical clustering algorithm adapted for ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2506.00277v1", "content": "... in this work, we introduce a novel adaptation of multilingual hierarchical Matryoshka embeddings and a hierarchical clustering algorithm adapted for ..."}
+{"idx": 5, "title": "HiMA: Hierarchical Quantum Microarchitecture for Qubit-Scaling", "date": "", "ddg_snippet": "HiMA: Hierarchical Quantum Microarchitecture for Qubit-Scaling and Quantum Process-Level Parallelism ... Process-Based Hierarchical Trigger Mechanism ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2408.11311v1", "content": "HiMA: Hierarchical Quantum Microarchitecture for Qubit-Scaling and Quantum Process-Level Parallelism ... Process-Based Hierarchical Trigger Mechanism ..."}
+{"idx": 6, "title": "US11392621B1 - Unsupervised information-based hierarchical", "date": "", "ddg_snippet": "the hierarchical cluster analyzer uses RIL to create a plurality of candidate sets, each formed from the union of randomly selected subsets ⁇ i .", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US11392621B1/en", "content": "the hierarchical cluster analyzer uses RIL to create a plurality of candidate sets, each formed from the union of randomly selected subsets ⁇ i ."}
+{"idx": 7, "title": "US7287180B1 - Hardware independent hierarchical cluster of", "date": "", "ddg_snippet": "... hierarchical cluster of heterogeneous media servers using a hierarchical command beat protocol to synchronize distributed parallel computing systems ...", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US7287180B1/en", "content": "... hierarchical cluster of heterogeneous media servers using a hierarchical command beat protocol to synchronize distributed parallel computing systems ..."}
+{"idx": 8, "title": "ACP - The use of hierarchical clustering for the design of", "date": "", "ddg_snippet": "The use of hierarchical clustering for the design of optimized monitoring networks The use of hierarchical clustering for the design of optimized ...", "subpage_snippet": "", "source": "acp.copernicus.org", "link": "https://acp.copernicus.org/articles/18/6543/2018/", "content": "The use of hierarchical clustering for the design of optimized monitoring networks The use of hierarchical clustering for the design of optimized ..."}
+{"idx": 9, "title": "A dockerized framework for hierarchical frequency-based", "date": "", "ddg_snippet": "Nevertheless, the high computational cost and memory usage of baseline hierarchical clustering algorithms render them inappropriate for the vast ...", "subpage_snippet": "", "source": "journalofcloudcomputing.springeropen.com", "link": "https://journalofcloudcomputing.springeropen.com/articles/10.1186/s13677-019-0150-y", "content": "Nevertheless, the high computational cost and memory usage of baseline hierarchical clustering algorithms render them inappropriate for the vast ..."}

data/sampled_jsons/generalized_static_Schrödinger_bridge_problem_Luo_Tseng_extension.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Linear Convergence of Sinkhorn's Algorithm for Generalized Static ...", "date": "", "ddg_snippet": "In this paper, we present the generalized static Schrödinger bridge problem , establish its Kantorovich dual, and show that the associated generalized Sinkhorn algorithm con-verges linearly in a dimension independent manner under mild assumptions on the general divergence functional f , weight matrix W and margin (r,c).", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=0hrkN07DuO", "content": "In this paper, we present the generalized static Schrödinger bridge problem , establish its Kantorovich dual, and show that the associated generalized Sinkhorn algorithm con-verges linearly in a dimension independent manner under mild assumptions on the general divergence functional f , weight matrix W and margin (r,c)."}
+{"idx": 1, "title": "Linear convergence of Sinkhorn's algorithm for generalized static ...", "date": "", "ddg_snippet": "The classical static Schrödinger Bridge (SSB) problem , which seeks the most likely stochastic evolution between two marginal probability measures, has been studied extensively in the optimal transport and statistical physics communities, and more recently in machine learning communities in the surge of generative models.", "subpage_snippet": "", "source": "par.nsf.gov", "link": "https://par.nsf.gov/biblio/10625991-linear-convergence-sinkhorn-algorithm-generalized-static-schrodinger-bridge", "content": "The classical static Schrödinger Bridge (SSB) problem , which seeks the most likely stochastic evolution between two marginal probability measures, has been studied extensively in the optimal transport and statistical physics communities, and more recently in machine learning communities in the surge of generative models."}
+{"idx": 2, "title": "The Schrödinger Bridge between Gaussian Measures has a Closed Form", "date": "", "ddg_snippet": "The static optimal transport $(\\\\mathrm{OT})$ problem between Gaussians seeks to recover an optimal map, or more generally a coupling, to morph a Gaussian into another. It has been well studied and applied to a wide variety of tasks. Here we focus on the dynamic formulation of OT, also known as the Schrödinger bridge (SB) problem , which has recently seen a surge of interest in machine ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2202.05722", "content": "The static optimal transport $(\\\\mathrm{OT})$ problem between Gaussians seeks to recover an optimal map, or more generally a coupling, to morph a Gaussian into another. It has been well studied and applied to a wide variety of tasks. Here we focus on the dynamic formulation of OT, also known as the Schrödinger bridge (SB) problem , which has recently seen a surge of interest in machine ..."}
+{"idx": 3, "title": "PDF Contraction and Reaction in Generalized Schrödinger Bridges", "date": "", "ddg_snippet": "What is a Schrödinger Bridge Problem (SBP) Most likely evolution between 2 distributional snapshots", "subpage_snippet": "", "source": "abhishekhalder.org", "link": "https://abhishekhalder.org/AlexisAdvancementSlides01302024.pdf", "content": "What is a Schrödinger Bridge Problem (SBP) Most likely evolution between 2 distributional snapshots"}
+{"idx": 4, "title": "Different Formulations of the Schrödinger Bridge Problem", "date": "", "ddg_snippet": "The Schrödinger bridge problem could be viewed as applying entropy maximizing principle on finding the most likely path of a particle given its spatial distribution at two time instances.", "subpage_snippet": "", "source": "www.otwiki.xyz", "link": "https://www.otwiki.xyz/SchrodingerBridge.html", "content": "The Schrödinger bridge problem could be viewed as applying entropy maximizing principle on finding the most likely path of a particle given its spatial distribution at two time instances."}
+{"idx": 5, "title": "PDF The Schrödinger Bridge between Gaussian Measures has a Closed For", "date": "", "ddg_snippet": "The goal of our paper is to continue this pursuit of closed-form solutions and thereby extending these advantages to SB-based learning methods. For an overview of the method, see Fig. 1. To this end, we make the following contribu-tions: As our central result, we derive the closed-form expres-sions for Gaussian Schrödinger bridges (GSBs), i.e., SBs between Gaussian measures. This is a ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v206/bunne23a/bunne23a.pdf", "content": "The goal of our paper is to continue this pursuit of closed-form solutions and thereby extending these advantages to SB-based learning methods. For an overview of the method, see Fig. 1. To this end, we make the following contribu-tions: As our central result, we derive the closed-form expres-sions for Gaussian Schrödinger bridges (GSBs), i.e., SBs between Gaussian measures. This is a ..."}
+{"idx": 6, "title": "PDF Schrödinger Bridge Problem", "date": "", "ddg_snippet": "Machine-learning approaches for the empirical Schrodinger bridge problem Stochastic control liaisons: Richard Sinkhorn meets Gaspard Monge on a Schroedinger bridge", "subpage_snippet": "", "source": "bangyan101.github.io", "link": "https://bangyan101.github.io/files/sbp.pdf", "content": "Machine-learning approaches for the empirical Schrodinger bridge problem Stochastic control liaisons: Richard Sinkhorn meets Gaspard Monge on a Schroedinger bridge"}
+{"idx": 7, "title": "An Optimal Transport Approach for the Schrödinger Bridge Problem and ...", "date": "", "ddg_snippet": "This paper exploit the equivalence between the Schrödinger Bridge problem (Léonard in J Funct Anal 262:1879-1920, 2012; Nelson in Phys Rev 150:1079, 1966; Schrödinger in Über die umkehrung der naturgesetze. Verlag Akademie der wissenschaften in kommission bei Walter de Gruyter u, Company, 1931) and the entropy penalized optimal transport (Cuturi in: Advances in neural information ...", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s10915-020-01325-7", "content": "This paper exploit the equivalence between the Schrödinger Bridge problem (Léonard in J Funct Anal 262:1879-1920, 2012; Nelson in Phys Rev 150:1079, 1966; Schrödinger in Über die umkehrung der naturgesetze. Verlag Akademie der wissenschaften in kommission bei Walter de Gruyter u, Company, 1931) and the entropy penalized optimal transport (Cuturi in: Advances in neural information ..."}
+{"idx": 8, "title": "[2209.09893] Deep Generalized Schrödinger Bridge - arXiv.org", "date": "", "ddg_snippet": "We show that our proposed objective function provides necessary and sufficient conditions to the mean-field problem . Our method, named Deep Generalized Schrödinger Bridge (DeepGSB), not only outperforms prior methods in solving classical population navigation MFGs, but is also capable of solving 1000-dimensional opinion depolarization, setting ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2209.09893", "content": "We show that our proposed objective function provides necessary and sufficient conditions to the mean-field problem . Our method, named Deep Generalized Schrödinger Bridge (DeepGSB), not only outperforms prior methods in solving classical population navigation MFGs, but is also capable of solving 1000-dimensional opinion depolarization, setting ..."}
+{"idx": 9, "title": "Linear convergence of Sinkhorn's algorithm for generalized static ...", "date": "", "ddg_snippet": "Poster Linear convergence of Sinkhorn's algorithm for generalized static Schrödinger bridge Rahul Choudhary · Hanbaek Lyu West Exhibition Hall B2-B3 #W-506", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46671", "content": "Poster Linear convergence of Sinkhorn's algorithm for generalized static Schrödinger bridge Rahul Choudhary · Hanbaek Lyu West Exhibition Hall B2-B3 #W-506"}

data/sampled_jsons/language_model_initialization_pretraining_speech_models_benefits_warm_start_cold_start_year_2023.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Moshi: a speech-text foundation model for real-time dialogue", "date": "", "ddg_snippet": "Starting from a text language model backbone, Moshi generates speech as tokens from the residual quantizer of a neural audio codec, while modeling ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2410.00037v2", "content": "Starting from a text language model backbone, Moshi generates speech as tokens from the residual quantizer of a neural audio codec, while modeling ..."}
+{"idx": 1, "title": "Your Pretrained Model Tells the Difficulty Itself: A", "date": "", "ddg_snippet": "Curriculum learning is a widely adopted training strategy in natural language processing (NLP), where models are exposed to examples organized by ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.09758v1", "content": "Curriculum learning is a widely adopted training strategy in natural language processing (NLP), where models are exposed to examples organized by ..."}
+{"idx": 2, "title": "Scaling Properties of Speech Language Models", "date": "", "ddg_snippet": "The general GSLM pipeline is composed of three separately trained models : (i) a speech tokenizer, (ii) a language model , and (iii) a vocoder (token ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2404.00685v2", "content": "The general GSLM pipeline is composed of three separately trained models : (i) a speech tokenizer, (ii) a language model , and (iii) a vocoder (token ..."}
+{"idx": 3, "title": "US12299579B2 - Adversarial pretraining of machine learning", "date": "", "ddg_snippet": "the acts can also include performing one or more initial pretraining iterations of a learning process to train the machine learning model with the ...", "subpage_snippet": "", "source": "patents.google.com", "link": "https://patents.google.com/patent/US12299579B2/en", "content": "the acts can also include performing one or more initial pretraining iterations of a learning process to train the machine learning model with the ..."}
+{"idx": 4, "title": "Research Projects Ideas Natural Language Processing –", "date": "", "ddg_snippet": "What all do audio transformer models hear? Probing Acoustic Representations for Language Delivery and its Structure 13.", "subpage_snippet": "", "source": "projectsinventory.com", "link": "https://projectsinventory.com/research-projects-ideas-natural-language-processing/", "content": "What all do audio transformer models hear? Probing Acoustic Representations for Language Delivery and its Structure 13."}
+{"idx": 5, "title": "Improving Contextual ASR via Multi-grained Fusion with Large", "date": "", "ddg_snippet": "While end-to-end Automatic Speech Recognition (ASR) models have shown impressive performance in transcribing general speech , they often struggle to ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2507.12252v1", "content": "While end-to-end Automatic Speech Recognition (ASR) models have shown impressive performance in transcribing general speech , they often struggle to ..."}
+{"idx": 6, "title": "GRAFT: Gradient-Aware Fast MaxVol Technique for Dynamic Data", "date": "", "ddg_snippet": "... its effectiveness is tightly coupled to the fidelity of the proxy model , which may not always generalize well to the main model ’s learning dynamics.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.13653v1", "content": "... its effectiveness is tightly coupled to the fidelity of the proxy model , which may not always generalize well to the main model ’s learning dynamics."}
+{"idx": 7, "title": "Downloads", "date": "", "ddg_snippet": "Addressing Resource Scarcity across Sign Languages with Multilingual Pretraining and Unified-Vocabulary Datasets ... Labels: On the Benefit of ...", "subpage_snippet": "", "source": "neurips.cc", "link": "https://neurips.cc/Downloads/2022", "content": "Addressing Resource Scarcity across Sign Languages with Multilingual Pretraining and Unified-Vocabulary Datasets ... Labels: On the Benefit of ..."}
+{"idx": 8, "title": "Yu Zhang - ACL Anthology", "date": "", "ddg_snippet": "... model to stably condense style information into a compact latent space; 2) the Style and Duration Language Model (S&D-LM) concurrently predicts ...", "subpage_snippet": "", "source": "aclanthology.org", "link": "https://aclanthology.org/people/y/yu-zhang/", "content": "... model to stably condense style information into a compact latent space; 2) the Style and Duration Language Model (S&D-LM) concurrently predicts ..."}
+{"idx": 9, "title": "NeurIPS 2022 Papers", "date": "", "ddg_snippet": "TGEA 2.0: A Large-Scale Diagnostically Annotated Dataset with Benchmark Tasks for Text Generation of Pretrained Language Models", "subpage_snippet": "", "source": "nips.cc", "link": "https://nips.cc/virtual/2022/papers.html", "content": "TGEA 2.0: A Large-Scale Diagnostically Annotated Dataset with Benchmark Tasks for Text Generation of Pretrained Language Models"}

data/sampled_jsons/parallel_Picard_method_algorithm_1_log-concave_sampling_loop_structure.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Parallel Sampling of Diffusion Models", "date": "", "ddg_snippet": "by A Shih · 2023 · Cited by 96 — In Algorithm 1 we present the complete procedure of ParaDiGMS, incorporating sliding window over a batch, up-front sampling of noise, and tolerance of Picard ... 14 pages", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper_files/paper/2023/file/0d1986a61e30e5fa408c81216a616e20-Paper-Conference.pdf", "content": "by A Shih · 2023 · Cited by 96 — In Algorithm 1 we present the complete procedure of ParaDiGMS, incorporating sliding window over a batch, up-front sampling of noise, and tolerance of Picard ... 14 pages"}
+{"idx": 1, "title": "Parallel Simulation for Log-concave Sampling and Score-based ...", "date": "", "ddg_snippet": "Parallel Picard Method for Strongly Log-concave Sampling In this section, we present parallel Picard methods for strongly log-concave sampling ( Algorithm 1 ) and show it holds improved convergence rate w.r.t. the KL divergence and total variance (Theorem 4.2 and Corollary 4.3).", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/attachment?id=qtuxDy2qEB&name=pdf", "content": "Parallel Picard Method for Strongly Log-concave Sampling In this section, we present parallel Picard methods for strongly log-concave sampling ( Algorithm 1 ) and show it holds improved convergence rate w.r.t. the KL divergence and total variance (Theorem 4.2 and Corollary 4.3)."}
+{"idx": 2, "title": "Parallel simulation for sampling under isoperimetry and score ...", "date": "", "ddg_snippet": "Balance between time and Picard directions. Realed works in scientific computation. 4 Parallel Picard method for sampling under isoperimetry 4. 1 Algorithm 4.2 Theoretical Guarantees 4.3 Proof sketch of Theorem 4.2: Performance analysis of Algorithm 1 5 Parallel Picard method for sampling of diffusion models", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.07435", "content": "Balance between time and Picard directions. Realed works in scientific computation. 4 Parallel Picard method for sampling under isoperimetry 4. 1 Algorithm 4.2 Theoretical Guarantees 4.3 Proof sketch of Theorem 4.2: Performance analysis of Algorithm 1 5 Parallel Picard method for sampling of diffusion models"}
+{"idx": 3, "title": "PCM : Picard Consistency Model for Fast Parallel Sampling of ...", "date": "", "ddg_snippet": "To address this, a parallel sampling method based on Picard iteration was introduced, effectively reducing se-quential steps while ensuring exact convergence to the orig-inal output.", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2025/papers/So_PCM__Picard_Consistency_Model_for_Fast_Parallel_Sampling_of_CVPR_2025_paper.pdf", "content": "To address this, a parallel sampling method based on Picard iteration was introduced, effectively reducing se-quential steps while ensuring exact convergence to the orig-inal output."}
+{"idx": 4, "title": "Fast parallel sampling under isoperimetry", "date": "", "ddg_snippet": "1 . Introduction In this paper, we study the problem of designing fast parallel algorithms for sampling from continuous distributions π(x) ∝ exp(−V (x)) over x ∈ Rd. Designing eficient sampling algorithms is a ubiquitous problem, but the focus of most prior works has been to minimize sequential eficiency criteria, such as the total number of arithmetic operations or total queries to V ...", "subpage_snippet": "", "source": "proceedings.mlr.press", "link": "https://proceedings.mlr.press/v247/anari24a/anari24a.pdf", "content": "1 . Introduction In this paper, we study the problem of designing fast parallel algorithms for sampling from continuous distributions π(x) ∝ exp(−V (x)) over x ∈ Rd. Designing eficient sampling algorithms is a ubiquitous problem, but the focus of most prior works has been to minimize sequential eficiency criteria, such as the total number of arithmetic operations or total queries to V ..."}
+{"idx": 5, "title": "Log-Concave Sampling", "date": "", "ddg_snippet": "1 Introduction Log-concave sampling involves answering the question of — given a smooth function V — how many queries are needed to give an approximate sample from π ∝ exp(−V ) on Rd. This field is increasingly important in fields as diverse as optimization, operations research, physical simulation, and machine learn-ing, including the burgeoning field of score-generative models for ...", "subpage_snippet": "", "source": "michaeltang.xyz", "link": "http://michaeltang.xyz/papers/LogConcave.pdf", "content": "1 Introduction Log-concave sampling involves answering the question of — given a smooth function V — how many queries are needed to give an approximate sample from π ∝ exp(−V ) on Rd. This field is increasingly important in fields as diverse as optimization, operations research, physical simulation, and machine learn-ing, including the burgeoning field of score-generative models for ..."}
+{"idx": 6, "title": "Parallel simulation for sampling under isoperimetry and score ...", "date": "", "ddg_snippet": "Sep 25, 2024 · Keywords: parallel sampling , log-concave sampling , diffusion model, score-based generative modeling, ddpm TL;DR: We propose new parallel algorithms for sampling under isoperimetry and diffusion models, and rigorously prove that our algorithms enjoy $\\ log d$ iteration complexity.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=6Gb7VfTKY7", "content": "Sep 25, 2024 · Keywords: parallel sampling , log-concave sampling , diffusion model, score-based generative modeling, ddpm TL;DR: We propose new parallel algorithms for sampling under isoperimetry and diffusion models, and rigorously prove that our algorithms enjoy $\\ log d$ iteration complexity."}
+{"idx": 7, "title": "Parallel Simulation for Log-concave Sampling and Score ...", "date": "", "ddg_snippet": "Our parallel Picard method for strongly log-concave sampling is summarized in Algorithm 1. ... It keeps same parallel structure as that illustrated in Figure 1.", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/43916", "content": "Our parallel Picard method for strongly log-concave sampling is summarized in Algorithm 1. ... It keeps same parallel structure as that illustrated in Figure 1."}
+{"idx": 8, "title": "Parallel Simulation for Log-concave Sampling and Score-based ...", "date": "", "ddg_snippet": "Algorithm 1 Parallel Picard Method for sampling . 1: Input: x0 ∼ µ0, approximate score function s ≈ ∇f, the number of the iterations in outer loop J, the ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/pdf?id=qtuxDy2qEB", "content": "Algorithm 1 Parallel Picard Method for sampling . 1: Input: x0 ∼ µ0, approximate score function s ≈ ∇f, the number of the iterations in outer loop J, the ..."}
+{"idx": 9, "title": "Parallel simulation for sampling under isoperimetry and ...", "date": "", "ddg_snippet": "by H Zhou · 2024 · Cited by 4 — Our parallel Picard method for sampling under isoperimetry is summarized in Algorithm 1 . In Lines 1 –3, we generate the noises and fix them. In ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/pdf/2412.07435?", "content": "by H Zhou · 2024 · Cited by 4 — Our parallel Picard method for sampling under isoperimetry is summarized in Algorithm 1 . In Lines 1 –3, we generate the noises and fix them. In ..."}

data/sampled_jsons/per-instance_privacy_OR_per-example_privacy_machine_unlearning_year_2022.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "May 24, 2025 · We present a principled, per-instance approach to quantifying the difficulty of unlearning via fine-tuning. We begin by sharpening an analysis of noisy gradient descent for unlearning (Chien et al., 2024), obtaining a better utility- unlearning tradeoff by replacing worst-case privacy loss bounds with per-instance privacy losses (Thudi et al., 2024), each of which bounds the (Renyi) divergence ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/abs/2505.18786", "content": "May 24, 2025 · We present a principled, per-instance approach to quantifying the difficulty of unlearning via fine-tuning. We begin by sharpening an analysis of noisy gradient descent for unlearning (Chien et al., 2024), obtaining a better utility- unlearning tradeoff by replacing worst-case privacy loss bounds with per-instance privacy losses (Thudi et al., 2024), each of which bounds the (Renyi) divergence ..."}
+{"idx": 1, "title": "Leveraging Per-Example Privacy for Machine Unlearning", "date": "", "ddg_snippet": "Our results show that per-instance privacy levels computed from training dynamics reliably predict unlearning difficulty, offering a principled and practical way to assess unlearning performance.", "subpage_snippet": "", "source": "research.google", "link": "https://research.google/pubs/leveraging-per-example-privacy-for-machine-unlearning/", "content": "Our results show that per-instance privacy levels computed from training dynamics reliably predict unlearning difficulty, offering a principled and practical way to assess unlearning performance."}
+{"idx": 2, "title": "Privately Publishable Per-instance Privacy", "date": "", "ddg_snippet": "The remaining challenge is that the per-instance privacy loss is a function of the entire dataset; publishing it directly would negate the purpose of privately training a model in the first place! In this paper, we propose a methodology to privately release the per-instance privacy losses associated with private empirical risk minimization.", "subpage_snippet": "", "source": "proceedings.neurips.cc", "link": "https://proceedings.neurips.cc/paper/2021/file/9087b0efc7c7acd1ef7e153678809c77-Paper.pdf", "content": "The remaining challenge is that the per-instance privacy loss is a function of the entire dataset; publishing it directly would negate the purpose of privately training a model in the first place! In this paper, we propose a methodology to privately release the per-instance privacy losses associated with private empirical risk minimization."}
+{"idx": 3, "title": "Manage data privacy and protection - Microsoft Priva and ... Privately Publishable Per-instance Privacy - NIPS Understanding AI and Data Privacy: Key Principles Privately Publishable Per-instance Privacy - OpenReview", "date": "", "ddg_snippet": "At least 71% of countries/regions have passed or introduced data privacy legislation, according to the United Nations. Chances are good that your organization is based in, or has customers or employees in, regions with data privacy laws. A prominent example of a data privacy law with broad impact is the European Union's General Data Protection Regulation (GDPR). Many organizations are subject to multiple regulations that themselves are frequently updated. As the regulatory landscape expands, it's never been more critical for organizations to safeguard personal data while staying on top of changes. Failure to comply with data privacy laws and regulations can result in considerable financial penalties, legal and business repercussions, and erosion of your customers' trust. Data privacy and data protection go hand in hand. You can't have data privacy without data protection. Data protection helps protect personal data stored and managed by your organization from external threats and leakage. Data privacy provides another layer of sophisticated protection, which helps honor the purpose of personal data use and respects a data subject's rights throughout the data lifecycle. To help organizations regardless of size or location fortify their data privacy and protection posture, we offer robust and scalable solutions in Microsoft Priva and Microsoft Purview. See full list on learn.microsoft.com Microsoft Priva and Microsoft Purview provide a unified platform to help you comply with data privacy regulations. The complementary features in Purview risk and compliance solutions and Priva privacy management solutions help you assess the personal data within your organization, and provide automation and scalability to help reduce the complexity... See full list on learn.microsoft.com Use the guidance in these articles to help you assess risks and take appropriate action to protect personal data in your organization's environment. This guide comprises four overarching steps to help you understand how and when to use the appropriate Microsoft solution for meeting your organization's data privacy obligations. The steps in this solution are: 1.Assess your organization's data and risks: Start your journey by understanding your data and possible risks. 2.Protect and govern your data: Identify, categorize, and manage the data you need to protect. 3.Stay on track with privacy regulations: Monitor your progress in completing assessments and stay up-to-date as regulations change. 4.Respond to data privacy incidents and subject requests: Set up alerts so you can respond to privacy risks and automate your management of data subject requests. See full list on learn.microsoft.com See full list on learn.microsoft.com In this paper, we analyze the per-instance privacy loss of releasing a private empirical risk minimizer learned via objective perturbation, and propose a group of methods to privately and accurately publish the pDP losses at little to no additional privacy cost. 5 days ago · AI is now part of customer service, product design, operations, and decision making. That reach brings real benefits, and it also surfaces personal and sensitive data in new places. It raises the question: How do we ship useful AI while protecting people and meeting laws? This guide helps you understand AI and data privacy as one practice through core principles, common pitfalls, practical ... Nov 9, 2021 · We calculate the per-instance privacy loss of releasing a private empirical risk minimizer via objective perturbation, and propose methods to privately and accurately publish the per-instance privacy losses at little to no additional privacy cost.", "subpage_snippet": "", "source": "learn.microsoft.com", "link": "https://learn.microsoft.com/en-us/microsoft-365/solutions/data-privacy-protection?view=o365-worldwide", "content": "At least 71% of countries/regions have passed or introduced data privacy legislation, according to the United Nations. Chances are good that your organization is based in, or has customers or employees in, regions with data privacy laws. A prominent example of a data privacy law with broad impact is the European Union's General Data Protection Regulation (GDPR). Many organizations are subject to multiple regulations that themselves are frequently updated. As the regulatory landscape expands, it's never been more critical for organizations to safeguard personal data while staying on top of changes. Failure to comply with data privacy laws and regulations can result in considerable financial penalties, legal and business repercussions, and erosion of your customers' trust. Data privacy and data protection go hand in hand. You can't have data privacy without data protection. Data protection helps protect personal data stored and managed by your organization from external threats and leakage. Data privacy provides another layer of sophisticated protection, which helps honor the purpose of personal data use and respects a data subject's rights throughout the data lifecycle. To help organizations regardless of size or location fortify their data privacy and protection posture, we offer robust and scalable solutions in Microsoft Priva and Microsoft Purview. See full list on learn.microsoft.com Microsoft Priva and Microsoft Purview provide a unified platform to help you comply with data privacy regulations. The complementary features in Purview risk and compliance solutions and Priva privacy management solutions help you assess the personal data within your organization, and provide automation and scalability to help reduce the complexity... See full list on learn.microsoft.com Use the guidance in these articles to help you assess risks and take appropriate action to protect personal data in your organization's environment. This guide comprises four overarching steps to help you understand how and when to use the appropriate Microsoft solution for meeting your organization's data privacy obligations. The steps in this solution are: 1.Assess your organization's data and risks: Start your journey by understanding your data and possible risks. 2.Protect and govern your data: Identify, categorize, and manage the data you need to protect. 3.Stay on track with privacy regulations: Monitor your progress in completing assessments and stay up-to-date as regulations change. 4.Respond to data privacy incidents and subject requests: Set up alerts so you can respond to privacy risks and automate your management of data subject requests. See full list on learn.microsoft.com See full list on learn.microsoft.com In this paper, we analyze the per-instance privacy loss of releasing a private empirical risk minimizer learned via objective perturbation, and propose a group of methods to privately and accurately publish the pDP losses at little to no additional privacy cost. 5 days ago · AI is now part of customer service, product design, operations, and decision making. That reach brings real benefits, and it also surfaces personal and sensitive data in new places. It raises the question: How do we ship useful AI while protecting people and meeting laws? This guide helps you understand AI and data privacy as one practice through core principles, common pitfalls, practical ... Nov 9, 2021 · We calculate the per-instance privacy loss of releasing a private empirical risk minimizer via objective perturbation, and propose methods to privately and accurately publish the per-instance privacy losses at little to no additional privacy cost."}
+{"idx": 4, "title": "Privately Publishable Per-instance Privacy - NIPS", "date": "", "ddg_snippet": "In this paper, we analyze the per-instance privacy loss of releasing a private empirical risk minimizer learned via objective perturbation, and propose a group of methods to privately and accurately publish the pDP losses at little to no additional privacy cost.", "subpage_snippet": "", "source": "papers.nips.cc", "link": "https://papers.nips.cc/paper/2021/hash/9087b0efc7c7acd1ef7e153678809c77-Abstract.html", "content": "In this paper, we analyze the per-instance privacy loss of releasing a private empirical risk minimizer learned via objective perturbation, and propose a group of methods to privately and accurately publish the pDP losses at little to no additional privacy cost."}
+{"idx": 5, "title": "Understanding AI and Data Privacy: Key Principles", "date": "", "ddg_snippet": "5 days ago · AI is now part of customer service, product design, operations, and decision making. That reach brings real benefits, and it also surfaces personal and sensitive data in new places. It raises the question: How do we ship useful AI while protecting people and meeting laws? This guide helps you understand AI and data privacy as one practice through core principles, common pitfalls, practical ...", "subpage_snippet": "", "source": "www.protecto.ai", "link": "https://www.protecto.ai/blog/understanding-ai-and-data-privacy", "content": "5 days ago · AI is now part of customer service, product design, operations, and decision making. That reach brings real benefits, and it also surfaces personal and sensitive data in new places. It raises the question: How do we ship useful AI while protecting people and meeting laws? This guide helps you understand AI and data privacy as one practice through core principles, common pitfalls, practical ..."}
+{"idx": 6, "title": "Privately Publishable Per-instance Privacy - OpenReview", "date": "", "ddg_snippet": "Nov 9, 2021 · We calculate the per-instance privacy loss of releasing a private empirical risk minimizer via objective perturbation, and propose methods to privately and accurately publish the per-instance privacy losses at little to no additional privacy cost.", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=pPbrtkTHe9", "content": "Nov 9, 2021 · We calculate the per-instance privacy loss of releasing a private empirical risk minimizer via objective perturbation, and propose methods to privately and accurately publish the per-instance privacy losses at little to no additional privacy cost."}
+{"idx": 7, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "25 Jun 2025 — TL;DR: We show that per-instance privacy levels, computed during training, provide a practical and reliable way to predict unlearning difficulty in fine-tuning- ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=0A4Y9qRnu9&noteId=Zd6KsMzKb8", "content": "25 Jun 2025 — TL;DR: We show that per-instance privacy levels, computed during training, provide a practical and reliable way to predict unlearning difficulty in fine-tuning- ..."}
+{"idx": 8, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "We further demonstrate that per-instance privacy losses correlate well with several existing data difficulty metrics, while also identifying harder groups of ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46697", "content": "We further demonstrate that per-instance privacy losses correlate well with several existing data difficulty metrics, while also identifying harder groups of ..."}
+{"idx": 9, "title": "Leveraging Per-Instance Privacy for Machine Unlearning", "date": "", "ddg_snippet": "24 May 2025 — Leveraging Per-Instance Privacy for Machine Unlearning . Report issue ... Langevin unlearning : A new perspective of noisy gradient descent for ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2505.18786v1", "content": "24 May 2025 — Leveraging Per-Instance Privacy for Machine Unlearning . Report issue ... Langevin unlearning : A new perspective of noisy gradient descent for ..."}

data/sampled_jsons/remote_sensing_dataset_image_size_resolution_benchmark_comparison.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "A Large Dataset and Benchmark Towards Remote Sensing ...", "date": "", "ddg_snippet": "29 Jan 2025 — This paper introduces NUDT4MSTAR , a large-scale SAR dataset for remote sensing target recognition in the wild, including 40 vehicle target types ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2501.13354v2", "content": "29 Jan 2025 — This paper introduces NUDT4MSTAR , a large-scale SAR dataset for remote sensing target recognition in the wild, including 40 vehicle target types ..."}
+{"idx": 1, "title": "Remote sensing image super-resolution and object detection", "date": "", "ddg_snippet": "by Y Wang · 2022 · Cited by 224 — This paper reviews current datasets and object detection methods (deep learning-based) for remote sensing images .", "subpage_snippet": "", "source": "www.sciencedirect.com", "link": "https://www.sciencedirect.com/science/article/abs/pii/S0957417422002524", "content": "by Y Wang · 2022 · Cited by 224 — This paper reviews current datasets and object detection methods (deep learning-based) for remote sensing images ."}
+{"idx": 2, "title": "Revisiting Pre-trained Remote Sensing Model Benchmarks", "date": "", "ddg_snippet": "by I Corley · 2024 · Cited by 34 — We present a set of benchmark results across seven geospa- tial machine learning datasets commonly used as down- stream tasks for testing pre-trained model ...", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/CVPR2024W/PBVS/papers/Corley_Revisiting_Pre-trained_Remote_Sensing_Model_Benchmarks_Resizing_and_Normalization_Matters_CVPRW_2024_paper.pdf", "content": "by I Corley · 2024 · Cited by 34 — We present a set of benchmark results across seven geospa- tial machine learning datasets commonly used as down- stream tasks for testing pre-trained model ..."}
+{"idx": 3, "title": "A Comprehensive Benchmark for Optical Remote Sensing ...", "date": "", "ddg_snippet": "We compare three pre-trained SR models using different datasets and architectures to showcase the effectiveness of our benchmark . One of the models, SR4RS [18], ...", "subpage_snippet": "", "source": "www.techrxiv.org", "link": "https://www.techrxiv.org/users/760184/articles/735467/master/file/data/opensrtest/opensrtest.pdf", "content": "We compare three pre-trained SR models using different datasets and architectures to showcase the effectiveness of our benchmark . One of the models, SR4RS [18], ..."}
+{"idx": 4, "title": "CHOICE: Benchmarking the Remote Sensing Capabilities ...", "date": "", "ddg_snippet": "13 May 2025 — We propose CHOICE , an extensive benchmark designed to objectively evaluate the hierarchical remote sensing capabilities of VLMs.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2411.18145v3", "content": "13 May 2025 — We propose CHOICE , an extensive benchmark designed to objectively evaluate the hierarchical remote sensing capabilities of VLMs."}
+{"idx": 5, "title": "ESAOpenSR/opensr-test: A comprehensive benchmark for ...", "date": "", "ddg_snippet": "'opensr-test' provides a fair approach for SR benchmark . We provide three datasets carefully crafted to minimize spatial and spectral misalignment.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/ESAOpenSR/opensr-test", "content": "'opensr-test' provides a fair approach for SR benchmark . We provide three datasets carefully crafted to minimize spatial and spectral misalignment."}
+{"idx": 6, "title": "Globe230k: A Benchmark Dense-Pixel Annotation Dataset ...", "date": "", "ddg_snippet": "by Q Shi · 2023 · Cited by 29 — Recent advances in Landsat, ASTER, SPOT, Sentinel-2 satellite and processing capabilities facilitateobservation to higher resolution , 30-m resolution images ...", "subpage_snippet": "", "source": "spj.science.org", "link": "https://spj.science.org/doi/10.34133/remotesensing.0078", "content": "by Q Shi · 2023 · Cited by 29 — Recent advances in Landsat, ASTER, SPOT, Sentinel-2 satellite and processing capabilities facilitateobservation to higher resolution , 30-m resolution images ..."}
+{"idx": 7, "title": "A Real-World Benchmark for Sentinel-2 Multi-Image Super- ...", "date": "", "ddg_snippet": "by P Kowaleczko · 2023 · Cited by 30 — In this paper, we address the problem of evaluating MISR for S-2 images by introducing a new Multi-image Sentinel-2 SR (MuS2) benchmark.", "subpage_snippet": "", "source": "www.nature.com", "link": "https://www.nature.com/articles/s41597-023-02538-9", "content": "by P Kowaleczko · 2023 · Cited by 30 — In this paper, we address the problem of evaluating MISR for S-2 images by introducing a new Multi-image Sentinel-2 SR (MuS2) benchmark."}
+{"idx": 8, "title": "A Comprehensive Survey of Super-Resolution Remote ...", "date": "", "ddg_snippet": "by AD Vu · 2025 — We classify and evaluate these datasets across multiple dimensions : image count, spatial resolution , sensor and platform diversity, temporal ...", "subpage_snippet": "", "source": "ieeexplore.ieee.org", "link": "https://ieeexplore.ieee.org/iel8/6287639/10820123/11126996.pdf", "content": "by AD Vu · 2025 — We classify and evaluate these datasets across multiple dimensions : image count, spatial resolution , sensor and platform diversity, temporal ..."}
+{"idx": 9, "title": "OpenEarthMap: A Benchmark Dataset for Global High- ...", "date": "", "ddg_snippet": "by J Xia · 2023 · Cited by 122 — Abstract. We introduce OpenEarthMap, a benchmark dataset , for global high- resolution land cover mapping. OpenEarth-. Map consists of 2.2 million segments of ... 11 pages", "subpage_snippet": "", "source": "openaccess.thecvf.com", "link": "https://openaccess.thecvf.com/content/WACV2023/papers/Xia_OpenEarthMap_A_Benchmark_Dataset_for_Global_High-Resolution_Land_Cover_Mapping_WACV_2023_paper.pdf", "content": "by J Xia · 2023 · Cited by 122 — Abstract. We introduce OpenEarthMap, a benchmark dataset , for global high- resolution land cover mapping. OpenEarth-. Map consists of 2.2 million segments of ... 11 pages"}

data/sampled_jsons/sieve_maximum_likelihood_estimator_convergence_rate_distribution_regression_manifold.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "A Likelihood Based Approach to Distribution Regression Using", "date": "", "ddg_snippet": "Convergence rates of the Sieve MLE.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": "arxiv.org", "link": "https://arxiv.org/pdf/2410.02025", "content": "Convergence rates of the Sieve MLE.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": 1, "title": "ICML Poster A Likelihood Based Approach to Distribution ...", "date": "", "ddg_snippet": "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": "icml.cc", "link": "https://icml.cc/virtual/2025/poster/46645", "content": "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": 2, "title": "Sieve Maximum Likelihood Estimation of Partially Linear...", "date": "", "ddg_snippet": "We consider the sieve maximum likelihood estimation approach that approximates the cumulative baseline hazard function and nonparametric covariate effect with the monotone splines and ‐splines, respectively.", "subpage_snippet": "", "source": "scispace.com", "link": "https://scispace.com/papers/sieve-maximum-likelihood-estimation-of-partially-linear-63ake3wt90vc", "content": "We consider the sieve maximum likelihood estimation approach that approximates the cumulative baseline hazard function and nonparametric covariate effect with the monotone splines and ‐splines, respectively."}
+{"idx": 3, "title": "Maximum likelihood estimation of a", "date": "", "ddg_snippet": "The rate of convergence of the maximum likelihood estimator over a class is intimately connected to the “size” of the , measured via the notion of the entropy.", "subpage_snippet": "", "source": "core.ac.uk", "link": "https://core.ac.uk/download/pdf/1333027.pdf", "content": "The rate of convergence of the maximum likelihood estimator over a class is intimately connected to the “size” of the , measured via the notion of the entropy."}
+{"idx": 4, "title": "Convergence Rates of a Class of Multivariate Density", "date": "", "ddg_snippet": "Convergence Rates for Density rees. Convergence Rate of Sieve MLE. Posterior Concentration Rate of the Bayesian Estimator .", "subpage_snippet": "", "source": "jmlr.org", "link": "https://jmlr.org/papers/volume24/20-060/20-060.pdf", "content": "Convergence Rates for Density rees. Convergence Rate of Sieve MLE. Posterior Concentration Rate of the Bayesian Estimator ."}
+{"idx": 5, "title": "Consistency and asymptotic normality of sieve ML estimators ... | colab", "date": "", "ddg_snippet": "This paper considers sieve maximum likelihood estimation of seminonparametric (SNP) models with an unknown density function as non-Euclidean parameter, next to a finite-dimensional parameter vector. The density function involved is modeled via an infinite series expansion...", "subpage_snippet": "", "source": "colab.ws", "link": "https://colab.ws/articles/10.1017/s0266466614000036", "content": "This paper considers sieve maximum likelihood estimation of seminonparametric (SNP) models with an unknown density function as non-Euclidean parameter, next to a finite-dimensional parameter vector. The density function involved is modeled via an infinite series expansion..."}
+{"idx": 6, "title": "Semiparametric Likelihood Estimation with Clayton-Oakes Model for...", "date": "", "ddg_snippet": "To determine the sieve maximum likelihood estimate , we propose an EM algorithm by treating the unobservable latent variable as missing values.For the maximum likelihood estimator of parameters β and γ, we have.", "subpage_snippet": "", "source": "www.gavinpublishers.com", "link": "https://www.gavinpublishers.com/article/view/semiparametric-likelihood-estimation-with-clayton-oakes-model-for-multivariate-current-status-data", "content": "To determine the sieve maximum likelihood estimate , we propose an EM algorithm by treating the unobservable latent variable as missing values.For the maximum likelihood estimator of parameters β and γ, we have."}
+{"idx": 7, "title": "Proportional odds regression and sieve maximum likelihood ...", "date": "", "ddg_snippet": "Dive into the research topics of 'Proportional odds regression and sieve maximum likelihood estimation '.", "subpage_snippet": "", "source": "experts.umn.edu", "link": "https://experts.umn.edu/en/publications/proportional-odds-regression-and-sieve-maximum-likelihood-estimat-2", "content": "Dive into the research topics of 'Proportional odds regression and sieve maximum likelihood estimation '."}
+{"idx": 8, "title": "logistic2ph: Sieve maximum likelihood estimator (SMLE) for...", "date": "", "ddg_snippet": "This function returns the sieve maximum likelihood estimators (SMLE) for the logistic regression model from Lotspeich et al.Specifies the convergence criterion in the EM algorithm. The default value is 1E-4. This argument is optional.", "subpage_snippet": "", "source": "rdrr.io", "link": "https://rdrr.io/cran/sleev/man/logistic2ph.html", "content": "This function returns the sieve maximum likelihood estimators (SMLE) for the logistic regression model from Lotspeich et al.Specifies the convergence criterion in the EM algorithm. The default value is 1E-4. This argument is optional."}
+{"idx": 9, "title": "Donsker-type theorems for nonparametric maximum likelihood ...", "date": "", "ddg_snippet": "Rufibach, K., Dümbgen, L.: Maximum likelihood estimation of a log-concave density. Basic properties and consistency (preprint) (2004).Wong W.H., Shen X. (1995) Probability inequalities for likelihood ratios and convergence rates of sieve MLEs.", "subpage_snippet": "", "source": "link.springer.com", "link": "https://link.springer.com/article/10.1007/s00440-006-0031-4", "content": "Rufibach, K., Dümbgen, L.: Maximum likelihood estimation of a log-concave density. Basic properties and consistency (preprint) (2004).Wong W.H., Shen X. (1995) Probability inequalities for likelihood ratios and convergence rates of sieve MLEs."}

data/sampled_jsons/sitearxiv.orghtml2412.06329_2.5_Score_Based_Denoising.jsonl

@@ -0,0 +1,3 @@
+{"idx": 0, "title": "Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "Second, we identify a post-training score based denoising technique that allows one to remove the noise portion of the generated samples.Another component in our sampling pipeline is the score based denoising step.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.06329v3", "content": "Second, we identify a post-training score based denoising technique that allows one to remove the noise portion of the generated samples.Another component in our sampling pipeline is the score based denoising step."}
+{"idx": 1, "title": "Normalizing Flows are Capable Generative Models", "date": "", "ddg_snippet": "Second, we identify a post-training score based denoising technique that allows one to remove the noise portion of the generated samples.Another component in our sampling pipeline is the score based denoising step.", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2412.06329v2", "content": "Second, we identify a post-training score based denoising technique that allows one to remove the noise portion of the generated samples.Another component in our sampling pipeline is the score based denoising step."}
+{"idx": 2, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}

data/sampled_jsons/sitegithub.com_facebookresearchMask2Formerblobmaintrain_net.py.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Mask 2 Former / train _ net . py at main · facebookresearch / Mask 2 Former", "date": "", "ddg_snippet": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask 2 Former / train _ net . py at main · facebookresearch / Mask 2 Former .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/Mask2Former/blob/main/train_net.py", "content": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask 2 Former / train _ net . py at main · facebookresearch / Mask 2 Former ."}
+{"idx": 1, "title": "GitHub - facebookresearch/Mask2Former: Code release for ... Mask2Former/GETTING_STARTED.md at main - GitHub GitHub - facebookresearch/MaskFormer: Per-Pixel ... MaskFormer/README.md at main · facebookresearch ... - GitHub MaskFormer/train_net.py at main · facebookresearch ... - GitHub Mask2Former/README.md at main · facebookresearch ... - GitHub", "date": "", "ddg_snippet": "Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar [arXiv] [Project] [BibTeX] Features •A single architecture for panoptic, instance and semantic segmentation. See full list on github . com •Add Google Colab demo. •Video instance segmentation is now supported! Please check our tech report for more details. See full list on github . com See Preparing Datasets for Mask2Former . See Getting Started with Mask2Former . Run our demo using Colab: Integrated into Huggingface Spaces 🤗 using Gradio. Try out the Web Demo: See full list on github . com We provide a large set of baseline results and trained models available for download in the Mask2Former Model Zoo . See full list on github . com Shield: The majority of Mask2Former is licensed under a MIT License. See full list on github . com If you use Mask2Former in your research or wish to refer to the baseline results published in the Model Zoo, please use the following BibTeX entry. If you find the code useful, please also consider the following BibTeX entry. See full list on github . com Code is largely based on MaskFormer (https:// github . com / facebookresearch /MaskFormer). See full list on github . com We provide a script train _ net . py , that is made to train all the configs provided in Mask2Former . To train a model with \" train _ net . py \", first setup the corresponding datasets following datasets/README.md, then run: Mask2Former Checkout Mask2Former , a universal architecture based on MaskFormer meta-architecture that achieves SOTA on panoptic, instance and semantic segmentation across four popular datasets (ADE20K, Cityscapes, COCO, Mapillary Vistas). Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/README.md at main · facebookresearch /MaskFormer Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/ train _ net . py at main · facebookresearch/MaskFormer Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask2Former /README.md at main · facebookresearch/ Mask2Former", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/Mask2Former", "content": "Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexander Kirillov, Rohit Girdhar [arXiv] [Project] [BibTeX] Features •A single architecture for panoptic, instance and semantic segmentation. See full list on github . com •Add Google Colab demo. •Video instance segmentation is now supported! Please check our tech report for more details. See full list on github . com See Preparing Datasets for Mask2Former . See Getting Started with Mask2Former . Run our demo using Colab: Integrated into Huggingface Spaces 🤗 using Gradio. Try out the Web Demo: See full list on github . com We provide a large set of baseline results and trained models available for download in the Mask2Former Model Zoo . See full list on github . com Shield: The majority of Mask2Former is licensed under a MIT License. See full list on github . com If you use Mask2Former in your research or wish to refer to the baseline results published in the Model Zoo, please use the following BibTeX entry. If you find the code useful, please also consider the following BibTeX entry. See full list on github . com Code is largely based on MaskFormer (https:// github . com / facebookresearch /MaskFormer). See full list on github . com We provide a script train _ net . py , that is made to train all the configs provided in Mask2Former . To train a model with \" train _ net . py \", first setup the corresponding datasets following datasets/README.md, then run: Mask2Former Checkout Mask2Former , a universal architecture based on MaskFormer meta-architecture that achieves SOTA on panoptic, instance and semantic segmentation across four popular datasets (ADE20K, Cityscapes, COCO, Mapillary Vistas). Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/README.md at main · facebookresearch /MaskFormer Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/ train _ net . py at main · facebookresearch/MaskFormer Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask2Former /README.md at main · facebookresearch/ Mask2Former"}
+{"idx": 2, "title": "MaskFormer/ train _ net . py at main · facebookresearch /MaskFormer", "date": "", "ddg_snippet": "Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/ train _ net . py at main · facebookresearch /MaskFormer.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/MaskFormer/blob/main/train_net.py", "content": "Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/ train _ net . py at main · facebookresearch /MaskFormer."}
+{"idx": 3, "title": "Mask2Former/GETTING_STARTED.md at main - GitHub", "date": "", "ddg_snippet": "We provide a script train _ net . py , that is made to train all the configs provided in Mask2Former . To train a model with \" train _ net . py \", first setup the corresponding datasets following datasets/README.md, then run:", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/Mask2Former/blob/master/GETTING_STARTED.md", "content": "We provide a script train _ net . py , that is made to train all the configs provided in Mask2Former . To train a model with \" train _ net . py \", first setup the corresponding datasets following datasets/README.md, then run:"}
+{"idx": 4, "title": "GitHub - facebookresearch/MaskFormer: Per-Pixel ...", "date": "", "ddg_snippet": "Mask2Former Checkout Mask2Former , a universal architecture based on MaskFormer meta-architecture that achieves SOTA on panoptic, instance and semantic segmentation across four popular datasets (ADE20K, Cityscapes, COCO, Mapillary Vistas).", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/MaskFormer", "content": "Mask2Former Checkout Mask2Former , a universal architecture based on MaskFormer meta-architecture that achieves SOTA on panoptic, instance and semantic segmentation across four popular datasets (ADE20K, Cityscapes, COCO, Mapillary Vistas)."}
+{"idx": 5, "title": "MaskFormer/README.md at main · facebookresearch ... - GitHub", "date": "", "ddg_snippet": "Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/README.md at main · facebookresearch /MaskFormer", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/MaskFormer/blob/main/README.md?plain=1", "content": "Per-Pixel Classification is Not All You Need for Semantic Segmentation (NeurIPS 2021, spotlight) - MaskFormer/README.md at main · facebookresearch /MaskFormer"}
+{"idx": 6, "title": "Mask2Former/README.md at main · facebookresearch ... - GitHub", "date": "", "ddg_snippet": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask2Former /README.md at main · facebookresearch/ Mask2Former", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/Mask2Former/blob/main/README.md?plain=1", "content": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask2Former /README.md at main · facebookresearch/ Mask2Former"}
+{"idx": 7, "title": "Mask 2 Former / train _ net _video. py at main ...", "date": "", "ddg_snippet": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask 2 Former / train _ net _video. py at main · facebookresearch / Mask 2 Former .", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/Mask2Former/blob/main/train_net_video.py", "content": "Code release for \"Masked-attention Mask Transformer for Universal Image Segmentation\" - Mask 2 Former / train _ net _video. py at main · facebookresearch / Mask 2 Former ."}
+{"idx": 8, "title": "TimeSformer/tools/ train _ net . py at main ...", "date": "", "ddg_snippet": "The official pytorch implementation of our paper \"Is Space-Time Attention All You Need for Video Understanding?\" - TimeSformer/tools/ train _ net . py at main · facebookresearch /TimeSformer.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/TimeSformer/blob/main/tools/train_net.py", "content": "The official pytorch implementation of our paper \"Is Space-Time Attention All You Need for Video Understanding?\" - TimeSformer/tools/ train _ net . py at main · facebookresearch /TimeSformer."}
+{"idx": 9, "title": "MaskFormer/ train _ net . py at master · facebookresearch /MaskFormer", "date": "", "ddg_snippet": "from mask_former import.logger = logging.getLogger(\"detectron2. trainer \"). # In the end of training , run an evaluation with TTA.", "subpage_snippet": "", "source": "github.com", "link": "https://github.com/facebookresearch/MaskFormer/blob/master/train_net.py", "content": "from mask_former import.logger = logging.getLogger(\"detectron2. trainer \"). # In the end of training , run an evaluation with TTA."}

data/sampled_jsons/sitegithub.comfiveaiunderstanding_safety_finetuning_minGPT_config_transformer_blocks_n_layer.jsonl

@@ -0,0 +1 @@
+{"idx": 0, "title": "", "date": "", "ddg_snippet": "", "subpage_snippet": "", "source": "", "link": "", "content": ""}

data/sampled_jsons/sitezhenxiao.com_SPRING_Improving_the_Throughput_of_Sharding_Blockchain_via_Deep_Reinforcement_Learn.jsonl

@@ -0,0 +1,8 @@
+{"idx": 0, "title": "SPRING: Improving the Throughput of Sharding Blockchain via ...", "date": "", "ddg_snippet": "In this paper, we present SPRING , the first deep - reinforcement - learning (DRL)- based shard-ing framework for state placement . SPRING formulates the state placement as a Markov Decision Process, which considers the cross- shard transaction ratio and workload balancing and employs DRL to learn the efective state placement policy.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/WWW_Spring_camera_ready.pdf", "content": "In this paper, we present SPRING , the first deep - reinforcement - learning (DRL)- based shard-ing framework for state placement . SPRING formulates the state placement as a Markov Decision Process, which considers the cross- shard transaction ratio and workload balancing and employs DRL to learn the efective state placement policy."}
+{"idx": 1, "title": "AERO: Enhancing Sharding Blockchain via Deep Reinforcement ...", "date": "", "ddg_snippet": "SPRING uses DRL to optimize state placement in the sharding blockchain , reducing CSTXs and improving the blockchain throughput . We choose SPRING to compare the efectiveness of account migration and account allocation in improving CSTXs and load balance.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/WWW_AERO_camera_ready.pdf", "content": "SPRING uses DRL to optimize state placement in the sharding blockchain , reducing CSTXs and improving the blockchain throughput . We choose SPRING to compare the efectiveness of account migration and account allocation in improving CSTXs and load balance."}
+{"idx": 2, "title": "A Dual-Agent Scheduler for Distributed Deep Learning Jobs", "date": "", "ddg_snippet": "Conclusion nA distributed deep learning job scheduler is responsible for reducing both job completion time and training cost. ethods either ignore or cannot handle the system uncertainty and policy cooperatio nContributions: Ø We propose a dual-agent structure to learn the ordering and placement policies.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/talks/KDD2023-Xing.pdf", "content": "Conclusion nA distributed deep learning job scheduler is responsible for reducing both job completion time and training cost. ethods either ignore or cannot handle the system uncertainty and policy cooperatio nContributions: Ø We propose a dual-agent structure to learn the ordering and placement policies."}
+{"idx": 3, "title": "A Data Flow Framework with High Throughput and Low Latency ...", "date": "", "ddg_snippet": "On the other hand, data distribution occupies the bandwidth of consensus nodes and determines block propagation latency. Thus, its improvement requires a solution of high- throughput data production and low-latency data distribution. These two mutually influencing processes co-determine the overall throughput of permissioned blockchains.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/ICDCS2023-Hu.pdf", "content": "On the other hand, data distribution occupies the bandwidth of consensus nodes and determines block propagation latency. Thus, its improvement requires a solution of high- throughput data production and low-latency data distribution. These two mutually influencing processes co-determine the overall throughput of permissioned blockchains."}
+{"idx": 4, "title": "Understanding the Weakness of Large Language Model ...", "date": "", "ddg_snippet": "by M Xing · 2024 · Cited by 48 — SPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement. In Proceedings of the. ACM on Web Conference ...", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/kdd-ads0905-xing.pdf", "content": "by M Xing · 2024 · Cited by 48 — SPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement. In Proceedings of the. ACM on Web Conference ..."}
+{"idx": 5, "title": "Papers - Zhen Xiao", "date": "", "ddg_snippet": "SPRING : Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement In Proceedings of the Web Conference 2024 (WWW 2024), May 2024.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/", "content": "SPRING : Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement In Proceedings of the Web Conference 2024 (WWW 2024), May 2024."}
+{"idx": 6, "title": "Presto: Optimizing Cross-Shard Transactions in Sharded ...", "date": "", "ddg_snippet": "In this paper, we introduce Presto, a protocol designed for the account- state - based blockchain , which reduces the latency of handling cross- shard transactions. Presto leverages the concept of opti-mistic pre-execution along with pending tree to optimize cross- shard transaction processing.", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/papers/srds-2024.pdf", "content": "In this paper, we introduce Presto, a protocol designed for the account- state - based blockchain , which reduces the latency of handling cross- shard transactions. Presto leverages the concept of opti-mistic pre-execution along with pending tree to optimize cross- shard transaction processing."}
+{"idx": 7, "title": "Talks - Zhen Xiao", "date": "", "ddg_snippet": "Conference and Colloquium Presentations \"A Dual-Agent Scheduler for Distributed Deep Learning Jobs on Public Cloud via Reinforcement Learning \", Proc. of 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2023), Long Beach, CA, USA, August 2023. \"A Data Flow Framework with High Throughput and Low Latency for Permissioned Blockchains\", Proc. of 43rd IEEE International ...", "subpage_snippet": "", "source": "zhenxiao.com", "link": "http://zhenxiao.com/talks/", "content": "Conference and Colloquium Presentations \"A Dual-Agent Scheduler for Distributed Deep Learning Jobs on Public Cloud via Reinforcement Learning \", Proc. of 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2023), Long Beach, CA, USA, August 2023. \"A Data Flow Framework with High Throughput and Low Latency for Permissioned Blockchains\", Proc. of 43rd IEEE International ..."}

data/sampled_jsons/tightening_privacy_accounting_bounds_group_unlearning_differential_privacy_2024_2025.jsonl

@@ -0,0 +1,10 @@
+{"idx": 0, "title": "Public Data Assisted Differentially Private In-Context ...", "date": "", "ddg_snippet": "13 Sept 2025 — Through experiments, we demonstrate that our approach significantly improves the utility of private ICL with the assistance of public data .", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2509.10932v1", "content": "13 Sept 2025 — Through experiments, we demonstrate that our approach significantly improves the utility of private ICL with the assistance of public data ."}
+{"idx": 1, "title": "Convergent Privacy Loss of Noisy-SGD without Convexity ...", "date": "", "ddg_snippet": "by E Chien · Cited by 4 — We study the Differential Privacy (DP) guarantee of hidden-state Noisy-SGD algorithms over a bounded domain. Standard privacy analysis for ...", "subpage_snippet": "", "source": "openreview.net", "link": "https://openreview.net/forum?id=kjmLabjSE2", "content": "by E Chien · Cited by 4 — We study the Differential Privacy (DP) guarantee of hidden-state Noisy-SGD algorithms over a bounded domain. Standard privacy analysis for ..."}
+{"idx": 2, "title": "Unfair Unlearning? Accounting for Fairness in Machine ...", "date": "", "ddg_snippet": "by GO Barbulescu · 2018 — In this paper, we propose a principled variational framework for fair unlearning , which balances three competing objectives: (1) forgetting specific datapoints, ...", "subpage_snippet": "", "source": "kdd2025.kdd.org", "link": "https://kdd2025.kdd.org/wp-content/uploads/2025/07/paper_23.pdf", "content": "by GO Barbulescu · 2018 — In this paper, we propose a principled variational framework for fair unlearning , which balances three competing objectives: (1) forgetting specific datapoints, ..."}
+{"idx": 3, "title": "Stream-Native Machine Unlearning", "date": "", "ddg_snippet": "30 Aug 2025 — The Memory Pair offers a practical and theoretically grounded foundation for stream-native privacy -preservingunlearning. Future work will focus ...", "subpage_snippet": "", "source": "arxiv.org", "link": "https://arxiv.org/html/2508.10193v2", "content": "30 Aug 2025 — The Memory Pair offers a practical and theoretically grounded foundation for stream-native privacy -preservingunlearning. Future work will focus ..."}
+{"idx": 4, "title": "Free Record-Level Privacy Risk Evaluation Through ...", "date": "", "ddg_snippet": "by J Pollock · Cited by 1 — As the goal of the privacy auditing is to provide a tight esti- mate of the differential privacy guarantees, prior works typi- cally rely on ... 21 pages", "subpage_snippet": "", "source": "www.usenix.org", "link": "https://www.usenix.org/system/files/usenixsecurity25-pollock.pdf", "content": "by J Pollock · Cited by 1 — As the goal of the privacy auditing is to provide a tight esti- mate of the differential privacy guarantees, prior works typi- cally rely on ... 21 pages"}
+{"idx": 5, "title": "COLT 2025 Book of Abstracts", "date": "", "ddg_snippet": "Our analysis can be seen as a fingerprinting argument, one of the main techniques used to prove lower bounds in differential privacy . Most fingerprinting ...", "subpage_snippet": "", "source": "learningtheory.org", "link": "https://learningtheory.org/colt2025/abstracts.html", "content": "Our analysis can be seen as a fingerprinting argument, one of the main techniques used to prove lower bounds in differential privacy . Most fingerprinting ..."}
+{"idx": 6, "title": "Track: Poster Session 5", "date": "", "ddg_snippet": "Federated learning is a promising distributed machine learning paradigm that can effectively exploit large-scale data without exposing users' privacy . However, ...", "subpage_snippet": "", "source": "icml.cc", "link": "https://icml.cc/virtual/2024/session/35595", "content": "Federated learning is a promising distributed machine learning paradigm that can effectively exploit large-scale data without exposing users' privacy . However, ..."}
+{"idx": 7, "title": "Publications - Di Wang", "date": "", "ddg_snippet": "Growing concerns over data privacy and security emphasize the importance of machine unlearning to remove targeted data's influence from trained models.", "subpage_snippet": "", "source": "shao3wangdi.github.io", "link": "https://shao3wangdi.github.io/publications.html", "content": "Growing concerns over data privacy and security emphasize the importance of machine unlearning to remove targeted data's influence from trained models."}
+{"idx": 8, "title": "On the Interplays Between Fairness, Interpretability, and ...", "date": "", "ddg_snippet": "by J Ferry · 2025 · Cited by 12 — 4.1.1 | Group Fairness and Differential Privacy Are. Theoretically Incompatible. It is provably impossible to build ML models strictly ...", "subpage_snippet": "", "source": "onlinelibrary.wiley.com", "link": "https://onlinelibrary.wiley.com/doi/pdf/10.1111/coin.70113", "content": "by J Ferry · 2025 · Cited by 12 — 4.1.1 | Group Fairness and Differential Privacy Are. Theoretically Incompatible. It is provably impossible to build ML models strictly ..."}
+{"idx": 9, "title": "Publications", "date": "", "ddg_snippet": "by Z Wu · 2025 · Cited by 1 — As machine learning (ML) has been widely developed in real-world applications, the privacy of ML models draws an increasing concern. In this paper, we study how ...", "subpage_snippet": "", "source": "jerrylife.github.io", "link": "https://jerrylife.github.io/publications/", "content": "by Z Wu · 2025 · Cited by 1 — As machine learning (ML) has been widely developed in real-world applications, the privacy of ML models draws an increasing concern. In this paper, we study how ..."}