arxiv_id
float64
1.5k
2.51k
title
stringlengths
9
178
authors
stringlengths
2
22.8k
categories
stringlengths
4
146
summary
stringlengths
103
1.92k
published
stringdate
2015-02-06 10:44:00
2025-07-10 17:59:58
comments
stringlengths
2
417
journal_ref
stringclasses
321 values
doi
stringclasses
398 values
ss_title
stringlengths
8
159
ss_authors
stringlengths
11
8.38k
ss_year
float64
2.02k
2.03k
ss_venue
stringclasses
281 values
ss_citationCount
float64
0
134k
ss_referenceCount
float64
0
429
ss_fieldsOfStudy
stringclasses
47 values
2,410.01679
VinePPO: Refining Credit Assignment in RL Training of LLMs
['Amirhossein Kazemnejad', 'Milad Aghajohari', 'Eva Portelance', 'Alessandro Sordoni', 'Siva Reddy', 'Aaron Courville', 'Nicolas Le Roux']
['cs.LG', 'cs.CL']
Large language models (LLMs) are increasingly applied to complex reasoning tasks that require executing several complex steps before receiving any reward. Properly assigning credit to these steps is essential for enhancing model performance. Proximal Policy Optimization (PPO), a common reinforcement learning (RL) algor...
2024-10-02T15:49:30Z
Accepted at ICML 2025; 12 pages and 22 pages Appendix
null
null
null
null
null
null
null
null
null
2,410.0168
PHI-S: Distribution Balancing for Label-Free Multi-Teacher Distillation
['Mike Ranzinger', 'Jon Barker', 'Greg Heinrich', 'Pavlo Molchanov', 'Bryan Catanzaro', 'Andrew Tao']
['cs.LG', 'cs.AI', 'cs.CV']
Various visual foundation models have distinct strengths and weaknesses, both of which can be improved through heterogeneous multi-teacher knowledge distillation without labels, termed "agglomerative models." We build upon this body of work by studying the effect of the teachers' activation statistics, particularly the...
2024-10-02T15:50:35Z
null
null
null
PHI-S: Distribution Balancing for Label-Free Multi-Teacher Distillation
['Michael Ranzinger', 'Jon Barker', 'Greg Heinrich', 'Pavlo Molchanov', 'Bryan Catanzaro', 'Andrew Tao']
2,024
arXiv.org
5
51
['Computer Science']
2,410.01691
FactAlign: Long-form Factuality Alignment of Large Language Models
['Chao-Wei Huang', 'Yun-Nung Chen']
['cs.CL', 'cs.AI']
Large language models have demonstrated significant potential as the next-generation information access engines. However, their reliability is hindered by issues of hallucination and generating non-factual content. This is particularly problematic in long-form responses, where assessing and ensuring factual accuracy is...
2024-10-02T16:03:13Z
Accepted to EMNLP 2024 Findings
null
null
null
null
null
null
null
null
null
2,410.01744
Leopard: A Vision Language Model For Text-Rich Multi-Image Tasks
['Mengzhao Jia', 'Wenhao Yu', 'Kaixin Ma', 'Tianqing Fang', 'Zhihan Zhang', 'Siru Ouyang', 'Hongming Zhang', 'Dong Yu', 'Meng Jiang']
['cs.CV', 'cs.CL']
Text-rich images, where text serves as the central visual element guiding the overall understanding, are prevalent in real-world applications, such as presentation slides, scanned documents, and webpage snapshots. Tasks involving multiple text-rich images are especially challenging, as they require not only understandi...
2024-10-02T16:55:01Z
Our code is available at https://github.com/tencent-ailab/Leopard
null
null
Leopard: A Vision Language Model For Text-Rich Multi-Image Tasks
['Mengzhao Jia', 'Wenhao Yu', 'Kaixin Ma', 'Tianqing Fang', 'Zhihan Zhang', 'Siru Ouyang', 'Hongming Zhang', 'Meng Jiang', 'Dong Yu']
2,024
Trans. Mach. Learn. Res.
7
79
['Computer Science']
2,410.01912
A Spark of Vision-Language Intelligence: 2-Dimensional Autoregressive Transformer for Efficient Finegrained Image Generation
['Liang Chen', 'Sinan Tan', 'Zefan Cai', 'Weichu Xie', 'Haozhe Zhao', 'Yichi Zhang', 'Junyang Lin', 'Jinze Bai', 'Tianyu Liu', 'Baobao Chang']
['cs.CV', 'cs.AI', 'cs.CL']
This work tackles the information loss bottleneck of vector-quantization (VQ) autoregressive image generation by introducing a novel model architecture called the 2-Dimensional Autoregression (DnD) Transformer. The DnD-Transformer predicts more codes for an image by introducing a new autoregression direction, \textit{m...
2024-10-02T18:10:05Z
25 pages, 20 figures, code is open at https://github.com/chenllliang/DnD-Transformer
null
null
A Spark of Vision-Language Intelligence: 2-Dimensional Autoregressive Transformer for Efficient Finegrained Image Generation
['Liang Chen', 'Sinan Tan', 'Zefan Cai', 'Weichu Xie', 'Haozhe Zhao', 'Yichi Zhang', 'Junyang Lin', 'Jinze Bai', 'Tianyu Liu', 'Baobao Chang']
2,024
International Conference on Learning Representations
4
47
['Computer Science']
2,410.02073
Depth Pro: Sharp Monocular Metric Depth in Less Than a Second
['Aleksei Bochkovskii', 'Amaël Delaunoy', 'Hugo Germain', 'Marcel Santos', 'Yichao Zhou', 'Stephan R. Richter', 'Vladlen Koltun']
['cs.CV', 'cs.LG']
We present a foundation model for zero-shot metric monocular depth estimation. Our model, Depth Pro, synthesizes high-resolution depth maps with unparalleled sharpness and high-frequency details. The predictions are metric, with absolute scale, without relying on the availability of metadata such as camera intrinsics. ...
2024-10-02T22:42:20Z
Published at ICLR 2025. Code and weights available at https://github.com/apple/ml-depth-pro
null
null
null
null
null
null
null
null
null
2,410.02082
FARM: Functional Group-Aware Representations for Small Molecules
['Thao Nguyen', 'Kuan-Hao Huang', 'Ge Liu', 'Martin D. Burke', 'Ying Diao', 'Heng Ji']
['cs.LG', 'q-bio.QM']
We introduce Functional Group-Aware Representations for Small Molecules (FARM), a novel foundation model designed to bridge the gap between SMILES, natural language, and molecular graphs. The key innovation of FARM lies in its functional group-aware tokenization, which directly incorporates functional group information...
2024-10-02T23:04:58Z
Preprint
null
null
FARM: Functional Group-Aware Representations for Small Molecules
['Thao Nguyen', 'Kuan-Hao Huang', 'Ge Liu', 'Martin Burke', 'Ying Diao', 'Heng Ji']
2,024
arXiv.org
1
42
['Computer Science', 'Biology']
2,410.02089
RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning
['Jonas Gehring', 'Kunhao Zheng', 'Jade Copet', 'Vegard Mella', 'Quentin Carbonneaux', 'Taco Cohen', 'Gabriel Synnaeve']
['cs.CL', 'cs.AI']
Large language models (LLMs) deployed as agents solve user-specified tasks over multiple steps while keeping the required manual engagement to a minimum. Crucially, such LLMs need to ground their generations in any feedback obtained to reliably achieve the desired outcomes. We propose an end-to-end reinforcement learni...
2024-10-02T23:25:17Z
Add repair model ablation, update related work
null
null
RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning
['Jonas Gehring', 'Kunhao Zheng', 'Jade Copet', 'Vegard Mella', 'Taco Cohen', 'Gabriele Synnaeve']
2,024
arXiv.org
36
48
['Computer Science']
2,410.02131
Boosting Masked ECG-Text Auto-Encoders as Discriminative Learners
['Hung Manh Pham', 'Aaqib Saeed', 'Dong Ma']
['cs.LG', 'cs.CL']
The accurate interpretation of Electrocardiogram (ECG) signals is pivotal for diagnosing cardiovascular diseases. Integrating ECG signals with accompanying textual reports further holds immense potential to enhance clinical diagnostics by combining physiological data and qualitative insights. However, this integration ...
2024-10-03T01:24:09Z
Accepted at ICML 2025
null
null
null
null
null
null
null
null
null
2,410.02197
Beyond Bradley-Terry Models: A General Preference Model for Language Model Alignment
['Yifan Zhang', 'Ge Zhang', 'Yue Wu', 'Kangping Xu', 'Quanquan Gu']
['cs.AI', 'cs.CL', 'cs.LG']
Modeling human preferences is crucial for aligning foundation models with human values. Traditional reward modeling methods, such as the Bradley-Terry (BT) reward model, fall short in expressiveness, particularly in addressing intransitive preferences. In this paper, we introduce preference embedding, an approach that ...
2024-10-03T04:22:55Z
Accepted to the 42nd International Conference on Machine Learning (ICML 2025)
null
null
Beyond Bradley-Terry Models: A General Preference Model for Language Model Alignment
['Yifan Zhang', 'Ge Zhang', 'Yue Wu', 'Kangping Xu', 'Quanquan Gu']
2,024
null
2
62
['Computer Science']
2,410.02249
Spiking Neural Network as Adaptive Event Stream Slicer
['Jiahang Cao', 'Mingyuan Sun', 'Ziqing Wang', 'Hao Cheng', 'Qiang Zhang', 'Shibo Zhou', 'Renjing Xu']
['cs.CV', 'cs.NE']
Event-based cameras are attracting significant interest as they provide rich edge information, high dynamic range, and high temporal resolution. Many state-of-the-art event-based algorithms rely on splitting the events into fixed groups, resulting in the omission of crucial temporal information, particularly when deali...
2024-10-03T06:41:10Z
Accepted to NeurIPS 2024
null
null
Spiking Neural Network as Adaptive Event Stream Slicer
['Jiahang Cao', 'Mingyuan Sun', 'Ziqing Wang', 'Haotai Cheng', 'Qiang Zhang', 'Shibo Zhou', 'Renjing Xu']
2,024
Neural Information Processing Systems
2
50
['Computer Science']
2,410.0225
Probabilistic road classification in historical maps using synthetic data and deep learning
['Dominik J. Mühlematter', 'Sebastian Schweizer', 'Chenjing Jiao', 'Xue Xia', 'Magnus Heitzler', 'Lorenz Hurni']
['cs.CV', 'cs.LG']
Historical maps are invaluable for analyzing long-term changes in transportation and spatial development, offering a rich source of data for evolutionary studies. However, digitizing and classifying road networks from these maps is often expensive and time-consuming, limiting their widespread use. Recent advancements i...
2024-10-03T06:43:09Z
null
null
null
Probabilistic road classification in historical maps using synthetic data and deep learning
['Dominik J. Mühlematter', 'Sebastian Schweizer', 'C. Jiao', 'Xue Xia', 'M. Heitzler', 'L. Hurni']
2,024
arXiv.org
0
71
['Computer Science']
2,410.02367
SageAttention: Accurate 8-Bit Attention for Plug-and-play Inference Acceleration
['Jintao Zhang', 'Jia Wei', 'Haofeng Huang', 'Pengle Zhang', 'Jun Zhu', 'Jianfei Chen']
['cs.LG']
The transformer architecture predominates across various models. As the heart of the transformer, attention has a computational complexity of $O(N^2)$, compared to $O(N)$ for linear transformations. When handling large sequence lengths, attention becomes the primary time-consuming component. Although quantization has p...
2024-10-03T10:25:23Z
@inproceedings{zhang2025sageattention, title={SageAttention: Accurate 8-Bit Attention for Plug-and-play Inference Acceleration}, author={Zhang, Jintao and Wei, Jia and Zhang, Pengle and Zhu, Jun and Chen, Jianfei}, booktitle={International Conference on Learning Representations (ICLR)}, year={2025} }
The Thirteenth International Conference on Learning Representations (ICLR 2025)
null
SageAttention: Accurate 8-Bit Attention for Plug-and-play Inference Acceleration
['Jintao Zhang', 'Jia Wei', 'Pengle Zhang', 'Jun Zhu', 'Jianfei Chen']
2,024
International Conference on Learning Representations
39
79
['Computer Science']
2,410.02381
MetaMetrics: Calibrating Metrics For Generation Tasks Using Human Preferences
['Genta Indra Winata', 'David Anugraha', 'Lucky Susanto', 'Garry Kuwanto', 'Derry Tanti Wijaya']
['cs.CL', 'cs.AI', 'cs.CV', 'cs.LG']
Understanding the quality of a performance evaluation metric is crucial for ensuring that model outputs align with human preferences. However, it remains unclear how well each metric captures the diverse aspects of these preferences, as metrics often excel in one particular area but not across all dimensions. To addres...
2024-10-03T11:01:25Z
Accepted to ICLR 2025
null
null
MetaMetrics: Calibrating Metrics For Generation Tasks Using Human Preferences
['Genta Indra Winata', 'David Anugraha', 'Lucky Susanto', 'Garry Kuwanto', 'Derry Tanti Wijaya']
2,024
International Conference on Learning Representations
11
85
['Computer Science']
2,410.02416
Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models
['Seyedmorteza Sadat', 'Otmar Hilliges', 'Romann M. Weber']
['cs.LG', 'cs.CV']
Classifier-free guidance (CFG) is crucial for improving both generation quality and alignment between the input condition and final output in diffusion models. While a high guidance scale is generally required to enhance these aspects, it also causes oversaturation and unrealistic artifacts. In this paper, we revisit t...
2024-10-03T12:06:29Z
Published as a conference paper at ICLR 2025
The Thirteenth International Conference on Learning Representations (ICLR 2025)
null
null
null
null
null
null
null
null
2,410.0244
Optimizing Adaptive Attacks against Watermarks for Language Models
['Abdulrahman Diaa', 'Toluwani Aremu', 'Nils Lukas']
['cs.CR', 'cs.AI']
Large Language Models (LLMs) can be misused to spread unwanted content at scale. Content watermarking deters misuse by hiding messages in content, enabling its detection using a secret watermarking key. Robustness is a core security property, stating that evading detection requires (significant) degradation of the cont...
2024-10-03T12:37:39Z
To appear at the International Conference on Machine Learning (ICML'25)
null
null
null
null
null
null
null
null
null
2,410.02503
Mixed-Session Conversation with Egocentric Memory
['Jihyoung Jang', 'Taeyoung Kim', 'Hyounghun Kim']
['cs.CL', 'cs.AI']
Recently introduced dialogue systems have demonstrated high usability. However, they still fall short of reflecting real-world conversation scenarios. Current dialogue systems exhibit an inability to replicate the dynamic, continuous, long-term interactions involving multiple partners. This shortfall arises because the...
2024-10-03T14:06:43Z
EMNLP Findings 2024 (30 pages); Project website: https://mixed-session.github.io/
null
null
null
null
null
null
null
null
null
2,410.02525
Contextual Document Embeddings
['John X. Morris', 'Alexander M. Rush']
['cs.CL', 'cs.AI']
Dense document embeddings are central to neural retrieval. The dominant paradigm is to train and construct embeddings by running encoders directly on individual documents. In this work, we argue that these embeddings, while effective, are implicitly out-of-context for targeted use cases of retrieval, and that a context...
2024-10-03T14:33:34Z
null
null
null
Contextual Document Embeddings
['John X. Morris', 'Alexander M. Rush']
2,024
International Conference on Learning Representations
9
58
['Computer Science']
2,410.02653
Measuring and Improving Persuasiveness of Large Language Models
['Somesh Singh', 'Yaman K Singla', 'Harini SI', 'Balaji Krishnamurthy']
['cs.CL', 'cs.CV']
LLMs are increasingly being used in workflows involving generating content to be consumed by humans (e.g., marketing) and also in directly interacting with humans (e.g., through chatbots). The development of such systems that are capable of generating verifiably persuasive messages presents both opportunities and chall...
2024-10-03T16:36:35Z
null
null
null
null
null
null
null
null
null
null
2,410.0266
How to Train Long-Context Language Models (Effectively)
['Tianyu Gao', 'Alexander Wettig', 'Howard Yen', 'Danqi Chen']
['cs.CL', 'cs.LG']
We study continued training and supervised fine-tuning (SFT) of a language model (LM) to make effective use of long-context information. We first establish a reliable evaluation protocol to guide model development -- instead of perplexity or simple needle-in-a-haystack (NIAH) tests, we use a broad set of long-context d...
2024-10-03T16:46:52Z
Accepted to ACL 2025. Our code, data, and models are available at https://github.com/princeton-nlp/ProLong
null
null
How to Train Long-Context Language Models (Effectively)
['Tianyu Gao', 'Alexander Wettig', 'Howard Yen', 'Danqi Chen']
2,024
arXiv.org
48
104
['Computer Science']
2,410.02675
FAN: Fourier Analysis Networks
['Yihong Dong', 'Ge Li', 'Yongding Tao', 'Xue Jiang', 'Kechi Zhang', 'Jia Li', 'Jinliang Deng', 'Jing Su', 'Jun Zhang', 'Jingjing Xu']
['cs.LG', 'cs.AI', 'cs.CL']
Despite the remarkable successes of general-purpose neural networks, such as MLPs and Transformers, we find that they exhibit notable shortcomings in modeling and reasoning about periodic phenomena, achieving only marginal performance within the training domain and failing to generalize effectively to out-of-domain (OO...
2024-10-03T17:02:21Z
null
null
null
null
null
null
null
null
null
null
2,410.02678
Distilling an End-to-End Voice Assistant Without Instruction Training Data
['William Held', 'Ella Li', 'Michael Ryan', 'Weiyan Shi', 'Yanzhe Zhang', 'Diyi Yang']
['cs.CL', 'cs.AI']
Voice assistants, such as Siri and Google Assistant, typically model audio and text separately, resulting in lost speech information and increased complexity. Recent efforts to address this with end-to-end Speech Large Language Models (LLMs) trained with supervised finetuning (SFT) have led to models ``forgetting" ca...
2024-10-03T17:04:48Z
null
null
null
null
null
null
null
null
null
null
2,410.02705
ControlAR: Controllable Image Generation with Autoregressive Models
['Zongming Li', 'Tianheng Cheng', 'Shoufa Chen', 'Peize Sun', 'Haocheng Shen', 'Longjin Ran', 'Xiaoxin Chen', 'Wenyu Liu', 'Xinggang Wang']
['cs.CV']
Autoregressive (AR) models have reformulated image generation as next-token prediction, demonstrating remarkable potential and emerging as strong competitors to diffusion models. However, control-to-image generation, akin to ControlNet, remains largely unexplored within AR models. Although a natural approach, inspired ...
2024-10-03T17:28:07Z
To appear in ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.02712
LLaVA-Critic: Learning to Evaluate Multimodal Models
['Tianyi Xiong', 'Xiyao Wang', 'Dong Guo', 'Qinghao Ye', 'Haoqi Fan', 'Quanquan Gu', 'Heng Huang', 'Chunyuan Li']
['cs.CV', 'cs.CL']
We introduce LLaVA-Critic, the first open-source large multimodal model (LMM) designed as a generalist evaluator to assess performance across a wide range of multimodal tasks. LLaVA-Critic is trained using a high-quality critic instruction-following dataset that incorporates diverse evaluation criteria and scenarios. O...
2024-10-03T17:36:33Z
Accepted by CVPR 2025; Project Page: https://llava-vl.github.io/blog/2024-10-03-llava-critic
null
null
null
null
null
null
null
null
null
2,410.02713
Video Instruction Tuning With Synthetic Data
['Yuanhan Zhang', 'Jinming Wu', 'Wei Li', 'Bo Li', 'Zejun Ma', 'Ziwei Liu', 'Chunyuan Li']
['cs.CV', 'cs.CL']
The development of video large multimodal models (LMMs) has been hindered by the difficulty of curating large amounts of high-quality raw data from the web. To address this, we propose an alternative approach by creating a high-quality synthetic dataset specifically for video instruction-following, namely LLaVA-Video-1...
2024-10-03T17:36:49Z
Project page: https://llava-vl.github.io/blog/2024-09-30-llava-video/
null
null
null
null
null
null
null
null
null
2,410.02743
MA-RLHF: Reinforcement Learning from Human Feedback with Macro Actions
['Yekun Chai', 'Haoran Sun', 'Huang Fang', 'Shuohuan Wang', 'Yu Sun', 'Hua Wu']
['cs.CL']
Reinforcement learning from human feedback (RLHF) has demonstrated effectiveness in aligning large language models (LLMs) with human preferences. However, token-level RLHF suffers from the credit assignment problem over long sequences, where delayed rewards make it challenging for the model to discern which actions con...
2024-10-03T17:55:13Z
null
null
null
null
null
null
null
null
null
null
2,410.02745
AVG-LLaVA: A Large Multimodal Model with Adaptive Visual Granularity
['Zhibin Lan', 'Liqiang Niu', 'Fandong Meng', 'Wenbo Li', 'Jie Zhou', 'Jinsong Su']
['cs.CV', 'cs.AI', 'cs.CL']
Recently, when dealing with high-resolution images, dominant LMMs usually divide them into multiple local images and one global image, which will lead to a large number of visual tokens. In this work, we introduce AVG-LLaVA, an LMM that can adaptively select the appropriate visual granularity based on the input image a...
2024-09-20T10:50:21Z
Preprint
null
null
null
null
null
null
null
null
null
2,410.02749
Training Language Models on Synthetic Edit Sequences Improves Code Synthesis
['Ulyana Piterbarg', 'Lerrel Pinto', 'Rob Fergus']
['cs.LG', 'cs.CL']
Software engineers mainly write code by editing existing programs. In contrast, language models (LMs) autoregressively synthesize programs in a single pass. One explanation for this is the scarcity of sequential edit data. While high-quality instruction data for code synthesis is scarce, edit data for synthesis is even...
2024-10-03T17:57:22Z
ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.0276
Erasing Conceptual Knowledge from Language Models
['Rohit Gandikota', 'Sheridan Feucht', 'Samuel Marks', 'David Bau']
['cs.CL', 'cs.LG']
In this work, we propose Erasure of Language Memory (ELM), an approach for concept-level unlearning built on the principle of matching the distribution defined by an introspective classifier. Our key insight is that effective unlearning should leverage the model's ability to evaluate its own knowledge, using the model ...
2024-10-03T17:59:30Z
Project Page: https://elm.baulab.info
null
null
null
null
null
null
null
null
null
2,410.02761
FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models
['Zhipei Xu', 'Xuanyu Zhang', 'Runyi Li', 'Zecheng Tang', 'Qing Huang', 'Jian Zhang']
['cs.CV', 'cs.AI']
The rapid development of generative AI is a double-edged sword, which not only facilitates content creation but also makes image manipulation easier and more difficult to detect. Although current image forgery detection and localization (IFDL) methods are generally effective, they tend to face two challenges: \textbf{1...
2024-10-03T17:59:34Z
Accepted by ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.02884
LLaMA-Berry: Pairwise Optimization for O1-like Olympiad-Level Mathematical Reasoning
['Di Zhang', 'Jianbo Wu', 'Jingdi Lei', 'Tong Che', 'Jiatong Li', 'Tong Xie', 'Xiaoshui Huang', 'Shufei Zhang', 'Marco Pavone', 'Yuqiang Li', 'Wanli Ouyang', 'Dongzhan Zhou']
['cs.AI', 'cs.CL']
This paper presents an advanced mathematical problem-solving framework, LLaMA-Berry, for enhancing the mathematical reasoning ability of Large Language Models (LLMs). The framework combines Monte Carlo Tree Search (MCTS) with iterative Self-Refine to optimize the reasoning path and utilizes a pairwise reward model to e...
2024-10-03T18:12:29Z
null
null
null
LLaMA-Berry: Pairwise Optimization for O1-like Olympiad-Level Mathematical Reasoning
['Di Zhang', 'Jianbo Wu', 'Jingdi Lei', 'Tong Che', 'Jiatong Li', 'Tong Xie', 'Xiaoshui Huang', 'Shufei Zhang', 'Marco Pavone', 'Yuqiang Li', 'Wanli Ouyang', 'Dongzhan Zhou']
2,024
arXiv.org
61
67
['Computer Science']
2,410.02907
NNetNav: Unsupervised Learning of Browser Agents Through Environment Interaction in the Wild
['Shikhar Murty', 'Hao Zhu', 'Dzmitry Bahdanau', 'Christopher D. Manning']
['cs.CL']
We introduce NNetNav, a method for unsupervised interaction with websites that generates synthetic demonstrations for training browser agents. Given any website, NNetNav produces these demonstrations by retroactively labeling action sequences from an exploration policy. Most work on training browser agents has relied o...
2024-10-03T18:56:51Z
Code, Data and Models available at https://www.nnetnav.dev
null
null
null
null
null
null
null
null
null
2,410.03051
AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark
['Wenhao Chai', 'Enxin Song', 'Yilun Du', 'Chenlin Meng', 'Vashisht Madhavan', 'Omer Bar-Tal', 'Jenq-Neng Hwang', 'Saining Xie', 'Christopher D. Manning']
['cs.CV']
Video detailed captioning is a key task which aims to generate comprehensive and coherent textual descriptions of video content, benefiting both video understanding and generation. In this paper, we propose AuroraCap, a video captioner based on a large multimodal model. We follow the simplest architecture design withou...
2024-10-04T00:13:54Z
Accepted to ICLR 2025. Code, docs, weight, benchmark and training data are all avaliable at https://rese1f.github.io/aurora-web/
null
null
null
null
null
null
null
null
null
2,410.03075
Multilingual Topic Classification in X: Dataset and Analysis
['Dimosthenis Antypas', 'Asahi Ushio', 'Francesco Barbieri', 'Jose Camacho-Collados']
['cs.CL']
In the dynamic realm of social media, diverse topics are discussed daily, transcending linguistic boundaries. However, the complexities of understanding and categorising this content across various languages remain an important challenge with traditional techniques like topic modelling often struggling to accommodate t...
2024-10-04T01:37:26Z
Accepted at EMNLP 2024
null
null
null
null
null
null
null
null
null
2,410.03115
X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale
['Haoran Xu', 'Kenton Murray', 'Philipp Koehn', 'Hieu Hoang', 'Akiko Eriguchi', 'Huda Khayrallah']
['cs.CL']
Large language models (LLMs) have achieved remarkable success across various NLP tasks with a focus on English due to English-centric pre-training and limited multilingual data. In this work, we focus on the problem of translation, and while some multilingual LLMs claim to support for hundreds of languages, models ofte...
2024-10-04T03:17:27Z
Published as a conference paper at ICLR 2025 (spotlight)
null
null
X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale
['Haoran Xu', 'Kenton Murray', 'Philipp Koehn', 'Hieu D. Hoang', 'Akiko Eriguchi', 'Huda Khayrallah']
2,024
International Conference on Learning Representations
15
71
['Computer Science']
2,410.0316
Redefining Temporal Modeling in Video Diffusion: The Vectorized Timestep Approach
['Yaofang Liu', 'Yumeng Ren', 'Xiaodong Cun', 'Aitor Artola', 'Yang Liu', 'Tieyong Zeng', 'Raymond H. Chan', 'Jean-michel Morel']
['cs.CV', 'cs.LG']
Diffusion models have revolutionized image generation, and their extension to video generation has shown promise. However, current video diffusion models~(VDMs) rely on a scalar timestep variable applied at the clip level, which limits their ability to model complex temporal dependencies needed for various tasks like i...
2024-10-04T05:47:39Z
Code at https://github.com/Yaofang-Liu/FVDM
null
null
null
null
null
null
null
null
null
2,410.0324
Beyond Film Subtitles: Is YouTube the Best Approximation of Spoken Vocabulary?
['Adam Nohejl', 'Frederikus Hudi', 'Eunike Andriani Kardinata', 'Shintaro Ozaki', 'Maria Angelica Riera Machin', 'Hongyu Sun', 'Justin Vasselli', 'Taro Watanabe']
['cs.CL']
Word frequency is a key variable in psycholinguistics, useful for modeling human familiarity with words even in the era of large language models (LLMs). Frequency in film subtitles has proved to be a particularly good approximation of everyday language exposure. For many languages, however, film subtitles are not easil...
2024-10-04T09:04:20Z
Accepted to COLING 2025. 9 pages, 3 figures
null
null
null
null
null
null
null
null
null
2,410.0329
Grounded-VideoLLM: Sharpening Fine-grained Temporal Grounding in Video Large Language Models
['Haibo Wang', 'Zhiyang Xu', 'Yu Cheng', 'Shizhe Diao', 'Yufan Zhou', 'Yixin Cao', 'Qifan Wang', 'Weifeng Ge', 'Lifu Huang']
['cs.CV', 'cs.AI']
Video Large Language Models (Video-LLMs) have demonstrated remarkable capabilities in coarse-grained video understanding, however, they struggle with fine-grained temporal grounding. In this paper, we introduce Grounded-VideoLLM, a novel Video-LLM adept at perceiving and reasoning over specific video moments in a fine-...
2024-10-04T10:04:37Z
null
null
null
Grounded-VideoLLM: Sharpening Fine-grained Temporal Grounding in Video Large Language Models
['Haibo Wang', 'Zhiyang Xu', 'Yu Cheng', 'Shizhe Diao', 'Yufan Zhou', 'Yixin Cao', 'Qifan Wang', 'Weifeng Ge', 'Lifu Huang']
2,024
arXiv.org
26
58
['Computer Science']
2,410.03355
LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding
['Doohyuk Jang', 'Sihwan Park', 'June Yong Yang', 'Yeonsung Jung', 'Jihun Yun', 'Souvik Kundu', 'Sung-Yub Kim', 'Eunho Yang']
['cs.CV', 'cs.AI']
Auto-Regressive (AR) models have recently gained prominence in image generation, often matching or even surpassing the performance of diffusion models. However, one major limitation of AR models is their sequential nature, which processes tokens one at a time, slowing down generation compared to models like GANs or dif...
2024-10-04T12:21:03Z
30 pages, 13 figures, Accepted to ICLR 2025 (poster)
null
null
null
null
null
null
null
null
null
2,410.03524
Steering Large Language Models between Code Execution and Textual Reasoning
['Yongchao Chen', 'Harsh Jhamtani', 'Srinagesh Sharma', 'Chuchu Fan', 'Chi Wang']
['cs.CL']
While a lot of recent research focuses on enhancing the textual reasoning capabilities of Large Language Models (LLMs) by optimizing the multi-agent framework or reasoning chains, several benchmark tasks can be solved with 100\% success through direct coding, which is more scalable and avoids the computational overhead...
2024-10-04T15:44:47Z
32 pages, 12 figures, 12 tables
The Thirteenth International Conference on Learning Representations (ICLR'2025)
null
null
null
null
null
null
null
null
2,410.03553
Structure-Enhanced Protein Instruction Tuning: Towards General-Purpose Protein Understanding with LLMs
['Wei Wu', 'Chao Wang', 'Liyi Chen', 'Mingze Yin', 'Yiheng Zhu', 'Kun Fu', 'Jieping Ye', 'Hui Xiong', 'Zheng Wang']
['cs.CL', 'q-bio.BM']
Proteins, as essential biomolecules, play a central role in biological processes, including metabolic reactions and DNA replication. Accurate prediction of their properties and functions is crucial in biological applications. Recent development of protein language models (pLMs) with supervised fine tuning provides a pr...
2024-10-04T16:02:50Z
Accepted by KDD2025
null
10.1145/3711896.3737138
Structure-Enhanced Protein Instruction Tuning: Towards General-Purpose Protein Understanding
['Wei Wu', 'Chao Wang', 'Liyi Chen', 'Mingze Yin', 'Yiheng Zhu', 'Kun Fu', 'Jieping Ye', 'Hui Xiong', 'Zheng Wang']
2,024
arXiv.org
1
116
['Computer Science', 'Biology']
2,410.03617
What Matters for Model Merging at Scale?
['Prateek Yadav', 'Tu Vu', 'Jonathan Lai', 'Alexandra Chronopoulou', 'Manaal Faruqui', 'Mohit Bansal', 'Tsendsuren Munkhdalai']
['cs.LG', 'cs.AI', 'cs.CL']
Model merging aims to combine multiple expert models into a more capable single model, offering benefits such as reduced storage and serving costs, improved generalization, and support for decentralized model development. Despite its promise, previous studies have primarily focused on merging a few small models. This l...
2024-10-04T17:17:19Z
20 Pages, 7 Figures, 4 Tables
null
null
null
null
null
null
null
null
null
2,410.0373
Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs
['Mehdi Ali', 'Michael Fromm', 'Klaudia Thellmann', 'Jan Ebert', 'Alexander Arno Weber', 'Richard Rutmann', 'Charvi Jain', 'Max Lübbering', 'Daniel Steinigen', 'Johannes Leveling', 'Katrin Klug', 'Jasper Schulze Buschhoff', 'Lena Jurkschat', 'Hammam Abdelwahab', 'Benny Jörg Stein', 'Karl-Heinz Sylla', 'Pavel Denisov', ...
['cs.CL', 'cs.AI', 'cs.LG']
We present two multilingual LLMs designed to embrace Europe's linguistic diversity by supporting all 24 official languages of the European Union. Trained on a dataset comprising around 60% non-English data and utilizing a custom multilingual tokenizer, our models address the limitations of existing LLMs that predominan...
2024-09-30T16:05:38Z
null
null
null
null
null
null
null
null
null
null
2,410.03742
Beyond Scalar Reward Model: Learning Generative Judge from Preference Data
['Ziyi Ye', 'Xiangsheng Li', 'Qiuchi Li', 'Qingyao Ai', 'Yujia Zhou', 'Wei Shen', 'Dong Yan', 'Yiqun Liu']
['cs.CL', 'cs.AI', 'cs.LG']
Learning from preference feedback is a common practice for aligning large language models~(LLMs) with human value. Conventionally, preference data is learned and encoded into a scalar reward model that connects a value head with an LLM to produce a scalar score as preference or reward. However, scalar models lack inter...
2024-10-01T07:38:58Z
null
null
null
null
null
null
null
null
null
null
2,410.0375
SQFT: Low-cost Model Adaptation in Low-precision Sparse Foundation Models
['Juan Pablo Muñoz', 'Jinjie Yuan', 'Nilesh Jain']
['cs.LG', 'cs.AI', 'cs.CL']
Large pre-trained models (LPMs), such as large language models, have become ubiquitous and are employed in many applications. These models are often adapted to a desired domain or downstream task through a fine-tuning stage. This paper proposes SQFT, an end-to-end solution for low-precision sparse parameter-efficient f...
2024-10-01T19:49:35Z
To be published in EMNLP-24 Findings
null
null
SQFT: Low-cost Model Adaptation in Low-precision Sparse Foundation Models
['J. P. Munoz', 'Jinjie Yuan', 'Nilesh Jain']
2,024
Conference on Empirical Methods in Natural Language Processing
3
28
['Computer Science']
2,410.03804
Mixture of Attentions For Speculative Decoding
['Matthieu Zimmer', 'Milan Gritta', 'Gerasimos Lampouras', 'Haitham Bou Ammar', 'Jun Wang']
['cs.CL', 'cs.AI', 'cs.LG']
The growth in the number of parameters of Large Language Models (LLMs) has led to a significant surge in computational requirements, making them challenging and costly to deploy. Speculative decoding (SD) leverages smaller models to efficiently propose future tokens, which are then verified by the LLM in parallel. Smal...
2024-10-04T10:25:52Z
Accepted at International Conference on Learning Representations (ICLR 2025)
null
null
Mixture of Attentions For Speculative Decoding
['Matthieu Zimmer', 'Milan Gritta', 'Gerasimos Lampouras', 'Haitham Bou-Ammar', 'Jun Wang']
2,024
International Conference on Learning Representations
6
41
['Computer Science']
2,410.03825
MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
['Junyi Zhang', 'Charles Herrmann', 'Junhwa Hur', 'Varun Jampani', 'Trevor Darrell', 'Forrester Cole', 'Deqing Sun', 'Ming-Hsuan Yang']
['cs.CV']
Estimating geometry from dynamic scenes, where objects move and deform over time, remains a core challenge in computer vision. Current approaches often rely on multi-stage pipelines or global optimizations that decompose the problem into subtasks, like depth and flow, leading to complex systems prone to errors. In this...
2024-10-04T18:00:07Z
Accepted by ICLR 25, Project page: https://monst3r-project.github.io/
null
null
MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
['Junyi Zhang', 'Charles Herrmann', 'Junhwa Hur', 'Varun Jampani', 'Trevor Darrell', 'Forrester Cole', 'Deqing Sun', 'Ming-Hsuan Yang']
2,024
International Conference on Learning Representations
96
77
['Computer Science']
2,410.0393
Reverb: Open-Source ASR and Diarization from Rev
['Nishchal Bhandari', 'Danny Chen', 'Miguel Ángel del Río Fernández', 'Natalie Delworth', 'Jennifer Drexler Fox', 'Migüel Jetté', 'Quinten McNamara', 'Corey Miller', 'Ondřej Novotný', 'Ján Profant', 'Nan Qin', 'Martin Ratajczak', 'Jean-Philippe Robichaud']
['cs.CL', 'cs.SD', 'eess.AS']
Today, we are open-sourcing our core speech recognition and diarization models for non-commercial use. We are releasing both a full production pipeline for developers as well as pared-down research models for experimentation. Rev hopes that these releases will spur research and innovation in the fast-moving domain of v...
2024-10-04T21:13:58Z
null
null
null
null
null
null
null
null
null
null
2,410.0396
SwiftKV: Fast Prefill-Optimized Inference with Knowledge-Preserving Model Transformation
['Aurick Qiao', 'Zhewei Yao', 'Samyam Rajbhandari', 'Yuxiong He']
['cs.LG', 'cs.AI', 'cs.CL']
LLM inference for enterprise applications, such as summarization, RAG, and code-generation, typically observe much longer prompt than generations, leading to high prefill cost and response latency. We present SwiftKV, a novel model transformation and distillation procedure targeted at reducing the prefill compute (in F...
2024-10-04T22:45:26Z
null
null
null
null
null
null
null
null
null
null
2,410.04133
An Electrocardiogram Foundation Model Built on over 10 Million Recordings with External Evaluation across Multiple Domains
['Jun Li', 'Aaron Aguirre', 'Junior Moura', 'Che Liu', 'Lanhai Zhong', 'Chenxi Sun', 'Gari Clifford', 'Brandon Westover', 'Shenda Hong']
['cs.LG', 'cs.AI', 'eess.SP']
Artificial intelligence (AI) has demonstrated significant potential in ECG analysis and cardiovascular disease assessment. Recently, foundation models have played a remarkable role in advancing medical AI. The development of an ECG foundation model holds the promise of elevating AI-ECG research to new heights. However,...
2024-10-05T12:12:02Z
Code: https://github.com/PKUDigitalHealth/ECGFounder
null
null
null
null
null
null
null
null
null
2,410.04223
Multimodal Large Language Models for Inverse Molecular Design with Retrosynthetic Planning
['Gang Liu', 'Michael Sun', 'Wojciech Matusik', 'Meng Jiang', 'Jie Chen']
['cs.LG', 'physics.chem-ph', 'q-bio.BM']
While large language models (LLMs) have integrated images, adapting them to graphs remains challenging, limiting their applications in materials and drug design. This difficulty stems from the need for coherent autoregressive generation across texts and graphs. To address this, we introduce Llamole, the first multimoda...
2024-10-05T16:35:32Z
27 pages, 11 figures, 4 tables
null
null
Multimodal Large Language Models for Inverse Molecular Design with Retrosynthetic Planning
['Gang Liu', 'Michael Sun', 'Wojciech Matusik', 'Meng Jiang', 'Jie Chen']
2,024
International Conference on Learning Representations
9
47
['Computer Science', 'Physics', 'Biology']
2,410.04269
RoQLlama: A Lightweight Romanian Adapted Language Model
['George-Andrei Dima', 'Andrei-Marius Avram', 'Cristian-George Crăciun', 'Dumitru-Clementin Cercel']
['cs.CL']
The remarkable achievements obtained by open-source large language models (LLMs) in recent years have predominantly been concentrated on tasks involving the English language. In this paper, we aim to advance the performance of Llama2 models on Romanian tasks. We tackle the problem of reduced computing resources by usin...
2024-10-05T19:14:11Z
Accepted at EMNLP Findings 2024 (short papers)
null
null
null
null
null
null
null
null
null
2,410.04415
Geometric Analysis of Reasoning Trajectories: A Phase Space Approach to Understanding Valid and Invalid Multi-Hop Reasoning in LLMs
['Javier Marin']
['cs.AI', 'cs.LG']
This paper proposes a novel approach to analyzing multi-hop reasoning in language models through Hamiltonian mechanics. We map reasoning chains in embedding spaces to Hamiltonian systems, defining a function that balances reasoning progression (kinetic energy) against question relevance (potential energy). Analyzing re...
2024-10-06T09:09:14Z
null
null
null
null
null
null
null
null
null
null
2,410.04456
SWEb: A Large Web Dataset for the Scandinavian Languages
['Tobias Norlund', 'Tim Isbister', 'Amaru Cuba Gyllensten', 'Paul Dos Santos', 'Danila Petrelli', 'Ariel Ekgren', 'Magnus Sahlgren']
['cs.CL']
This paper presents the hitherto largest pretraining dataset for the Scandinavian languages: the Scandinavian WEb (SWEb), comprising over one trillion tokens. The paper details the collection and processing pipeline, and introduces a novel model-based text extractor that significantly reduces complexity in comparison w...
2024-10-06T11:55:15Z
null
null
null
SWEb: A Large Web Dataset for the Scandinavian Languages
['Tobias Norlund', 'T. Isbister', 'Amaru Cuba Gyllensten', 'Paul Gabriel dos Santos', 'Danila Petrelli', 'Ariel Ekgren', 'Magnus Sahlgren']
2,024
arXiv.org
0
24
['Computer Science']
2,410.04587
Hammer: Robust Function-Calling for On-Device Language Models via Function Masking
['Qiqiang Lin', 'Muning Wen', 'Qiuying Peng', 'Guanyu Nie', 'Junwei Liao', 'Jun Wang', 'Xiaoyun Mo', 'Jiamu Zhou', 'Cheng Cheng', 'Yin Zhao', 'Jun Wang', 'Weinan Zhang']
['cs.LG', 'cs.AI', 'cs.SE']
Large language models have demonstrated impressive value in performing as autonomous agents when equipped with external tools and API calls. Nonetheless, effectively harnessing their potential for executing complex tasks crucially relies on enhancements in their function calling capabilities. This paper identifies a cr...
2024-10-06T18:57:46Z
null
null
null
Hammer: Robust Function-Calling for On-Device Language Models via Function Masking
['Qiqiang Lin', 'Muning Wen', 'Qiuying Peng', 'Guanyu Nie', 'Junwei Liao', 'Jun Wang', 'Xiaoyun Mo', 'Jiamu Zhou', 'Cheng Cheng', 'Yin Zhao', 'Weinan Zhang']
2,024
arXiv.org
21
27
['Computer Science']
2,410.04612
Regressing the Relative Future: Efficient Policy Optimization for Multi-turn RLHF
['Zhaolin Gao', 'Wenhao Zhan', 'Jonathan D. Chang', 'Gokul Swamy', 'Kianté Brantley', 'Jason D. Lee', 'Wen Sun']
['cs.LG', 'cs.AI', 'cs.CL']
Large Language Models (LLMs) have achieved remarkable success at tasks like summarization that involve a single turn of interaction. However, they can still struggle with multi-turn tasks like dialogue that require long-term planning. Previous works on multi-turn dialogue extend single-turn reinforcement learning from ...
2024-10-06T20:20:22Z
null
null
null
null
null
null
null
null
null
null
2,410.04803
Timer-XL: Long-Context Transformers for Unified Time Series Forecasting
['Yong Liu', 'Guo Qin', 'Xiangdong Huang', 'Jianmin Wang', 'Mingsheng Long']
['cs.LG', 'stat.ML']
We present Timer-XL, a causal Transformer for unified time series forecasting. To uniformly predict multidimensional time series, we generalize next token prediction, predominantly adopted for 1D token sequences, to multivariate next token prediction. The paradigm formulates various forecasting tasks as a long-context ...
2024-10-07T07:27:39Z
null
null
null
null
null
null
null
null
null
null
2,410.04932
OmniBooth: Learning Latent Control for Image Synthesis with Multi-modal Instruction
['Leheng Li', 'Weichao Qiu', 'Xu Yan', 'Jing He', 'Kaiqiang Zhou', 'Yingjie Cai', 'Qing Lian', 'Bingbing Liu', 'Ying-Cong Chen']
['cs.CV']
We present OmniBooth, an image generation framework that enables spatial control with instance-level multi-modal customization. For all instances, the multimodal instruction can be described through text prompts or image references. Given a set of user-defined masks and associated text or image guidance, our objective ...
2024-10-07T11:26:13Z
null
null
null
OmniBooth: Learning Latent Control for Image Synthesis with Multi-modal Instruction
['Leheng Li', 'Weichao Qiu', 'Xu Yan', 'Jing He', 'Kaiqiang Zhou', 'Yingjie Cai', 'Qing Lian', 'Bingbing Liu', 'Ying-Cong Chen']
2,024
arXiv.org
1
42
['Computer Science']
2,410.05077
ZEBRA: Zero-Shot Example-Based Retrieval Augmentation for Commonsense Question Answering
['Francesco Maria Molfese', 'Simone Conia', 'Riccardo Orlando', 'Roberto Navigli']
['cs.CL']
Current Large Language Models (LLMs) have shown strong reasoning capabilities in commonsense question answering benchmarks, but the process underlying their success remains largely opaque. As a consequence, recent approaches have equipped LLMs with mechanisms for knowledge retrieval, reasoning and introspection, not on...
2024-10-07T14:31:43Z
Accepted at EMNLP 2024 Main Conference
null
null
null
null
null
null
null
null
null
2,410.0516
VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks
['Ziyan Jiang', 'Rui Meng', 'Xinyi Yang', 'Semih Yavuz', 'Yingbo Zhou', 'Wenhu Chen']
['cs.CV', 'cs.AI', 'cs.CL']
Embedding models have been crucial in enabling various downstream tasks such as semantic similarity, information retrieval, and clustering. Recently, there has been a surge of interest in developing universal text embedding models that can generalize across tasks (e.g., MTEB). However, progress in learning universal mu...
2024-10-07T16:14:05Z
Technical Report
null
null
VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks
['Ziyan Jiang', 'Rui Meng', 'Xinyi Yang', 'Semih Yavuz', 'Yingbo Zhou', 'Wenhu Chen']
2,024
International Conference on Learning Representations
29
81
['Computer Science']
2,410.05192
Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective
['Kaiyue Wen', 'Zhiyuan Li', 'Jason Wang', 'David Hall', 'Percy Liang', 'Tengyu Ma']
['cs.LG', 'cs.CL', 'stat.ML']
Training language models currently requires pre-determining a fixed compute budget because the typical cosine learning rate schedule depends on the total number of steps. In contrast, the Warmup-Stable-Decay (WSD) schedule uses a constant learning rate to produce a main branch of iterates that can in principle continue...
2024-10-07T16:49:39Z
45 pages,13 figures
null
null
null
null
null
null
null
null
null
2,410.0521
Preserving Multi-Modal Capabilities of Pre-trained VLMs for Improving Vision-Linguistic Compositionality
['Youngtaek Oh', 'Jae Won Cho', 'Dong-Jin Kim', 'In So Kweon', 'Junmo Kim']
['cs.CV', 'cs.AI', 'cs.CL']
In this paper, we propose a new method to enhance compositional understanding in pre-trained vision and language models (VLMs) without sacrificing performance in zero-shot multi-modal tasks. Traditional fine-tuning approaches often improve compositional reasoning at the cost of degrading multi-modal capabilities, prima...
2024-10-07T17:16:20Z
EMNLP 2024 (Long, Main). Project page: https://ytaek-oh.github.io/fsc-clip
null
null
null
null
null
null
null
null
null
2,410.05243
Navigating the Digital World as Humans Do: Universal Visual Grounding for GUI Agents
['Boyu Gou', 'Ruohan Wang', 'Boyuan Zheng', 'Yanan Xie', 'Cheng Chang', 'Yiheng Shu', 'Huan Sun', 'Yu Su']
['cs.AI', 'cs.CL', 'cs.CV']
Multimodal large language models (MLLMs) are transforming the capabilities of graphical user interface (GUI) agents, facilitating their transition from controlled simulations to complex, real-world applications across various platforms. However, the effectiveness of these agents hinges on the robustness of their ground...
2024-10-07T17:47:50Z
Accepted to ICLR 2025 (Oral). Project Homepage: https://osu-nlp-group.github.io/UGround/
null
null
Navigating the Digital World as Humans Do: Universal Visual Grounding for GUI Agents
['Boyu Gou', 'Ruohan Wang', 'Boyuan Zheng', 'Yanan Xie', 'Cheng Chang', 'Yiheng Shu', 'Huan Sun', 'Yu Su']
2,024
International Conference on Learning Representations
96
70
['Computer Science']
2,410.05255
Bridging SFT and DPO for Diffusion Model Alignment with Self-Sampling Preference Optimization
['Daoan Zhang', 'Guangchen Lan', 'Dong-Jun Han', 'Wenlin Yao', 'Xiaoman Pan', 'Hongming Zhang', 'Mingxiao Li', 'Pengcheng Chen', 'Yu Dong', 'Christopher Brinton', 'Jiebo Luo']
['cs.CV', 'cs.LG', 'I.2.6; I.2.10; I.4.0; I.5.0']
Existing post-training techniques are broadly categorized into supervised fine-tuning (SFT) and reinforcement learning (RL) methods; the former is stable during training but suffers from limited generalization, while the latter, despite its stronger generalization capability, relies on additional preference data or rew...
2024-10-07T17:56:53Z
null
null
null
SePPO: Semi-Policy Preference Optimization for Diffusion Alignment
['Daoan Zhang', 'Guangchen Lan', 'Dong-Jun Han', 'Wenlin Yao', 'Xiaoman Pan', 'Hongming Zhang', 'Mingxiao Li', 'Pengcheng Chen', 'Yu Dong', 'Christopher G. Brinton', 'Jiebo Luo']
2,024
arXiv.org
6
36
['Computer Science']
2,410.05258
Differential Transformer
['Tianzhu Ye', 'Li Dong', 'Yuqing Xia', 'Yutao Sun', 'Yi Zhu', 'Gao Huang', 'Furu Wei']
['cs.CL', 'cs.LG']
Transformer tends to overallocate attention to irrelevant context. In this work, we introduce Diff Transformer, which amplifies attention to the relevant context while canceling noise. Specifically, the differential attention mechanism calculates attention scores as the difference between two separate softmax attention...
2024-10-07T17:57:38Z
Accepted as an Oral Presentation at ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.05346
AnyAttack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models
['Jiaming Zhang', 'Junhong Ye', 'Xingjun Ma', 'Yige Li', 'Yunfan Yang', 'Yunhao Chen', 'Jitao Sang', 'Dit-Yan Yeung']
['cs.LG', 'cs.AI']
Due to their multimodal capabilities, Vision-Language Models (VLMs) have found numerous impactful applications in real-world scenarios. However, recent studies have revealed that VLMs are vulnerable to image-based adversarial attacks. Traditional targeted adversarial attacks require specific targets and labels, limitin...
2024-10-07T09:45:18Z
CVPR 2025
null
null
AnyAttack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models
['Jiaming Zhang', 'Junhong Ye', 'Xingjun Ma', 'Yige Li', 'Yunfan Yang', 'Jitao Sang', 'Dit-Yan Yeung']
2,024
null
0
35
['Computer Science']
2,410.05355
Falcon Mamba: The First Competitive Attention-free 7B Language Model
['Jingwei Zuo', 'Maksim Velikanov', 'Dhia Eddine Rhaiem', 'Ilyas Chahed', 'Younes Belkada', 'Guillaume Kunsch', 'Hakim Hacid']
['cs.CL', 'cs.AI']
In this technical report, we present Falcon Mamba 7B, a new base large language model based on the novel Mamba architecture. Falcon Mamba 7B is trained on 5.8 trillion tokens with carefully selected data mixtures. As a pure Mamba-based model, Falcon Mamba 7B surpasses leading open-weight models based on Transformers, s...
2024-10-07T15:40:45Z
null
null
null
null
null
null
null
null
null
null
2,410.05363
Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation
['Fanqing Meng', 'Jiaqi Liao', 'Xinyu Tan', 'Wenqi Shao', 'Quanfeng Lu', 'Kaipeng Zhang', 'Yu Cheng', 'Dianqi Li', 'Yu Qiao', 'Ping Luo']
['cs.CV']
Text-to-video (T2V) models like Sora have made significant strides in visualizing complex prompts, which is increasingly viewed as a promising path towards constructing the universal world simulator. Cognitive psychologists believe that the foundation for achieving this goal is the ability to understand intuitive physi...
2024-10-07T17:56:04Z
Project Page: https://phygenbench123.github.io/
null
null
Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation
['Fanqing Meng', 'Jiaqi Liao', 'Xinyu Tan', 'Wenqi Shao', 'Quanfeng Lu', 'Kaipeng Zhang', 'Yu Cheng', 'Dianqi Li', 'Yu Qiao', 'Ping Luo']
2,024
arXiv.org
27
38
['Computer Science']
2,410.0547
Image Watermarks are Removable Using Controllable Regeneration from Clean Noise
['Yepeng Liu', 'Yiren Song', 'Hai Ci', 'Yu Zhang', 'Haofan Wang', 'Mike Zheng Shou', 'Yuheng Bu']
['cs.CR', 'cs.AI', 'cs.CV']
Image watermark techniques provide an effective way to assert ownership, deter misuse, and trace content sources, which has become increasingly essential in the era of large generative models. A critical attribute of watermark techniques is their robustness against various manipulations. In this paper, we introduce a w...
2024-10-07T20:04:29Z
ICLR2025
null
null
null
null
null
null
null
null
null
2,410.05472
Neural machine translation system for Lezgian, Russian and Azerbaijani languages
['Alidar Asvarov', 'Andrey Grabovoy']
['cs.CL']
We release the first neural machine translation system for translation between Russian, Azerbaijani and the endangered Lezgian languages, as well as monolingual and parallel datasets collected and aligned for training and evaluating the system. Multiple experiments are conducted to identify how different sets of traini...
2024-10-07T20:08:10Z
null
null
10.1109/ISPRAS64596.2024.10899143
null
null
null
null
null
null
null
2,410.05474
R-Bench: Are your Large Multimodal Model Robust to Real-world Corruptions?
['Chunyi Li', 'Jianbo Zhang', 'Zicheng Zhang', 'Haoning Wu', 'Yuan Tian', 'Wei Sun', 'Guo Lu', 'Xiaohong Liu', 'Xiongkuo Min', 'Weisi Lin', 'Guangtao Zhai']
['cs.CV', 'cs.MM', 'eess.IV']
The outstanding performance of Large Multimodal Models (LMMs) has made them widely applied in vision-related tasks. However, various corruptions in the real world mean that images will not be as ideal as in simulations, presenting significant challenges for the practical application of LMMs. To address this issue, we i...
2024-10-07T20:12:08Z
null
null
null
null
null
null
null
null
null
null
2,410.0561
Structural Reasoning Improves Molecular Understanding of LLM
['Yunhui Jang', 'Jaehyung Kim', 'Sungsoo Ahn']
['cs.LG', 'cs.AI']
Recently, large language models (LLMs) have shown significant progress, approaching human perception levels. In this work, we demonstrate that despite these advances, LLMs still struggle to reason using molecular structural information. This gap is critical because many molecular properties, including functional groups...
2024-10-08T01:49:48Z
null
null
null
null
null
null
null
null
null
null
2,410.05643
TRACE: Temporal Grounding Video LLM via Causal Event Modeling
['Yongxin Guo', 'Jingyu Liu', 'Mingda Li', 'Qingbin Liu', 'Xi Chen', 'Xiaoying Tang']
['cs.CV']
Video Temporal Grounding (VTG) is a crucial capability for video understanding models and plays a vital role in downstream tasks such as video browsing and editing. To effectively handle various tasks simultaneously and enable zero-shot prediction, there is a growing trend in employing video LLMs for VTG tasks. However...
2024-10-08T02:46:30Z
ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.05677
T2V-Turbo-v2: Enhancing Video Generation Model Post-Training through Data, Reward, and Conditional Guidance Design
['Jiachen Li', 'Qian Long', 'Jian Zheng', 'Xiaofeng Gao', 'Robinson Piramuthu', 'Wenhu Chen', 'William Yang Wang']
['cs.CV', 'cs.AI']
In this paper, we focus on enhancing a diffusion-based text-to-video (T2V) model during the post-training phase by distilling a highly capable consistency model from a pretrained T2V model. Our proposed method, T2V-Turbo-v2, introduces a significant advancement by integrating various supervision signals, including high...
2024-10-08T04:30:06Z
Project Page: https://t2v-turbo-v2.github.io/
null
null
null
null
null
null
null
null
null
2,410.05954
Pyramidal Flow Matching for Efficient Video Generative Modeling
['Yang Jin', 'Zhicheng Sun', 'Ningyuan Li', 'Kun Xu', 'Kun Xu', 'Hao Jiang', 'Nan Zhuang', 'Quzhe Huang', 'Yang Song', 'Yadong Mu', 'Zhouchen Lin']
['cs.CV', 'cs.LG']
Video generation requires modeling a vast spatiotemporal space, which demands significant computational resources and data usage. To reduce the complexity, the prevailing approaches employ a cascaded architecture to avoid direct training with full resolution latent. Despite reducing computational demands, the separate ...
2024-10-08T12:10:37Z
ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.05993
Aria: An Open Multimodal Native Mixture-of-Experts Model
['Dongxu Li', 'Yudong Liu', 'Haoning Wu', 'Yue Wang', 'Zhiqi Shen', 'Bowen Qu', 'Xinyao Niu', 'Fan Zhou', 'Chengen Huang', 'Yanpeng Li', 'Chongyan Zhu', 'Xiaoyi Ren', 'Chao Li', 'Yifan Ye', 'Peng Liu', 'Lihuan Zhang', 'Hanshu Yan', 'Guoyin Wang', 'Bei Chen', 'Junnan Li']
['cs.CV']
Information comes in diverse modalities. Multimodal native AI models are essential to integrate real-world information and deliver comprehensive understanding. While proprietary multimodal native models exist, their lack of openness imposes obstacles for adoptions, let alone adaptations. To fill this gap, we introduce ...
2024-10-08T12:44:57Z
null
null
null
Aria: An Open Multimodal Native Mixture-of-Experts Model
['Dongxu Li', 'Yudong Liu', 'Haoning Wu', 'Yue Wang', 'Zhiqi Shen', 'Bowen Qu', 'Xinyao Niu', 'Guoyin Wang', 'Bei Chen', 'Junnan Li']
2,024
arXiv.org
65
24
['Computer Science']
2,410.06234
TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Data
['Jeremy Andrew Irvin', 'Emily Ruoyu Liu', 'Joyce Chuyi Chen', 'Ines Dormoy', 'Jinyoung Kim', 'Samar Khanna', 'Zhuo Zheng', 'Stefano Ermon']
['cs.CV', 'cs.AI', 'cs.LG']
Large vision and language assistants have enabled new capabilities for interpreting natural images. These approaches have recently been adapted to earth observation data, but they are only able to handle single image inputs, limiting their use for many real-world tasks. In this work, we develop a new vision and languag...
2024-10-08T17:45:51Z
Published at ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.06264
Think While You Generate: Discrete Diffusion with Planned Denoising
['Sulin Liu', 'Juno Nam', 'Andrew Campbell', 'Hannes Stärk', 'Yilun Xu', 'Tommi Jaakkola', 'Rafael Gómez-Bombarelli']
['cs.LG', 'cs.AI', 'cs.CL', 'cs.CV', 'stat.ML']
Discrete diffusion has achieved state-of-the-art performance, outperforming or approaching autoregressive models on standard benchmarks. In this work, we introduce Discrete Diffusion with Planned Denoising (DDPD), a novel framework that separates the generation process into two models: a planner and a denoiser. At infe...
2024-10-08T18:03:34Z
ICLR 2025
null
null
null
null
null
null
null
null
null
2,410.06364
Sketch to Adapt: Fine-Tunable Sketches for Efficient LLM Adaptation
['Tianyi Zhang', 'Junda Su', 'Aditya Desai', 'Oscar Wu', 'Zhaozhuo Xu', 'Anshumali Shrivastava']
['cs.LG']
Adapting pre-trained large language models (LLMs) is crucial but challenging due to their enormous size. Parameter-efficient fine-tuning (PEFT) techniques typically employ additive adapters applied to frozen model weights. To further reduce memory usage, model weights can be compressed through quantization. However, ex...
2024-10-08T20:58:24Z
null
null
null
Sketch to Adapt: Fine-Tunable Sketches for Efficient LLM Adaptation
['Tianyi Zhang', 'Junda Su', 'Aditya Desai', 'Oscar Wu', 'Zhaozhuo Xu', 'Anshumali Shrivastava']
2,024
null
0
61
['Computer Science']
2,410.06542
MedImageInsight: An Open-Source Embedding Model for General Domain Medical Imaging
['Noel C. F. Codella', 'Ying Jin', 'Shrey Jain', 'Yu Gu', 'Ho Hin Lee', 'Asma Ben Abacha', 'Alberto Santamaria-Pang', 'Will Guyman', 'Naiteek Sangani', 'Sheng Zhang', 'Hoifung Poon', 'Stephanie Hyland', 'Shruthi Bannur', 'Javier Alvarez-Valle', 'Xue Li', 'John Garrett', 'Alan McMillan', 'Gaurav Rajguru', 'Madhu Maddi',...
['eess.IV', 'cs.CV']
In this work, we present MedImageInsight, an open-source medical imaging embedding model. MedImageInsight is trained on medical images with associated text and labels across a diverse collection of domains, including X-Ray, CT, MRI, dermoscopy, OCT, fundus photography, ultrasound, histopathology, and mammography. Rigor...
2024-10-09T04:36:47Z
null
null
null
null
null
null
null
null
null
null
2,410.06551
InstantIR: Blind Image Restoration with Instant Generative Reference
['Jen-Yuan Huang', 'Haofan Wang', 'Qixun Wang', 'Xu Bai', 'Hao Ai', 'Peng Xing', 'Jen-Tse Huang']
['cs.CV', 'cs.AI', 'cs.LG']
Handling test-time unknown degradation is the major challenge in Blind Image Restoration (BIR), necessitating high model generalization. An effective strategy is to incorporate prior knowledge, either from human input or generative model. In this paper, we introduce Instant-reference Image Restoration (InstantIR), a no...
2024-10-09T05:15:29Z
null
null
null
InstantIR: Blind Image Restoration with Instant Generative Reference
['Jen-Yuan Huang', 'Haofan Wang', 'Qixun Wang', 'Xu Bai', 'Hao Ai', 'Peng Xing', 'Jen-Tse Huang']
2,024
arXiv.org
1
61
['Computer Science']
2,410.06577
Rodimus*: Breaking the Accuracy-Efficiency Trade-Off with Efficient Attentions
['Zhihao He', 'Hang Yu', 'Zi Gong', 'Shizhan Liu', 'Jianguo Li', 'Weiyao Lin']
['cs.CL']
Recent advancements in Transformer-based large language models (LLMs) have set new standards in natural language processing. However, the classical softmax attention incurs significant computational costs, leading to a $O(T)$ complexity for per-token generation, where $T$ represents the context length. This work explor...
2024-10-09T06:22:36Z
Accepted by ICLR 2025. Camera-ready Version
null
null
null
null
null
null
null
null
null
2,410.06581
Enhancing Legal Case Retrieval via Scaling High-quality Synthetic Query-Candidate Pairs
['Cheng Gao', 'Chaojun Xiao', 'Zhenghao Liu', 'Huimin Chen', 'Zhiyuan Liu', 'Maosong Sun']
['cs.IR']
Legal case retrieval (LCR) aims to provide similar cases as references for a given fact description. This task is crucial for promoting consistent judgments in similar cases, effectively enhancing judicial fairness and improving work efficiency for judges. However, existing works face two main challenges for real-world...
2024-10-09T06:26:39Z
15 pages, 3 figures, accepted by EMNLP 2024
null
null
null
null
null
null
null
null
null
2,410.06593
Towards Natural Image Matting in the Wild via Real-Scenario Prior
['Ruihao Xia', 'Yu Liang', 'Peng-Tao Jiang', 'Hao Zhang', 'Qianru Sun', 'Yang Tang', 'Bo Li', 'Pan Zhou']
['cs.CV']
Recent approaches attempt to adapt powerful interactive segmentation models, such as SAM, to interactive matting and fine-tune the models based on synthetic matting datasets. However, models trained on synthetic data fail to generalize to complex and occlusion scenes. We address this challenge by proposing a new mattin...
2024-10-09T06:43:19Z
null
null
null
null
null
null
null
null
null
null
2,410.06614
Pair-VPR: Place-Aware Pre-training and Contrastive Pair Classification for Visual Place Recognition with Vision Transformers
['Stephen Hausler', 'Peyman Moghadam']
['cs.RO', 'cs.AI', 'cs.CV']
In this work we propose a novel joint training method for Visual Place Recognition (VPR), which simultaneously learns a global descriptor and a pair classifier for re-ranking. The pair classifier can predict whether a given pair of images are from the same place or not. The network only comprises Vision Transformer com...
2024-10-09T07:09:46Z
null
null
10.1109/LRA.2025.3546512
Pair-VPR: Place-Aware Pre-Training and Contrastive Pair Classification for Visual Place Recognition With Vision Transformers
['Stephen Hausler', 'Peyman Moghadam']
2,024
IEEE Robotics and Automation Letters
4
49
['Computer Science']
2,410.06734
MimicTalk: Mimicking a personalized and expressive 3D talking face in minutes
['Zhenhui Ye', 'Tianyun Zhong', 'Yi Ren', 'Ziyue Jiang', 'Jiawei Huang', 'Rongjie Huang', 'Jinglin Liu', 'Jinzheng He', 'Chen Zhang', 'Zehan Wang', 'Xize Chen', 'Xiang Yin', 'Zhou Zhao']
['cs.CV']
Talking face generation (TFG) aims to animate a target identity's face to create realistic talking videos. Personalized TFG is a variant that emphasizes the perceptual identity similarity of the synthesized result (from the perspective of appearance and talking style). While previous works typically solve this problem ...
2024-10-09T10:12:37Z
Accepted by NeurIPS 2024
null
null
null
null
null
null
null
null
null
2,410.06885
F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching
['Yushen Chen', 'Zhikang Niu', 'Ziyang Ma', 'Keqi Deng', 'Chunhui Wang', 'Jian Zhao', 'Kai Yu', 'Xie Chen']
['eess.AS', 'cs.SD']
This paper introduces F5-TTS, a fully non-autoregressive text-to-speech system based on flow matching with Diffusion Transformer (DiT). Without requiring complex designs such as duration model, text encoder, and phoneme alignment, the text input is simply padded with filler tokens to the same length as input speech, an...
2024-10-09T13:46:34Z
17 pages, 9 tables, 3 figures
null
null
null
null
null
null
null
null
null
2,410.06961
Self-Boosting Large Language Models with Synthetic Preference Data
['Qingxiu Dong', 'Li Dong', 'Xingxing Zhang', 'Zhifang Sui', 'Furu Wei']
['cs.CL', 'cs.AI']
Through alignment with human preferences, Large Language Models (LLMs) have advanced significantly in generating honest, harmless, and helpful responses. However, collecting high-quality preference data is a resource-intensive and creativity-demanding process, especially for the continual improvement of LLMs. We introd...
2024-10-09T14:57:31Z
null
null
null
null
null
null
null
null
null
null
2,410.07002
CursorCore: Assist Programming through Aligning Anything
['Hao Jiang', 'Qi Liu', 'Rui Li', 'Shengyu Ye', 'Shijin Wang']
['cs.CL', 'cs.AI', 'cs.SE']
Large language models have been successfully applied to programming assistance tasks, such as code completion, code insertion, and instructional code editing. However, these applications remain insufficiently automated and struggle to effectively integrate various types of information during the programming process, in...
2024-10-09T15:45:52Z
null
null
null
null
null
null
null
null
null
null
2,410.07064
Data Selection via Optimal Control for Language Models
['Yuxian Gu', 'Li Dong', 'Hongning Wang', 'Yaru Hao', 'Qingxiu Dong', 'Furu Wei', 'Minlie Huang']
['cs.CL']
This work investigates the selection of high-quality pre-training data from massive corpora to enhance LMs' capabilities for downstream usage. We formulate data selection as a generalized Optimal Control problem, which can be solved theoretically by Pontryagin's Maximum Principle (PMP), yielding a set of necessary cond...
2024-10-09T17:06:57Z
ICLR 2025 Oral
null
null
Data Selection via Optimal Control for Language Models
['Yuxian Gu', 'Li Dong', 'Hongning Wang', 'Y. Hao', 'Qingxiu Dong', 'Furu Wei', 'Minlie Huang']
2,024
International Conference on Learning Representations
9
87
['Computer Science']
2,410.07095
MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering
['Jun Shern Chan', 'Neil Chowdhury', 'Oliver Jaffe', 'James Aung', 'Dane Sherburn', 'Evan Mays', 'Giulio Starace', 'Kevin Liu', 'Leon Maksin', 'Tejal Patwardhan', 'Lilian Weng', 'Aleksander Mądry']
['cs.CL']
We introduce MLE-bench, a benchmark for measuring how well AI agents perform at machine learning engineering. To this end, we curate 75 ML engineering-related competitions from Kaggle, creating a diverse set of challenging tasks that test real-world ML engineering skills such as training models, preparing datasets, and...
2024-10-09T17:34:27Z
10 pages, 17 pages appendix. Equal contribution by first seven authors, authors randomized. ICLR version
null
null
null
null
null
null
null
null
null
2,410.07133
EvolveDirector: Approaching Advanced Text-to-Image Generation with Large Vision-Language Models
['Rui Zhao', 'Hangjie Yuan', 'Yujie Wei', 'Shiwei Zhang', 'Yuchao Gu', 'Lingmin Ran', 'Xiang Wang', 'Zhangjie Wu', 'Junhao Zhang', 'Yingya Zhang', 'Mike Zheng Shou']
['cs.CV']
Recent advancements in generation models have showcased remarkable capabilities in generating fantastic content. However, most of them are trained on proprietary high-quality data, and some models withhold their parameters and only provide accessible application programming interfaces (APIs), limiting their benefits fo...
2024-10-09T17:52:28Z
null
null
null
EvolveDirector: Approaching Advanced Text-to-Image Generation with Large Vision-Language Models
['Rui Zhao', 'Hangjie Yuan', 'Yujie Wei', 'Shiwei Zhang', 'Yuchao Gu', 'L. Ran', 'Xiang Wang', 'Zhangjie Wu', 'Junhao Zhang', 'Yingya Zhang', 'Mike Zheng Shou']
2,024
Neural Information Processing Systems
4
76
['Computer Science']
2,410.07153
CHASE: Learning Convex Hull Adaptive Shift for Skeleton-based Multi-Entity Action Recognition
['Yuhang Wen', 'Mengyuan Liu', 'Songtao Wu', 'Beichen Ding']
['cs.CV', 'cs.LG']
Skeleton-based multi-entity action recognition is a challenging task aiming to identify interactive actions or group activities involving multiple diverse entities. Existing models for individuals often fall short in this task due to the inherent distribution discrepancies among entity skeletons, leading to suboptimal ...
2024-10-09T17:55:43Z
NeurIPS 2024 Camera-ready Version. Project Website: https://necolizer.github.io/CHASE/
null
null
CHASE: Learning Convex Hull Adaptive Shift for Skeleton-based Multi-Entity Action Recognition
['Yuhang Wen', 'Mengyuan Liu', 'Songtao Wu', 'Beichen Ding']
2,024
Neural Information Processing Systems
1
102
['Computer Science']
2,410.07157
InstructG2I: Synthesizing Images from Multimodal Attributed Graphs
['Bowen Jin', 'Ziqi Pang', 'Bingjun Guo', 'Yu-Xiong Wang', 'Jiaxuan You', 'Jiawei Han']
['cs.AI', 'cs.CL', 'cs.CV', 'cs.LG', 'cs.SI']
In this paper, we approach an overlooked yet critical task Graph2Image: generating images from multimodal attributed graphs (MMAGs). This task poses significant challenges due to the explosion in graph size, dependencies among graph entities, and the need for controllability in graph conditions. To address these challe...
2024-10-09T17:56:15Z
16 pages
NeurIPs 2024
null
null
null
null
null
null
null
null
2,410.07163
Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
['Chongyu Fan', 'Jiancheng Liu', 'Licong Lin', 'Jinghan Jia', 'Ruiqi Zhang', 'Song Mei', 'Sijia Liu']
['cs.CL', 'cs.AI', 'cs.LG']
This work studies the problem of large language model (LLM) unlearning, aiming to remove unwanted data influences (e.g., copyrighted or harmful content) while preserving model utility. Despite the increasing demand for unlearning, a technically-grounded optimization framework is lacking. Gradient ascent (GA)-type metho...
2024-10-09T17:58:12Z
null
null
null
null
null
null
null
null
null
null
2,410.07167
Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate
['Qidong Huang', 'Xiaoyi Dong', 'Pan Zhang', 'Yuhang Zang', 'Yuhang Cao', 'Jiaqi Wang', 'Dahua Lin', 'Weiming Zhang', 'Nenghai Yu']
['cs.CV', 'cs.CL']
We present the Modality Integration Rate (MIR), an effective, robust, and generalized metric to indicate the multi-modal pre-training quality of Large Vision Language Models (LVLMs). Large-scale pre-training plays a critical role in building capable LVLMs, while evaluating its training quality without the costly superv...
2024-10-09T17:59:04Z
Project page: https://github.com/shikiw/Modality-Integration-Rate
null
null
Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate
['Qidong Huang', 'Xiao-wen Dong', 'Pan Zhang', 'Yuhang Zang', 'Yuhang Cao', 'Jiaqi Wang', 'Dahua Lin', 'Weiming Zhang', 'Neng H. Yu']
2,024
arXiv.org
9
53
['Computer Science']
2,410.07168
Sylber: Syllabic Embedding Representation of Speech from Raw Audio
['Cheol Jun Cho', 'Nicholas Lee', 'Akshat Gupta', 'Dhruv Agarwal', 'Ethan Chen', 'Alan W Black', 'Gopala K. Anumanchipalli']
['cs.CL', 'cs.SD', 'eess.AS']
Syllables are compositional units of spoken language that efficiently structure human speech perception and production. However, current neural speech representations lack such structure, resulting in dense token sequences that are costly to process. To bridge this gap, we propose a new model, Sylber, that produces spe...
2024-10-09T17:59:04Z
Accepted at ICLR 2025
null
null
Sylber: Syllabic Embedding Representation of Speech from Raw Audio
['Cheol Jun Cho', 'Nicholas Lee', 'Akshat Gupta', 'Dhruv Agarwal', 'Ethan Chen', 'Alan W. Black', 'G. Anumanchipalli']
2,024
International Conference on Learning Representations
4
69
['Computer Science', 'Engineering']
2,410.07169
VIP: Vision Instructed Pre-training for Robotic Manipulation
['Zhuoling Li', 'Liangliang Ren', 'Jinrong Yang', 'Yong Zhao', 'Xiaoyang Wu', 'Zhenhua Xu', 'Xiang Bai', 'Hengshuang Zhao']
['cs.RO']
The effectiveness of scaling up training data in robotic manipulation is still limited. A primary challenge in manipulation is the tasks are diverse, and the trained policy would be confused if the task targets are not specified clearly. Existing works primarily rely on text instruction to describe targets. However, we...
2024-10-09T17:59:06Z
null
null
null
VIP: Vision Instructed Pre-training for Robotic Manipulation
['Zhuoling Li', 'Liangliang Ren', 'Jinrong Yang', 'Yong Zhao', 'Xiaoyang Wu', 'Zhenhua Xu', 'Xiang Bai', 'Hengshuang Zhao']
2,024
null
0
0
['Computer Science']
2,410.07171
IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation
['Xinchen Zhang', 'Ling Yang', 'Guohao Li', 'Yaqi Cai', 'Jiake Xie', 'Yong Tang', 'Yujiu Yang', 'Mengdi Wang', 'Bin Cui']
['cs.CV']
Advanced diffusion models like RPG, Stable Diffusion 3 and FLUX have made notable strides in compositional text-to-image generation. However, these methods typically exhibit distinct strengths for compositional generation, with some excelling in handling attribute binding and others in spatial relationships. This dispa...
2024-10-09T17:59:13Z
ICLR 2025. Project: https://github.com/YangLing0818/IterComp
null
null
null
null
null
null
null
null
null
2,410.07173
Better Language Models Exhibit Higher Visual Alignment
['Jona Ruthardt', 'Gertjan J. Burghouts', 'Serge Belongie', 'Yuki M. Asano']
['cs.CL', 'cs.AI', 'cs.CV']
How well do text-only Large Language Models (LLMs) naturally align with the visual world? We provide the first direct analysis by utilizing frozen text representations in a discriminative vision-language model framework and measuring zero-shot generalization on unseen classes. We find decoder-based LLMs exhibit high in...
2024-10-09T17:59:33Z
null
null
null
Better Language Models Exhibit Higher Visual Alignment
['Jona Ruthardt', 'G. Burghouts', 'Serge J. Belongie', 'Yuki M. Asano']
2,024
null
0
58
['Computer Science']