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
2,502.11187
TituLLMs: A Family of Bangla LLMs with Comprehensive Benchmarking
['Shahriar Kabir Nahin', 'Rabindra Nath Nandi', 'Sagor Sarker', 'Quazi Sarwar Muhtaseem', 'Md Kowsher', 'Apu Chandraw Shill', 'Md Ibrahim', 'Mehadi Hasan Menon', 'Tareq Al Muntasir', 'Firoj Alam']
['cs.CL', 'cs.AI', '68T50', 'F.2.2; I.2.7']
In this paper, we present TituLLMs, the first large pretrained Bangla LLMs, available in 1b and 3b parameter sizes. Due to computational constraints during both training and inference, we focused on smaller models. To train TituLLMs, we collected a pretraining dataset of approximately ~37 billion tokens. We extended th...
2025-02-16T16:22:23Z
LLMs, Benchmarking, Large Language Models, Bangla, BanglaLLMs
null
null
TituLLMs: A Family of Bangla LLMs with Comprehensive Benchmarking
['Shahriar Kabir Nahin', 'R. N. Nandi', 'Sagor Sarker', 'Quazi Sarwar Muhtaseem', 'Md Kowsher', 'Apu Chandraw Shill', 'Md Ibrahim', 'Mehadi Hasan Menon', 'Tareq Al Muntasir', 'Firoj Alam']
2,025
arXiv.org
0
86
['Computer Science']
2,502.11191
Primus: A Pioneering Collection of Open-Source Datasets for Cybersecurity LLM Training
['Yao-Ching Yu', 'Tsun-Han Chiang', 'Cheng-Wei Tsai', 'Chien-Ming Huang', 'Wen-Kwang Tsao']
['cs.CR', 'cs.AI', 'cs.CL']
Large Language Models (LLMs) have shown remarkable advancements in specialized fields such as finance, law, and medicine. However, in cybersecurity, we have noticed a lack of open-source datasets, with a particular lack of high-quality cybersecurity pretraining corpora, even though much research indicates that LLMs acq...
2025-02-16T16:34:49Z
null
null
null
null
null
null
null
null
null
null
2,502.11223
Asymmetric Conflict and Synergy in Post-training for LLM-based Multilingual Machine Translation
['Tong Zheng', 'Yan Wen', 'Huiwen Bao', 'Junfeng Guo', 'Heng Huang']
['cs.CL']
The emergence of Large Language Models (LLMs) has advanced the multilingual machine translation (MMT), yet the Curse of Multilinguality (CoM) remains a major challenge. Existing work in LLM-based MMT typically mitigates this issue via scaling up training and computation budget, which raises a critical question: Is scal...
2025-02-16T18:06:58Z
22 pages
null
null
null
null
null
null
null
null
null
2,502.11275
Cuckoo: An IE Free Rider Hatched by Massive Nutrition in LLM's Nest
['Letian Peng', 'Zilong Wang', 'Feng Yao', 'Jingbo Shang']
['cs.CL']
Massive high-quality data, both pre-training raw texts and post-training annotations, have been carefully prepared to incubate advanced large language models (LLMs). In contrast, for information extraction (IE), pre-training data, such as BIO-tagged sequences, are hard to scale up. We show that IE models can act as fre...
2025-02-16T21:32:20Z
null
null
null
null
null
null
null
null
null
null
2,502.11431
Any Information Is Just Worth One Single Screenshot: Unifying Search With Visualized Information Retrieval
['Ze Liu', 'Zhengyang Liang', 'Junjie Zhou', 'Zheng Liu', 'Defu Lian']
['cs.CL']
With the popularity of multimodal techniques, it receives growing interests to acquire useful information in visual forms. In this work, we formally define an emerging IR paradigm called \textit{Visualized Information Retrieval}, or \textbf{Vis-IR}, where multimodal information, such as texts, images, tables and charts...
2025-02-17T04:40:15Z
null
null
null
null
null
null
null
null
null
null
2,502.11492
Why Vision Language Models Struggle with Visual Arithmetic? Towards Enhanced Chart and Geometry Understanding
['Kung-Hsiang Huang', 'Can Qin', 'Haoyi Qiu', 'Philippe Laban', 'Shafiq Joty', 'Caiming Xiong', 'Chien-Sheng Wu']
['cs.AI', 'cs.CL', 'cs.CV']
Vision Language Models (VLMs) have achieved remarkable progress in multimodal tasks, yet they often struggle with visual arithmetic, seemingly simple capabilities like object counting or length comparison, which are essential for relevant complex tasks like chart understanding and geometric reasoning. In this work, we ...
2025-02-17T06:54:49Z
Code and data are available at https://github.com/SalesforceAIResearch/CogAlign
null
null
Why Vision Language Models Struggle with Visual Arithmetic? Towards Enhanced Chart and Geometry Understanding
['Kung-Hsiang Huang', 'Can Qin', 'Haoyi Qiu', 'Philippe Laban', 'Shafiq Joty', 'Caiming Xiong', 'Chien-Sheng Wu']
2,025
arXiv.org
5
41
['Computer Science']
2,502.1152
AURORA:Automated Training Framework of Universal Process Reward Models via Ensemble Prompting and Reverse Verification
['Xiaoyu Tan', 'Tianchu Yao', 'Chao Qu', 'Bin Li', 'Minghao Yang', 'Dakuan Lu', 'Haozhe Wang', 'Xihe Qiu', 'Wei Chu', 'Yinghui Xu', 'Yuan Qi']
['cs.CL']
The reasoning capabilities of advanced large language models (LLMs) like o1 have revolutionized artificial intelligence applications. Nevertheless, evaluating and optimizing complex reasoning processes remain significant challenges due to diverse policy distributions and the inherent limitations of human effort and acc...
2025-02-17T07:41:27Z
Under Review
null
null
AURORA:Automated Training Framework of Universal Process Reward Models via Ensemble Prompting and Reverse Verification
['Xiaoyu Tan', 'Tianchu Yao', 'Chao Qu', 'Bin Li', 'Minghao Yang', 'Dakuan Lu', 'Haozhe Wang', 'Xihe Qiu', 'Wei Chu', 'Yinghui Xu', 'Yuan Qi']
2,025
arXiv.org
2
54
['Computer Science']
2,502.11537
Uncovering Untapped Potential in Sample-Efficient World Model Agents
['Lior Cohen', 'Kaixin Wang', 'Bingyi Kang', 'Uri Gadot', 'Shie Mannor']
['cs.LG', 'cs.AI']
World model (WM) agents enable sample-efficient reinforcement learning by learning policies entirely from simulated experience. However, existing token-based world models (TBWMs) are limited to visual inputs and discrete actions, restricting their adoption and applicability. Moreover, although both intrinsic motivation...
2025-02-17T08:06:10Z
null
null
null
Uncovering Untapped Potential in Sample-Efficient World Model Agents
['Lior Cohen', 'Kaixin Wang', 'Bingyi Kang', 'Uri Gadot', 'Shie Mannor']
2,025
null
0
62
['Computer Science']
2,502.11689
Improve LLM-as-a-Judge Ability as a General Ability
['Jiachen Yu', 'Shaoning Sun', 'Xiaohui Hu', 'Jiaxu Yan', 'Kaidong Yu', 'Xuelong Li']
['cs.CL']
LLM-as-a-Judge leverages the generative and reasoning capabilities of large language models (LLMs) to evaluate LLM responses across diverse scenarios, providing accurate preference signals. This approach plays a vital role in aligning LLMs with human values, ensuring ethical and reliable AI outputs that align with soci...
2025-02-17T11:28:43Z
null
null
null
null
null
null
null
null
null
null
2,502.12025
SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities
['Fengqing Jiang', 'Zhangchen Xu', 'Yuetai Li', 'Luyao Niu', 'Zhen Xiang', 'Bo Li', 'Bill Yuchen Lin', 'Radha Poovendran']
['cs.AI', 'cs.CL']
Emerging large reasoning models (LRMs), such as DeepSeek-R1 models, leverage long chain-of-thought (CoT) reasoning to generate structured intermediate steps, enhancing their reasoning capabilities. However, long CoT does not inherently guarantee safe outputs, potentially leading to harmful consequences such as the intr...
2025-02-17T16:57:56Z
null
null
null
SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities
['Fengqing Jiang', 'Zhangchen Xu', 'Yuetai Li', 'Luyao Niu', 'Zhen Xiang', 'Bo Li', 'Bill Yuchen Lin', 'Radha Poovendran']
2,025
arXiv.org
28
44
['Computer Science']
2,502.1208
HumanGif: Single-View Human Diffusion with Generative Prior
['Shoukang Hu', 'Takuya Narihira', 'Kazumi Fukuda', 'Ryosuke Sawata', 'Takashi Shibuya', 'Yuki Mitsufuji']
['cs.CV']
Previous 3D human creation methods have made significant progress in synthesizing view-consistent and temporally aligned results from sparse-view images or monocular videos. However, it remains challenging to produce perpetually realistic, view-consistent, and temporally coherent human avatars from a single image, as l...
2025-02-17T17:55:27Z
Project page: https://skhu101.github.io/HumanGif/
null
null
null
null
null
null
null
null
null
2,502.12082
AdaSplash: Adaptive Sparse Flash Attention
['Nuno Gonçalves', 'Marcos Treviso', 'André F. T. Martins']
['cs.CL', 'cs.LG']
The computational cost of softmax-based attention in transformers limits their applicability to long-context tasks. Adaptive sparsity, of which $\alpha$-entmax attention is an example, offers a flexible data-dependent alternative, but existing implementations are inefficient and do not leverage the sparsity to obtain r...
2025-02-17T17:56:23Z
Accepted as spotlight in ICML 2025
null
null
null
null
null
null
null
null
null
2,502.1213
Scaling Autonomous Agents via Automatic Reward Modeling And Planning
['Zhenfang Chen', 'Delin Chen', 'Rui Sun', 'Wenjun Liu', 'Chuang Gan']
['cs.AI']
Large language models (LLMs) have demonstrated remarkable capabilities across a range of text-generation tasks. However, LLMs still struggle with problems requiring multi-step decision-making and environmental feedback, such as online shopping, scientific reasoning, and mathematical problem-solving. Unlike pure text da...
2025-02-17T18:49:25Z
ICLR2025, Project page: https://armap-agent.github.io
null
null
null
null
null
null
null
null
null
2,502.12135
MagicArticulate: Make Your 3D Models Articulation-Ready
['Chaoyue Song', 'Jianfeng Zhang', 'Xiu Li', 'Fan Yang', 'Yiwen Chen', 'Zhongcong Xu', 'Jun Hao Liew', 'Xiaoyang Guo', 'Fayao Liu', 'Jiashi Feng', 'Guosheng Lin']
['cs.CV', 'cs.GR']
With the explosive growth of 3D content creation, there is an increasing demand for automatically converting static 3D models into articulation-ready versions that support realistic animation. Traditional approaches rely heavily on manual annotation, which is both time-consuming and labor-intensive. Moreover, the lack ...
2025-02-17T18:53:27Z
Project: https://chaoyuesong.github.io/MagicArticulate
null
null
null
null
null
null
null
null
null
2,502.12138
FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Views
['Shangzhan Zhang', 'Jianyuan Wang', 'Yinghao Xu', 'Nan Xue', 'Christian Rupprecht', 'Xiaowei Zhou', 'Yujun Shen', 'Gordon Wetzstein']
['cs.CV']
We present FLARE, a feed-forward model designed to infer high-quality camera poses and 3D geometry from uncalibrated sparse-view images (i.e., as few as 2-8 inputs), which is a challenging yet practical setting in real-world applications. Our solution features a cascaded learning paradigm with camera pose serving as th...
2025-02-17T18:54:05Z
CVPR 2025. Website: https://zhanghe3z.github.io/FLARE/
null
null
null
null
null
null
null
null
null
2,502.12143
Small Models Struggle to Learn from Strong Reasoners
['Yuetai Li', 'Xiang Yue', 'Zhangchen Xu', 'Fengqing Jiang', 'Luyao Niu', 'Bill Yuchen Lin', 'Bhaskar Ramasubramanian', 'Radha Poovendran']
['cs.AI']
Large language models (LLMs) excel in complex reasoning tasks, and distilling their reasoning capabilities into smaller models has shown promise. However, we uncover an interesting phenomenon, which we term the Small Model Learnability Gap: small models ($\leq$3B parameters) do not consistently benefit from long chain-...
2025-02-17T18:56:15Z
null
null
null
null
null
null
null
null
null
null
2,502.12147
Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction
['Xiang Fu', 'Brandon M. Wood', 'Luis Barroso-Luque', 'Daniel S. Levine', 'Meng Gao', 'Misko Dzamba', 'C. Lawrence Zitnick']
['physics.comp-ph', 'cs.LG']
Machine learning interatomic potentials (MLIPs) have become increasingly effective at approximating quantum mechanical calculations at a fraction of the computational cost. However, lower errors on held out test sets do not always translate to improved results on downstream physical property prediction tasks. In this p...
2025-02-17T18:57:32Z
20 pages, 14 figures, 6 tables
null
null
null
null
null
null
null
null
null
2,502.12148
HermesFlow: Seamlessly Closing the Gap in Multimodal Understanding and Generation
['Ling Yang', 'Xinchen Zhang', 'Ye Tian', 'Chenming Shang', 'Minghao Xu', 'Wentao Zhang', 'Bin Cui']
['cs.CV']
The remarkable success of the autoregressive paradigm has made significant advancement in Multimodal Large Language Models (MLLMs), with powerful models like Show-o, Transfusion and Emu3 achieving notable progress in unified image understanding and generation. For the first time, we uncover a common phenomenon: the und...
2025-02-17T18:57:51Z
Code: https://github.com/Gen-Verse/HermesFlow
null
null
null
null
null
null
null
null
null
2,502.1217
MUDDFormer: Breaking Residual Bottlenecks in Transformers via Multiway Dynamic Dense Connections
['Da Xiao', 'Qingye Meng', 'Shengping Li', 'Xingyuan Yuan']
['cs.LG', 'cs.AI', 'cs.CL']
We propose MUltiway Dynamic Dense (MUDD) connections, a simple yet effective method to address the limitations of residual connections and enhance cross-layer information flow in Transformers. Unlike existing dense connection approaches with static and shared connection weights, MUDD generates connection weights dynami...
2025-02-13T10:26:27Z
Accepted to the 42nd International Conference on Machine Learning (ICML'25)
null
null
null
null
null
null
null
null
null
2,502.12202
To Think or Not to Think: Exploring the Unthinking Vulnerability in Large Reasoning Models
['Zihao Zhu', 'Hongbao Zhang', 'Ruotong Wang', 'Ke Xu', 'Siwei Lyu', 'Baoyuan Wu']
['cs.CL', 'cs.AI', 'cs.LG']
Large Reasoning Models (LRMs) are designed to solve complex tasks by generating explicit reasoning traces before producing final answers. However, we reveal a critical vulnerability in LRMs -- termed Unthinking Vulnerability -- wherein the thinking process can be bypassed by manipulating special delimiter tokens. It is...
2025-02-16T10:45:56Z
39 pages, 13 tables, 14 figures
null
null
null
null
null
null
null
null
null
2,502.12221
ReF Decompile: Relabeling and Function Call Enhanced Decompile
['Yunlong Feng', 'Bohan Li', 'Xiaoming Shi', 'Qingfu Zhu', 'Wanxiang Che']
['cs.SE']
The goal of decompilation is to convert compiled low-level code (e.g., assembly code) back into high-level programming languages, enabling analysis in scenarios where source code is unavailable. This task supports various reverse engineering applications, such as vulnerability identification, malware analysis, and lega...
2025-02-17T12:38:57Z
null
null
null
null
null
null
null
null
null
null
2,502.12342
REAL-MM-RAG: A Real-World Multi-Modal Retrieval Benchmark
['Navve Wasserman', 'Roi Pony', 'Oshri Naparstek', 'Adi Raz Goldfarb', 'Eli Schwartz', 'Udi Barzelay', 'Leonid Karlinsky']
['cs.IR', 'cs.CV']
Accurate multi-modal document retrieval is crucial for Retrieval-Augmented Generation (RAG), yet existing benchmarks do not fully capture real-world challenges with their current design. We introduce REAL-MM-RAG, an automatically generated benchmark designed to address four key properties essential for real-world retri...
2025-02-17T22:10:47Z
null
null
null
REAL-MM-RAG: A Real-World Multi-Modal Retrieval Benchmark
['Navve Wasserman', 'Roi Pony', 'O. Naparstek', 'Adi Raz Goldfarb', 'Eli Schwartz', 'Udi Barzelay', 'Leonid Karlinsky']
2,025
arXiv.org
3
38
['Computer Science']
2,502.12404
WMT24++: Expanding the Language Coverage of WMT24 to 55 Languages & Dialects
['Daniel Deutsch', 'Eleftheria Briakou', 'Isaac Caswell', 'Mara Finkelstein', 'Rebecca Galor', 'Juraj Juraska', 'Geza Kovacs', 'Alison Lui', 'Ricardo Rei', 'Jason Riesa', 'Shruti Rijhwani', 'Parker Riley', 'Elizabeth Salesky', 'Firas Trabelsi', 'Stephanie Winkler', 'Biao Zhang', 'Markus Freitag']
['cs.CL']
As large language models (LLM) become more and more capable in languages other than English, it is important to collect benchmark datasets in order to evaluate their multilingual performance, including on tasks like machine translation (MT). In this work, we extend the WMT24 dataset to cover 55 languages by collecting ...
2025-02-18T00:39:30Z
null
null
null
WMT24++: Expanding the Language Coverage of WMT24 to 55 Languages & Dialects
['Daniel Deutsch', 'Eleftheria Briakou', 'Isaac Caswell', 'Mara Finkelstein', 'Rebecca Galor', 'Juraj Juraska', 'Geza Kovacs', 'Alison Lui', 'Ricardo Rei', 'Jason Riesa', 'Shruti Rijhwani', 'Parker Riley', 'Elizabeth Salesky', 'Firas Trabelsi', 'Stephanie Winkler', 'Biao Zhang', 'Markus Freitag']
2,025
arXiv.org
11
0
['Computer Science']
2,502.12485
Safe at the Margins: A General Approach to Safety Alignment in Low-Resource English Languages -- A Singlish Case Study
['Isaac Lim', 'Shaun Khoo', 'Roy Ka-Wei Lee', 'Watson Chua', 'Jia Yi Goh', 'Jessica Foo']
['cs.CL', 'cs.AI']
Ensuring the safety of Large Language Models (LLMs) in diverse linguistic settings remains challenging, particularly for low-resource languages. Existing safety alignment methods are English-centric, limiting their effectiveness. We systematically compare Supervised Fine-Tuning (SFT), Direct Preference Optimization (DP...
2025-02-18T03:11:06Z
null
null
null
Safe at the Margins: A General Approach to Safety Alignment in Low-Resource English Languages - A Singlish Case Study
['Isaac Lim', 'Shaun Khoo', 'W. Chua', 'Goh Jiayi', 'Jessica Foo']
2,025
arXiv.org
0
27
['Computer Science']
2,502.12524
YOLOv12: Attention-Centric Real-Time Object Detectors
['Yunjie Tian', 'Qixiang Ye', 'David Doermann']
['cs.CV', 'cs.AI']
Enhancing the network architecture of the YOLO framework has been crucial for a long time, but has focused on CNN-based improvements despite the proven superiority of attention mechanisms in modeling capabilities. This is because attention-based models cannot match the speed of CNN-based models. This paper proposes an ...
2025-02-18T04:20:14Z
https://github.com/sunsmarterjie/yolov12
null
null
null
null
null
null
null
null
null
2,502.12572
TechSinger: Technique Controllable Multilingual Singing Voice Synthesis via Flow Matching
['Wenxiang Guo', 'Yu Zhang', 'Changhao Pan', 'Rongjie Huang', 'Li Tang', 'Ruiqi Li', 'Zhiqing Hong', 'Yongqi Wang', 'Zhou Zhao']
['cs.SD']
Singing voice synthesis has made remarkable progress in generating natural and high-quality voices. However, existing methods rarely provide precise control over vocal techniques such as intensity, mixed voice, falsetto, bubble, and breathy tones, thus limiting the expressive potential of synthetic voices. We introduce...
2025-02-18T06:25:07Z
Accepted by AAAI 2025
null
null
null
null
null
null
null
null
null
2,502.12579
CHATS: Combining Human-Aligned Optimization and Test-Time Sampling for Text-to-Image Generation
['Minghao Fu', 'Guo-Hua Wang', 'Liangfu Cao', 'Qing-Guo Chen', 'Zhao Xu', 'Weihua Luo', 'Kaifu Zhang']
['cs.CV']
Diffusion models have emerged as a dominant approach for text-to-image generation. Key components such as the human preference alignment and classifier-free guidance play a crucial role in ensuring generation quality. However, their independent application in current text-to-image models continues to face significant c...
2025-02-18T06:31:08Z
ICML 2025. The code is publicly available at https://github.com/AIDC-AI/CHATS
null
null
null
null
null
null
null
null
null
2,502.12614
Label Drop for Multi-Aspect Relation Modeling in Universal Information Extraction
['Lu Yang', 'Jiajia Li', 'En Ci', 'Lefei Zhang', 'Zuchao Li', 'Ping Wang']
['cs.CL', 'cs.AI']
Universal Information Extraction (UIE) has garnered significant attention due to its ability to address model explosion problems effectively. Extractive UIE can achieve strong performance using a relatively small model, making it widely adopted. Extractive UIEs generally rely on task instructions for different tasks, i...
2025-02-18T07:53:26Z
Accepted to NAACL-main 2025
null
null
null
null
null
null
null
null
null
2,502.12671
Baichuan-M1: Pushing the Medical Capability of Large Language Models
['Bingning Wang', 'Haizhou Zhao', 'Huozhi Zhou', 'Liang Song', 'Mingyu Xu', 'Wei Cheng', 'Xiangrong Zeng', 'Yupeng Zhang', 'Yuqi Huo', 'Zecheng Wang', 'Zhengyun Zhao', 'Da Pan', 'Fei Kou', 'Fei Li', 'Fuzhong Chen', 'Guosheng Dong', 'Han Liu', 'Hongda Zhang', 'Jin He', 'Jinjie Yang', 'Kangxi Wu', 'Kegeng Wu', 'Lei Su', ...
['cs.CL']
The current generation of large language models (LLMs) is typically designed for broad, general-purpose applications, while domain-specific LLMs, especially in vertical fields like medicine, remain relatively scarce. In particular, the development of highly efficient and practical LLMs for the medical domain is challen...
2025-02-18T09:21:12Z
33 pages, technical report
null
null
Baichuan-M1: Pushing the Medical Capability of Large Language Models
['Bingning Wang', 'Haizhou Zhao', 'Huozhi Zhou', 'Liang Song', 'Mingyu Xu', 'Wei Cheng', 'Xiangrong Zeng', 'Yupeng Zhang', 'Yuqi Huo', 'Zecheng Wang', 'Zhengyun Zhao', 'Da Pan', 'Fan Yang', 'Fei Kou', 'Fei Li', 'Fuzhong Chen', 'Guosheng Dong', 'Han Liu', 'Hongda Zhang', 'Jin He', 'Jinjie Yang', 'Kangxi Wu', 'Kegeng Wu'...
2,025
arXiv.org
10
74
['Computer Science']
2,502.12759
High-Fidelity Music Vocoder using Neural Audio Codecs
['Luca A. Lanzendörfer', 'Florian Grötschla', 'Michael Ungersböck', 'Roger Wattenhofer']
['cs.SD', 'cs.LG']
While neural vocoders have made significant progress in high-fidelity speech synthesis, their application on polyphonic music has remained underexplored. In this work, we propose DisCoder, a neural vocoder that leverages a generative adversarial encoder-decoder architecture informed by a neural audio codec to reconstru...
2025-02-18T11:25:46Z
Accepted at ICASSP 2025
null
null
High-Fidelity Music Vocoder using Neural Audio Codecs
['Luca A. Lanzendörfer', 'Florian Grötschla', 'Michael Ungersböck', 'R. Wattenhofer']
2,025
IEEE International Conference on Acoustics, Speech, and Signal Processing
1
0
['Computer Science']
2,502.12835
Subword models struggle with word learning, but surprisal hides it
['Bastian Bunzeck', 'Sina Zarrieß']
['cs.CL']
We study word learning in subword and character language models with the psycholinguistic lexical decision task. While subword LMs struggle to discern words and non-words with high accuracy, character LMs solve this task easily and consistently. Only when supplied with further contexts do subword LMs perform similarly ...
2025-02-18T13:09:16Z
Accepted to ACL 2025 (Main)
null
null
null
null
null
null
null
null
null
2,502.12892
Archetypal SAE: Adaptive and Stable Dictionary Learning for Concept Extraction in Large Vision Models
['Thomas Fel', 'Ekdeep Singh Lubana', 'Jacob S. Prince', 'Matthew Kowal', 'Victor Boutin', 'Isabel Papadimitriou', 'Binxu Wang', 'Martin Wattenberg', 'Demba Ba', 'Talia Konkle']
['cs.CV']
Sparse Autoencoders (SAEs) have emerged as a powerful framework for machine learning interpretability, enabling the unsupervised decomposition of model representations into a dictionary of abstract, human-interpretable concepts. However, we reveal a fundamental limitation: existing SAEs exhibit severe instability, as i...
2025-02-18T14:29:11Z
null
Proceedings of the 42nd International Conference on Machine Learning (ICML), 2025
null
Archetypal SAE: Adaptive and Stable Dictionary Learning for Concept Extraction in Large Vision Models
['Thomas Fel', 'Ekdeep Singh Lubana', 'Jacob S. Prince', 'Matthew Kowal', 'Victor Boutin', 'Isabel Papadimitriou', 'Binxu Wang', 'Martin Wattenberg', 'Demba Ba', 'Talia Konkle']
2,025
arXiv.org
8
173
['Computer Science']
2,502.129
Soundwave: Less is More for Speech-Text Alignment in LLMs
['Yuhao Zhang', 'Zhiheng Liu', 'Fan Bu', 'Ruiyu Zhang', 'Benyou Wang', 'Haizhou Li']
['cs.CL', 'cs.AI', 'cs.SD']
Existing end-to-end speech large language models (LLMs) usually rely on large-scale annotated data for training, while data-efficient training has not been discussed in depth. We focus on two fundamental problems between speech and text: the representation space gap and sequence length inconsistency. We propose Soundwa...
2025-02-18T14:36:39Z
null
null
null
null
null
null
null
null
null
null
2,502.12982
Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs
['Longxu Dou', 'Qian Liu', 'Fan Zhou', 'Changyu Chen', 'Zili Wang', 'Ziqi Jin', 'Zichen Liu', 'Tongyao Zhu', 'Cunxiao Du', 'Penghui Yang', 'Haonan Wang', 'Jiaheng Liu', 'Yongchi Zhao', 'Xiachong Feng', 'Xin Mao', 'Man Tsung Yeung', 'Kunat Pipatanakul', 'Fajri Koto', 'Min Si Thu', 'Hynek Kydlíček', 'Zeyi Liu', 'Qunshu L...
['cs.CL', 'cs.AI', 'cs.LG']
Sailor2 is a family of cutting-edge multilingual language models for South-East Asian (SEA) languages, available in 1B, 8B, and 20B sizes to suit diverse applications. Building on Qwen2.5, Sailor2 undergoes continuous pre-training on 500B tokens (400B SEA-specific and 100B replay tokens) to support 13 SEA languages whi...
2025-02-18T16:04:57Z
49 pages, 16 figures. Technical Report of Sailor2: https://sea-sailor.github.io/blog/sailor2/
null
null
null
null
null
null
null
null
null
2,502.1299
Artificial Intelligence-derived Vascular Age from Photoplethysmography: A Novel Digital Biomarker for Cardiovascular Health
['Guangkun Nie', 'Qinghao Zhao', 'Gongzheng Tang', 'Yaxin Li', 'Shenda Hong']
['eess.SP']
With the increasing availability of wearable devices, photoplethysmography (PPG) has emerged as a promising non-invasive tool for monitoring human hemodynamics. We propose a deep learning framework to estimate vascular age (AI-vascular age) from PPG signals, incorporating a distribution-aware loss to address biases cau...
2025-02-18T16:12:28Z
null
null
null
null
null
null
null
null
null
null
2,502.13061
Robust Adaptation of Large Multimodal Models for Retrieval Augmented Hateful Meme Detection
['Jingbiao Mei', 'Jinghong Chen', 'Guangyu Yang', 'Weizhe Lin', 'Bill Byrne']
['cs.CL', 'cs.AI', 'cs.CV', 'cs.LG']
Hateful memes have become a significant concern on the Internet, necessitating robust automated detection systems. While LMMs have shown promise in hateful meme detection, they face notable challenges like sub-optimal performance and limited out-of-domain generalization capabilities. Recent studies further reveal the l...
2025-02-18T17:07:29Z
Preprint. Under Review
null
null
Robust Adaptation of Large Multimodal Models for Retrieval Augmented Hateful Meme Detection
['Jingbiao Mei', 'Jinghong Chen', 'Guangyu Yang', 'Weizhe Lin', 'Bill Byrne']
2,025
null
0
51
['Computer Science']
2,502.1313
Magma: A Foundation Model for Multimodal AI Agents
['Jianwei Yang', 'Reuben Tan', 'Qianhui Wu', 'Ruijie Zheng', 'Baolin Peng', 'Yongyuan Liang', 'Yu Gu', 'Mu Cai', 'Seonghyeon Ye', 'Joel Jang', 'Yuquan Deng', 'Lars Liden', 'Jianfeng Gao']
['cs.CV', 'cs.AI', 'cs.HC', 'cs.LG', 'cs.RO']
We present Magma, a foundation model that serves multimodal AI agentic tasks in both the digital and physical worlds. Magma is a significant extension of vision-language (VL) models in that it not only retains the VL understanding ability (verbal intelligence) of the latter, but is also equipped with the ability to pla...
2025-02-18T18:55:21Z
29 pages, 16 figures, technical report from MSR
null
null
null
null
null
null
null
null
null
2,502.13138
AIDE: AI-Driven Exploration in the Space of Code
['Zhengyao Jiang', 'Dominik Schmidt', 'Dhruv Srikanth', 'Dixing Xu', 'Ian Kaplan', 'Deniss Jacenko', 'Yuxiang Wu']
['cs.AI', 'cs.LG']
Machine learning, the foundation of modern artificial intelligence, has driven innovations that have fundamentally transformed the world. Yet, behind advancements lies a complex and often tedious process requiring labor and compute intensive iteration and experimentation. Engineers and scientists developing machine lea...
2025-02-18T18:57:21Z
null
null
null
null
null
null
null
null
null
null
2,502.13143
SoFar: Language-Grounded Orientation Bridges Spatial Reasoning and Object Manipulation
['Zekun Qi', 'Wenyao Zhang', 'Yufei Ding', 'Runpei Dong', 'Xinqiang Yu', 'Jingwen Li', 'Lingyun Xu', 'Baoyu Li', 'Xialin He', 'Guofan Fan', 'Jiazhao Zhang', 'Jiawei He', 'Jiayuan Gu', 'Xin Jin', 'Kaisheng Ma', 'Zhizheng Zhang', 'He Wang', 'Li Yi']
['cs.RO', 'cs.AI', 'cs.CV']
Spatial intelligence is a critical component of embodied AI, promoting robots to understand and interact with their environments. While recent advances have enhanced the ability of VLMs to perceive object locations and positional relationships, they still lack the capability to precisely understand object orientations-...
2025-02-18T18:59:02Z
Project page: https://qizekun.github.io/sofar/
null
null
SoFar: Language-Grounded Orientation Bridges Spatial Reasoning and Object Manipulation
['Zekun Qi', 'Wenyao Zhang', 'Yufei Ding', 'Runpei Dong', 'Xinqiang Yu', 'Jingwen Li', 'Lingyun Xu', 'Baoyu Li', 'Xialin He', 'Guo Fan', 'Jiazhao Zhang', 'Jiawei He', 'Jiayuan Gu', 'Xin Jin', 'Kaisheng Ma', 'Zhizheng Zhang', 'He Wang', 'Li Yi']
2,025
arXiv.org
7
170
['Computer Science']
2,502.13145
Multimodal Mamba: Decoder-only Multimodal State Space Model via Quadratic to Linear Distillation
['Bencheng Liao', 'Hongyuan Tao', 'Qian Zhang', 'Tianheng Cheng', 'Yingyue Li', 'Haoran Yin', 'Wenyu Liu', 'Xinggang Wang']
['cs.CV']
Recent Multimodal Large Language Models (MLLMs) have achieved remarkable performance but face deployment challenges due to their quadratic computational complexity, growing Key-Value cache requirements, and reliance on separate vision encoders. We propose mmMamba, a framework for developing linear-complexity native mul...
2025-02-18T18:59:57Z
Code and model are available at https://github.com/hustvl/mmMamba
null
null
null
null
null
null
null
null
null
2,502.13167
SmartLLM: Smart Contract Auditing using Custom Generative AI
['Jun Kevin', 'Pujianto Yugopuspito']
['cs.CR', 'cs.AI']
Smart contracts are essential to decentralized finance (DeFi) and blockchain ecosystems but are increasingly vulnerable to exploits due to coding errors and complex attack vectors. Traditional static analysis tools and existing vulnerability detection methods often fail to address these challenges comprehensively, lead...
2025-02-17T06:22:05Z
null
null
null
null
null
null
null
null
null
null
2,502.13252
Multilingual Language Model Pretraining using Machine-translated Data
['Jiayi Wang', 'Yao Lu', 'Maurice Weber', 'Max Ryabinin', 'David Adelani', 'Yihong Chen', 'Raphael Tang', 'Pontus Stenetorp']
['cs.CL']
High-resource languages such as English, enables the pretraining of high-quality large language models (LLMs). The same can not be said for most other languages as LLMs still underperform for non-English languages, likely due to a gap in the quality and diversity of the available multilingual pretraining corpora. In th...
2025-02-18T19:27:53Z
null
null
null
null
null
null
null
null
null
null
2,502.13398
GeLLMO: Generalizing Large Language Models for Multi-property Molecule Optimization
['Vishal Dey', 'Xiao Hu', 'Xia Ning']
['cs.LG', 'cs.AI', 'cs.CL', 'physics.chem-ph', 'q-bio.QM']
Despite recent advancements, most computational methods for molecule optimization are constrained to single- or double-property optimization tasks and suffer from poor scalability and generalizability to novel optimization tasks. Meanwhile, Large Language Models (LLMs) demonstrate remarkable out-of-domain generalizabil...
2025-02-19T03:14:11Z
Accepted to ACL Main 2025. Vishal Dey and Xiao Hu contributed equally to this paper
null
null
GeLLM3O: Generalizing Large Language Models for Multi-property Molecule Optimization
['Vishal Dey', 'Xiao Hu', 'Xia Ning']
2,025
arXiv.org
4
69
['Computer Science', 'Physics', 'Biology']
2,502.13449
Mol-LLaMA: Towards General Understanding of Molecules in Large Molecular Language Model
['Dongki Kim', 'Wonbin Lee', 'Sung Ju Hwang']
['cs.LG', 'physics.chem-ph']
Understanding molecules is key to understanding organisms and driving advances in drug discovery, requiring interdisciplinary knowledge across chemistry and biology. Although large molecular language models have achieved notable success in task transfer, they often struggle to accurately analyze molecular features due ...
2025-02-19T05:49:10Z
Project Page: https://mol-llama.github.io/
null
null
null
null
null
null
null
null
null
2,502.13458
ThinkGuard: Deliberative Slow Thinking Leads to Cautious Guardrails
['Xiaofei Wen', 'Wenxuan Zhou', 'Wenjie Jacky Mo', 'Muhao Chen']
['cs.CL', 'cs.AI', 'cs.CR', 'cs.LG']
Ensuring the safety of large language models (LLMs) is critical as they are deployed in real-world applications. Existing guardrails rely on rule-based filtering or single-pass classification, limiting their ability to handle nuanced safety violations. To address this, we propose ThinkGuard, a critique-augmented guardr...
2025-02-19T06:09:58Z
ACL 2025
null
null
ThinkGuard: Deliberative Slow Thinking Leads to Cautious Guardrails
['Xiaofei Wen', 'Wenxuan Zhou', 'W. Mo', 'Muhao Chen']
2,025
arXiv.org
7
42
['Computer Science']
2,502.13502
PLDR-LLMs Learn A Generalizable Tensor Operator That Can Replace Its Own Deep Neural Net At Inference
['Burc Gokden']
['cs.CL', 'cs.AI', 'cs.LG']
We show that Large Language Model from Power Law Decoder Representations (PLDR-LLM) is a foundational model whose deductive outputs are invariant tensors up to a small perturbation. PLDR-LLM learns a singularity condition for the deductive outputs that enable the once-inferred energy-curvature tensor $\mathbf{G}_{LM}$ ...
2025-02-19T07:43:36Z
15 pages, 1 figure, 12 tables, more ablation data included
null
null
null
null
null
null
null
null
null
2,502.1352
A Large and Balanced Corpus for Fine-grained Arabic Readability Assessment
['Khalid N. Elmadani', 'Nizar Habash', 'Hanada Taha-Thomure']
['cs.CL']
This paper introduces the Balanced Arabic Readability Evaluation Corpus (BAREC), a large-scale, fine-grained dataset for Arabic readability assessment. BAREC consists of 69,441 sentences spanning 1+ million words, carefully curated to cover 19 readability levels, from kindergarten to postgraduate comprehension. The cor...
2025-02-19T08:16:11Z
Accepted at ACL 2025 Findings
null
null
null
null
null
null
null
null
null
2,502.13603
Efficient Safety Retrofitting Against Jailbreaking for LLMs
['Dario Garcia-Gasulla', 'Adrian Tormos', 'Anna Arias-Duart', 'Daniel Hinjos', 'Oscar Molina-Sedano', 'Ashwin Kumar Gururajan', 'Maria Eugenia Cardello']
['cs.CL', 'cs.AI', 'cs.LG']
Direct Preference Optimization (DPO) is an efficient alignment technique that steers LLMs towards preferable outputs by training on preference data, bypassing the need for explicit reward models. Its simplicity enables easy adaptation to various domains and safety requirements. This paper examines DPO's effectiveness i...
2025-02-19T10:33:18Z
null
null
null
Efficient Safety Retrofitting Against Jailbreaking for LLMs
['Dario Garcia-Gasulla', 'Adrián Tormos', 'Anna Arias-Duart', 'Daniel Hinjos', 'Oscar Molina-Sedano', 'Ashwin Kumar Gururajan', 'Maria Eugenia Cardello']
2,025
arXiv.org
0
61
['Computer Science']
2,502.13656
Refining Sentence Embedding Model through Ranking Sentences Generation with Large Language Models
['Liyang He', 'Chenglong Liu', 'Rui Li', 'Zhenya Huang', 'Shulan Ruan', 'Jun Zhou', 'Enhong Chen']
['cs.CL']
Sentence embedding is essential for many NLP tasks, with contrastive learning methods achieving strong performance using annotated datasets like NLI. Yet, the reliance on manual labels limits scalability. Recent studies leverage large language models (LLMs) to generate sentence pairs, reducing annotation dependency. Ho...
2025-02-19T12:07:53Z
null
null
null
Refining Sentence Embedding Model through Ranking Sentences Generation with Large Language Models
['Liyang He', 'Chenglong Liu', 'Rui Li', 'Zhenya Huang', 'Shulan Ruan', 'Jun Zhou', 'Enhong Chen']
2,025
arXiv.org
1
65
['Computer Science']
2,502.13685
MoM: Linear Sequence Modeling with Mixture-of-Memories
['Jusen Du', 'Weigao Sun', 'Disen Lan', 'Jiaxi Hu', 'Yu Cheng']
['cs.CL', 'cs.AI', 'cs.LG']
Linear sequence modeling methods, such as linear attention, state space modeling, and linear RNNs, offer significant efficiency improvements by reducing the complexity of training and inference. However, these methods typically compress the entire input sequence into a single fixed-size memory state, which leads to sub...
2025-02-19T12:53:55Z
Technical report, 16 pages
null
null
null
null
null
null
null
null
null
2,502.13759
Geolocation with Real Human Gameplay Data: A Large-Scale Dataset and Human-Like Reasoning Framework
['Zirui Song', 'Jingpu Yang', 'Yuan Huang', 'Jonathan Tonglet', 'Zeyu Zhang', 'Tao Cheng', 'Meng Fang', 'Iryna Gurevych', 'Xiuying Chen']
['cs.CV']
Geolocation, the task of identifying an image's location, requires complex reasoning and is crucial for navigation, monitoring, and cultural preservation. However, current methods often produce coarse, imprecise, and non-interpretable localization. A major challenge lies in the quality and scale of existing geolocation...
2025-02-19T14:21:25Z
Update new version
null
null
Geolocation with Real Human Gameplay Data: A Large-Scale Dataset and Human-Like Reasoning Framework
['Zirui Song', 'Jingpu Yang', 'Yuan Huang', 'Jonathan Tonglet', 'Zeyu Zhang', 'Tao Cheng', 'Meng Fang', 'Iryna Gurevych', 'Xiuying Chen']
2,025
arXiv.org
5
64
['Computer Science']
2,502.13785
Helix-mRNA: A Hybrid Foundation Model For Full Sequence mRNA Therapeutics
['Matthew Wood', 'Mathieu Klop', 'Maxime Allard']
['q-bio.GN', 'cs.AI']
mRNA-based vaccines have become a major focus in the pharmaceutical industry. The coding sequence as well as the Untranslated Regions (UTRs) of an mRNA can strongly influence translation efficiency, stability, degradation, and other factors that collectively determine a vaccine's effectiveness. However, optimizing mRNA...
2025-02-19T14:51:41Z
8 pages, 3 figures, 3 tables
null
null
null
null
null
null
null
null
null
2,502.13898
GroundCap: A Visually Grounded Image Captioning Dataset
['Daniel A. P. Oliveira', 'Lourenço Teodoro', 'David Martins de Matos']
['cs.CV', 'cs.CL', 'I.2.10; I.2.7']
Current image captioning systems lack the ability to link descriptive text to specific visual elements, making their outputs difficult to verify. While recent approaches offer some grounding capabilities, they cannot track object identities across multiple references or ground both actions and objects simultaneously. W...
2025-02-19T17:31:59Z
37 pages
null
null
null
null
null
null
null
null
null
2,502.13917
TESS 2: A Large-Scale Generalist Diffusion Language Model
['Jaesung Tae', 'Hamish Ivison', 'Sachin Kumar', 'Arman Cohan']
['cs.CL']
We introduce TESS 2, a general instruction-following diffusion language model that outperforms contemporary instruction-tuned diffusion models, as well as matches and sometimes exceeds strong autoregressive (AR) models. We train TESS 2 by first adapting a strong AR model via continued pretraining with the usual cross-e...
2025-02-19T17:50:31Z
ACL 2025 camera-ready
null
null
TESS 2: A Large-Scale Generalist Diffusion Language Model
['Jaesung Tae', 'Hamish Ivison', 'Sachin Kumar', 'Arman Cohan']
2,025
arXiv.org
1
60
['Computer Science']
2,502.13922
LongPO: Long Context Self-Evolution of Large Language Models through Short-to-Long Preference Optimization
['Guanzheng Chen', 'Xin Li', 'Michael Qizhe Shieh', 'Lidong Bing']
['cs.CL', 'cs.LG']
Large Language Models (LLMs) have demonstrated remarkable capabilities through pretraining and alignment. However, superior short-context LLMs may underperform in long-context scenarios due to insufficient long-context alignment. This alignment process remains challenging due to the impracticality of human annotation f...
2025-02-19T17:59:03Z
ICLR 2025
null
null
null
null
null
null
null
null
null
2,502.13923
Qwen2.5-VL Technical Report
['Shuai Bai', 'Keqin Chen', 'Xuejing Liu', 'Jialin Wang', 'Wenbin Ge', 'Sibo Song', 'Kai Dang', 'Peng Wang', 'Shijie Wang', 'Jun Tang', 'Humen Zhong', 'Yuanzhi Zhu', 'Mingkun Yang', 'Zhaohai Li', 'Jianqiang Wan', 'Pengfei Wang', 'Wei Ding', 'Zheren Fu', 'Yiheng Xu', 'Jiabo Ye', 'Xi Zhang', 'Tianbao Xie', 'Zesen Cheng',...
['cs.CV', 'cs.CL']
We introduce Qwen2.5-VL, the latest flagship model of Qwen vision-language series, which demonstrates significant advancements in both foundational capabilities and innovative functionalities. Qwen2.5-VL achieves a major leap forward in understanding and interacting with the world through enhanced visual recognition, p...
2025-02-19T18:00:14Z
null
null
null
null
null
null
null
null
null
null
2,502.13967
FlexTok: Resampling Images into 1D Token Sequences of Flexible Length
['Roman Bachmann', 'Jesse Allardice', 'David Mizrahi', 'Enrico Fini', 'Oğuzhan Fatih Kar', 'Elmira Amirloo', 'Alaaeldin El-Nouby', 'Amir Zamir', 'Afshin Dehghan']
['cs.CV', 'cs.LG']
Image tokenization has enabled major advances in autoregressive image generation by providing compressed, discrete representations that are more efficient to process than raw pixels. While traditional approaches use 2D grid tokenization, recent methods like TiTok have shown that 1D tokenization can achieve high generat...
2025-02-19T18:59:44Z
ICML 2025. Project page at https://flextok.epfl.ch/
null
null
null
null
null
null
null
null
null
2,502.1399
Remote Sensing Semantic Segmentation Quality Assessment based on Vision Language Model
['Huiying Shi', 'Zhihong Tan', 'Zhihan Zhang', 'Hongchen Wei', 'Yaosi Hu', 'Yingxue Zhang', 'Zhenzhong Chen']
['eess.IV', 'cs.LG']
The complexity of scenes and variations in image quality result in significant variability in the performance of semantic segmentation methods of remote sensing imagery (RSI) in supervised real-world scenarios. This makes the evaluation of semantic segmentation quality in such scenarios an issue to be resolved. However...
2025-02-19T02:28:12Z
16 pages,6 figures
null
null
null
null
null
null
null
null
null
2,502.13991
Learning to Discover Regulatory Elements for Gene Expression Prediction
['Xingyu Su', 'Haiyang Yu', 'Degui Zhi', 'Shuiwang Ji']
['q-bio.GN', 'cs.AI']
We consider the problem of predicting gene expressions from DNA sequences. A key challenge of this task is to find the regulatory elements that control gene expressions. Here, we introduce Seq2Exp, a Sequence to Expression network explicitly designed to discover and extract regulatory elements that drive target gene ex...
2025-02-19T03:25:49Z
null
null
null
null
null
null
null
null
null
null
2,502.13995
FantasyID: Face Knowledge Enhanced ID-Preserving Video Generation
['Yunpeng Zhang', 'Qiang Wang', 'Fan Jiang', 'Yaqi Fan', 'Mu Xu', 'Yonggang Qi']
['cs.GR', 'cs.CV']
Tuning-free approaches adapting large-scale pre-trained video diffusion models for identity-preserving text-to-video generation (IPT2V) have gained popularity recently due to their efficacy and scalability. However, significant challenges remain to achieve satisfied facial dynamics while keeping the identity unchanged....
2025-02-19T06:50:27Z
null
null
null
FantasyID: Face Knowledge Enhanced ID-Preserving Video Generation
['Yunpeng Zhang', 'Qiang Wang', 'Fan Jiang', 'Yaqi Fan', 'Mu Xu', 'Yonggang Qi']
2,025
arXiv.org
4
0
['Computer Science']
2,502.14044
Enhancing Cognition and Explainability of Multimodal Foundation Models with Self-Synthesized Data
['Yucheng Shi', 'Quanzheng Li', 'Jin Sun', 'Xiang Li', 'Ninghao Liu']
['cs.CV', 'cs.LG']
Large Multimodal Models (LMMs), or Vision-Language Models (VLMs), have shown impressive capabilities in a wide range of visual tasks. However, they often struggle with fine-grained visual reasoning, failing to identify domain-specific objectives and provide justifiable explanations for their predictions. To address the...
2025-02-19T19:05:45Z
Accepted by ICLR 2025. Code: https://github.com/sycny/SelfSynthX
null
null
Enhancing Cognition and Explainability of Multimodal Foundation Models with Self-Synthesized Data
['Yucheng Shi', 'Quanzheng Li', 'Jin Sun', 'Xiang Li', 'Ninghao Liu']
2,025
International Conference on Learning Representations
3
55
['Computer Science']
2,502.14301
SEA-HELM: Southeast Asian Holistic Evaluation of Language Models
['Yosephine Susanto', 'Adithya Venkatadri Hulagadri', 'Jann Railey Montalan', 'Jian Gang Ngui', 'Xian Bin Yong', 'Weiqi Leong', 'Hamsawardhini Rengarajan', 'Peerat Limkonchotiwat', 'Yifan Mai', 'William Chandra Tjhi']
['cs.CL', 'cs.AI']
With the rapid emergence of novel capabilities in Large Language Models (LLMs), the need for rigorous multilingual and multicultural benchmarks that are integrated has become more pronounced. Though existing LLM benchmarks are capable of evaluating specific capabilities of LLMs in English as well as in various mid- to ...
2025-02-20T06:32:45Z
null
null
null
SEA-HELM: Southeast Asian Holistic Evaluation of Language Models
['Yosephine Susanto', 'Adithya Venkatadri Hulagadri', 'J. Montalan', 'Jian Gang Ngui', 'Xian Bin Yong', 'Weiqi Leong', 'Hamsawardhini Rengarajan', 'Peerat Limkonchotiwat', 'Yifan Mai', 'William-Chandra Tjhi']
2,025
arXiv.org
0
106
['Computer Science']
2,502.14377
RelaCtrl: Relevance-Guided Efficient Control for Diffusion Transformers
['Ke Cao', 'Jing Wang', 'Ao Ma', 'Jiasong Feng', 'Zhanjie Zhang', 'Xuanhua He', 'Shanyuan Liu', 'Bo Cheng', 'Dawei Leng', 'Yuhui Yin', 'Jie Zhang']
['cs.CV']
The Diffusion Transformer plays a pivotal role in advancing text-to-image and text-to-video generation, owing primarily to its inherent scalability. However, existing controlled diffusion transformer methods incur significant parameter and computational overheads and suffer from inefficient resource allocation due to t...
2025-02-20T09:10:05Z
Homepage: https://360cvgroup.github.io/RelaCtrl/ Github: https://github.com/360CVGroup/RelaCtrl
null
null
RelaCtrl: Relevance-Guided Efficient Control for Diffusion Transformers
['Ke Cao', 'Jing Wang', 'Ao Ma', 'Jiasong Feng', 'Zhanjie Zhang', 'Xuanhua He', 'Shanyuan Liu', 'Bo Cheng', 'Dawei Leng', 'Yuhui Yin', 'Jie Zhang']
2,025
arXiv.org
4
37
['Computer Science']
2,502.14429
Early-Exit and Instant Confidence Translation Quality Estimation
['Vilém Zouhar', 'Maike Züfle', 'Beni Egressy', 'Julius Cheng', 'Mrinmaya Sachan', 'Jan Niehues']
['cs.CL']
Quality estimation is omnipresent in machine translation, for both evaluation and generation. Unfortunately, quality estimation models are often opaque and computationally expensive, making them impractical to be part of large-scale pipelines. In this work, we tackle two connected challenges: (1) reducing the cost of q...
2025-02-20T10:27:13Z
null
null
null
Early-Exit and Instant Confidence Translation Quality Estimation
['Vilém Zouhar', 'Maike Zufle', 'Béni Egressy', 'Julius Cheng', 'Jan Niehues']
2,025
arXiv.org
1
0
['Computer Science']
2,502.14458
Llamba: Scaling Distilled Recurrent Models for Efficient Language Processing
['Aviv Bick', 'Tobias Katsch', 'Nimit Sohoni', 'Arjun Desai', 'Albert Gu']
['cs.LG', 'cs.AI']
We introduce Llamba, a family of efficient recurrent language models distilled from Llama-3.x into the Mamba architecture. The series includes Llamba-1B, Llamba-3B, and Llamba-8B, which achieve higher inference throughput and handle significantly larger batch sizes than Transformer-based models while maintaining compar...
2025-02-20T11:18:39Z
null
null
null
null
null
null
null
null
null
null
2,502.14502
How Much Knowledge Can You Pack into a LoRA Adapter without Harming LLM?
['Sergey Pletenev', 'Maria Marina', 'Daniil Moskovskiy', 'Vasily Konovalov', 'Pavel Braslavski', 'Alexander Panchenko', 'Mikhail Salnikov']
['cs.CL']
The performance of Large Language Models (LLMs) on many tasks is greatly limited by the knowledge learned during pre-training and stored in the model's parameters. Low-rank adaptation (LoRA) is a popular and efficient training technique for updating or domain-specific adaptation of LLMs. In this study, we investigate h...
2025-02-20T12:31:03Z
null
null
null
How Much Knowledge Can You Pack into a LoRA Adapter without Harming LLM?
['Sergey Pletenev', 'Maria Marina', 'Daniil Moskovskiy', 'Vasily Konovalov', 'Pavel Braslavski', 'Alexander Panchenko', 'M. Salnikov']
2,025
North American Chapter of the Association for Computational Linguistics
1
34
['Computer Science']
2,502.14561
Can LLMs Predict Citation Intent? An Experimental Analysis of In-context Learning and Fine-tuning on Open LLMs
['Paris Koloveas', 'Serafeim Chatzopoulos', 'Thanasis Vergoulis', 'Christos Tryfonopoulos']
['cs.CL', 'cs.DL']
This work investigates the ability of open Large Language Models (LLMs) to predict citation intent through in-context learning and fine-tuning. Unlike traditional approaches relying on domain-specific pre-trained models like SciBERT, we demonstrate that general-purpose LLMs can be adapted to this task with minimal task...
2025-02-20T13:45:42Z
null
null
null
null
null
null
null
null
null
null
2,502.14637
ReQFlow: Rectified Quaternion Flow for Efficient and High-Quality Protein Backbone Generation
['Angxiao Yue', 'Zichong Wang', 'Hongteng Xu']
['cs.LG', 'cs.AI']
Protein backbone generation plays a central role in de novo protein design and is significant for many biological and medical applications. Although diffusion and flow-based generative models provide potential solutions to this challenging task, they often generate proteins with undesired designability and suffer compu...
2025-02-20T15:20:37Z
null
null
null
null
null
null
null
null
null
null
2,502.14638
NAVIG: Natural Language-guided Analysis with Vision Language Models for Image Geo-localization
['Zheyuan Zhang', 'Runze Li', 'Tasnim Kabir', 'Jordan Boyd-Graber']
['cs.CL', 'cs.CV']
Image geo-localization is the task of predicting the specific location of an image and requires complex reasoning across visual, geographical, and cultural contexts. While prior Vision Language Models (VLMs) have the best accuracy at this task, there is a dearth of high-quality datasets and models for analytical reason...
2025-02-20T15:21:35Z
null
null
null
null
null
null
null
null
null
null
2,502.14669
AlphaMaze: Enhancing Large Language Models' Spatial Intelligence via GRPO
['Alan Dao', 'Dinh Bach Vu']
['cs.CL']
Large Language Models (LLMs) have demonstrated impressive capabilities in language processing, yet they often struggle with tasks requiring genuine visual spatial reasoning. In this paper, we introduce a novel two-stage training framework designed to equip standard LLMs with visual reasoning abilities for maze navigati...
2025-02-20T16:05:18Z
null
null
null
AlphaMaze: Enhancing Large Language Models' Spatial Intelligence via GRPO
['Alan Dao', 'Dinh Bach Vu']
2,025
arXiv.org
4
20
['Computer Science']
2,502.14673
ChunkFormer: Masked Chunking Conformer For Long-Form Speech Transcription
['Khanh Le', 'Tuan Vu Ho', 'Dung Tran', 'Duc Thanh Chau']
['cs.SD', 'eess.AS']
Deploying ASR models at an industrial scale poses significant challenges in hardware resource management, especially for long-form transcription tasks where audio may last for hours. Large Conformer models, despite their capabilities, are limited to processing only 15 minutes of audio on an 80GB GPU. Furthermore, varia...
2025-02-20T16:06:06Z
Accepted to ICASSP 2025
null
null
null
null
null
null
null
null
null
2,502.14706
Building reliable sim driving agents by scaling self-play
['Daphne Cornelisse', 'Aarav Pandya', 'Kevin Joseph', 'Joseph Suárez', 'Eugene Vinitsky']
['cs.AI', 'cs.RO']
Simulation agents are essential for designing and testing systems that interact with humans, such as autonomous vehicles (AVs). These agents serve various purposes, from benchmarking AV performance to stress-testing system limits, but all applications share one key requirement: reliability. To enable sound experimentat...
2025-02-20T16:30:45Z
v3
null
null
Building reliable sim driving agents by scaling self-play
['Daphne Cornelisse', 'Aarav Pandya', 'Kevin Joseph', "Joseph Su'arez", 'Eugene Vinitsky']
2,025
arXiv.org
1
42
['Computer Science']
2,502.14753
MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders
['Maya Varma', 'Ashwin Kumar', 'Rogier van der Sluijs', 'Sophie Ostmeier', 'Louis Blankemeier', 'Pierre Chambon', 'Christian Bluethgen', 'Jip Prince', 'Curtis Langlotz', 'Akshay Chaudhari']
['eess.IV', 'cs.AI', 'cs.CV']
Medical images are acquired at high resolutions with large fields of view in order to capture fine-grained features necessary for clinical decision-making. Consequently, training deep learning models on medical images can incur large computational costs. In this work, we address the challenge of downsizing medical imag...
2025-02-20T17:24:06Z
MIDL 2025 (Oral)
null
null
null
null
null
null
null
null
null
2,502.14786
SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
['Michael Tschannen', 'Alexey Gritsenko', 'Xiao Wang', 'Muhammad Ferjad Naeem', 'Ibrahim Alabdulmohsin', 'Nikhil Parthasarathy', 'Talfan Evans', 'Lucas Beyer', 'Ye Xia', 'Basil Mustafa', 'Olivier Hénaff', 'Jeremiah Harmsen', 'Andreas Steiner', 'Xiaohua Zhai']
['cs.CV', 'cs.AI']
We introduce SigLIP 2, a family of new multilingual vision-language encoders that build on the success of the original SigLIP. In this second iteration, we extend the original image-text training objective with several prior, independently developed techniques into a unified recipe -- this includes captioning-based pre...
2025-02-20T18:08:29Z
Model checkpoints are available at https://github.com/google-research/big_vision/tree/main/big_vision/configs/proj/image_text/README_siglip2.md
null
null
SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
['Michael Tschannen', 'Alexey Gritsenko', 'Xiao Wang', 'M. Naeem', 'Ibrahim M. Alabdulmohsin', 'Nikhil Parthasarathy', 'Talfan Evans', 'Lucas Beyer', 'Ye Xia', 'Basil Mustafa', "Olivier H'enaff", 'Jeremiah Harmsen', 'A. Steiner', 'Xiao-Qi Zhai']
2,025
arXiv.org
80
0
['Computer Science']
2,502.1483
Middle-Layer Representation Alignment for Cross-Lingual Transfer in Fine-Tuned LLMs
['Danni Liu', 'Jan Niehues']
['cs.CL', 'cs.AI']
While large language models demonstrate remarkable capabilities at task-specific applications through fine-tuning, extending these benefits across diverse languages is essential for broad accessibility. However, effective cross-lingual transfer is hindered by LLM performance gaps across languages and the scarcity of fi...
2025-02-20T18:45:43Z
ACL 2025
null
null
null
null
null
null
null
null
null
2,502.14837
Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs
['Tao Ji', 'Bin Guo', 'Yuanbin Wu', 'Qipeng Guo', 'Lixing Shen', 'Zhan Chen', 'Xipeng Qiu', 'Qi Zhang', 'Tao Gui']
['cs.CL', 'cs.AI']
Multi-head Latent Attention (MLA) is an innovative architecture proposed by DeepSeek, designed to ensure efficient and economical inference by significantly compressing the Key-Value (KV) cache into a latent vector. Compared to MLA, standard LLMs employing Multi-Head Attention (MHA) and its variants such as Grouped-Que...
2025-02-20T18:50:42Z
16 pages, 8 figures
null
null
null
null
null
null
null
null
null
2,502.14854
CLIPPER: Compression enables long-context synthetic data generation
['Chau Minh Pham', 'Yapei Chang', 'Mohit Iyyer']
['cs.CL']
LLM developers are increasingly reliant on synthetic data, but generating high-quality data for complex long-context reasoning tasks remains challenging. We introduce CLIPPER, a compression-based approach for generating synthetic data tailored to narrative claim verification - a task that requires reasoning over a book...
2025-02-20T18:58:03Z
null
null
null
CLIPPER: Compression enables long-context synthetic data generation
['Chau Minh Pham', 'Yapei Chang', 'Mohit Iyyer']
2,025
arXiv.org
1
59
['Computer Science']
2,502.14855
Prompt-to-Leaderboard
['Evan Frick', 'Connor Chen', 'Joseph Tennyson', 'Tianle Li', 'Wei-Lin Chiang', 'Anastasios N. Angelopoulos', 'Ion Stoica']
['cs.LG', 'cs.CL']
Large language model (LLM) evaluations typically rely on aggregated metrics like accuracy or human preference, averaging across users and prompts. This averaging obscures user- and prompt-specific variations in model performance. To address this, we propose Prompt-to-Leaderboard (P2L), a method that produces leaderboar...
2025-02-20T18:58:07Z
null
null
null
Prompt-to-Leaderboard
['Evan Frick', 'Connor Chen', 'Joseph Tennyson', 'Tianle Li', 'Wei-Lin Chiang', 'Anastasios N. Angelopoulos', 'Ion Stoica']
2,025
arXiv.org
7
42
['Computer Science']
2,502.14856
FR-Spec: Accelerating Large-Vocabulary Language Models via Frequency-Ranked Speculative Sampling
['Weilin Zhao', 'Tengyu Pan', 'Xu Han', 'Yudi Zhang', 'Ao Sun', 'Yuxiang Huang', 'Kaihuo Zhang', 'Weilun Zhao', 'Yuxuan Li', 'Jianyong Wang', 'Zhiyuan Liu', 'Maosong Sun']
['cs.CL', 'cs.AI', 'cs.LG']
Speculative sampling has emerged as an important technique for accelerating the auto-regressive generation process of large language models (LLMs) by utilizing a draft-then-verify mechanism to produce multiple tokens per forward pass. While state-of-the-art speculative sampling methods use only a single layer and a lan...
2025-02-20T18:58:10Z
null
null
null
FR-Spec: Accelerating Large-Vocabulary Language Models via Frequency-Ranked Speculative Sampling
['Weilin Zhao', 'Tengyu Pan', 'Xu Han', 'Yudi Zhang', 'Ao Sun', 'Yuxiang Huang', 'Kaihuo Zhang', 'Weilun Zhao', 'Yuxuan Li', 'Jianyong Wang', 'Zhiyuan Liu', 'Maosong Sun']
2,025
arXiv.org
2
49
['Computer Science']
2,502.14907
GneissWeb: Preparing High Quality Data for LLMs at Scale
['Hajar Emami Gohari', 'Swanand Ravindra Kadhe', 'Syed Yousaf Shah. Constantin Adam', 'Abdulhamid Adebayo', 'Praneet Adusumilli', 'Farhan Ahmed', 'Nathalie Baracaldo Angel', 'Santosh Borse', 'Yuan-Chi Chang', 'Xuan-Hong Dang', 'Nirmit Desai', 'Ravital Eres', 'Ran Iwamoto', 'Alexei Karve', 'Yan Koyfman', 'Wei-Han Lee', ...
['cs.CL', 'cs.AI']
Data quantity and quality play a vital role in determining the performance of Large Language Models (LLMs). High-quality data, in particular, can significantly boost the LLM's ability to generalize on a wide range of downstream tasks. Large pre-training datasets for leading LLMs remain inaccessible to the public, where...
2025-02-19T00:14:29Z
null
null
null
null
null
null
null
null
null
null
2,502.15011
CrossOver: 3D Scene Cross-Modal Alignment
['Sayan Deb Sarkar', 'Ondrej Miksik', 'Marc Pollefeys', 'Daniel Barath', 'Iro Armeni']
['cs.CV']
Multi-modal 3D object understanding has gained significant attention, yet current approaches often assume complete data availability and rigid alignment across all modalities. We present CrossOver, a novel framework for cross-modal 3D scene understanding via flexible, scene-level modality alignment. Unlike traditional ...
2025-02-20T20:05:30Z
Project Page: https://sayands.github.io/crossover/
null
null
CrossOver: 3D Scene Cross-Modal Alignment
['Sayan Deb Sarkar', 'O. Mikšík', 'Marc Pollefeys', 'Daniel Barath', 'Iro Armeni']
2,025
Computer Vision and Pattern Recognition
2
45
['Computer Science']
2,502.15167
M3-AGIQA: Multimodal, Multi-Round, Multi-Aspect AI-Generated Image Quality Assessment
['Chuan Cui', 'Kejiang Chen', 'Zhihua Wei', 'Wen Shen', 'Weiming Zhang', 'Nenghai Yu']
['cs.CV']
The rapid advancement of AI-generated image (AIGI) models presents new challenges for evaluating image quality, particularly across three aspects: perceptual quality, prompt correspondence, and authenticity. To address these challenges, we introduce M3-AGIQA, a comprehensive framework that leverages Multimodal Large La...
2025-02-21T03:05:45Z
24 pages. This work has been submitted to the ACM for possible publication
null
null
M3-AGIQA: Multimodal, Multi-Round, Multi-Aspect AI-Generated Image Quality Assessment
['Chuan Cui', 'Kejiang Chen', 'Zhihua Wei', 'Wen Shen', 'Weiming Zhang', 'Neng H. Yu']
2,025
arXiv.org
0
63
['Computer Science']
2,502.15168
mStyleDistance: Multilingual Style Embeddings and their Evaluation
['Justin Qiu', 'Jiacheng Zhu', 'Ajay Patel', 'Marianna Apidianaki', 'Chris Callison-Burch']
['cs.CL']
Style embeddings are useful for stylistic analysis and style transfer; however, only English style embeddings have been made available. We introduce Multilingual StyleDistance (mStyleDistance), a multilingual style embedding model trained using synthetic data and contrastive learning. We train the model on data from ni...
2025-02-21T03:11:41Z
arXiv admin note: substantial text overlap with arXiv:2410.12757
null
null
null
null
null
null
null
null
null
2,502.15392
Chitrarth: Bridging Vision and Language for a Billion People
['Shaharukh Khan', 'Ayush Tarun', 'Abhinav Ravi', 'Ali Faraz', 'Akshat Patidar', 'Praveen Kumar Pokala', 'Anagha Bhangare', 'Raja Kolla', 'Chandra Khatri', 'Shubham Agarwal']
['cs.AI', 'cs.CL', 'cs.CV']
Recent multimodal foundation models are primarily trained on English or high resource European language data, which hinders their applicability to other medium and low-resource languages. To address this limitation, we introduce Chitrarth (Chitra: Image; Artha: Meaning), an inclusive Vision-Language Model (VLM), specif...
2025-02-21T11:38:40Z
null
null
null
Chitrarth: Bridging Vision and Language for a Billion People
['Shaharukh Khan', 'Ayush K Tarun', 'Abhinav Ravi', 'Ali Faraz', 'Akshat Patidar', 'Praveen Pokala', 'Anagha Bhangare', 'Raja Kolla', 'Chandra Khatri', 'Shubham Agarwal']
2,025
IEEE International Conference on Acoustics, Speech, and Signal Processing
1
54
['Computer Science']
2,502.15429
Pub-Guard-LLM: Detecting Fraudulent Biomedical Articles with Reliable Explanations
['Lihu Chen', 'Shuojie Fu', 'Gabriel Freedman', 'Cemre Zor', 'Guy Martin', 'James Kinross', 'Uddhav Vaghela', 'Ovidiu Serban', 'Francesca Toni']
['cs.CL']
A significant and growing number of published scientific articles is found to involve fraudulent practices, posing a serious threat to the credibility and safety of research in fields such as medicine. We propose Pub-Guard-LLM, the first large language model-based system tailored to fraud detection of biomedical scient...
2025-02-21T12:54:56Z
long paper under review
null
null
null
null
null
null
null
null
null
2,502.15543
ParamMute: Suppressing Knowledge-Critical FFNs for Faithful Retrieval-Augmented Generation
['Pengcheng Huang', 'Zhenghao Liu', 'Yukun Yan', 'Haiyan Zhao', 'Xiaoyuan Yi', 'Hao Chen', 'Zhiyuan Liu', 'Maosong Sun', 'Tong Xiao', 'Ge Yu', 'Chenyan Xiong']
['cs.CL', 'cs.AI']
Large language models (LLMs) integrated with retrieval-augmented generation (RAG) have improved factuality by grounding outputs in external evidence. However, they remain susceptible to unfaithful generation, where outputs contradict retrieved context despite its relevance and accuracy. Existing approaches aiming to im...
2025-02-21T15:50:41Z
22 pages, 7 figures, 7 tables
null
null
ParamMute: Suppressing Knowledge-Critical FFNs for Faithful Retrieval-Augmented Generation
['Pengcheng Huang', 'Zhenghao Liu', 'Yukun Yan', 'Xiaoyuan Yi', 'Hao Chen', 'Zhiyuan Liu', 'Maosong Sun', 'Tong Xiao', 'Ge Yu', 'Chenyan Xiong']
2,025
null
2
62
['Computer Science']
2,502.15589
LightThinker: Thinking Step-by-Step Compression
['Jintian Zhang', 'Yuqi Zhu', 'Mengshu Sun', 'Yujie Luo', 'Shuofei Qiao', 'Lun Du', 'Da Zheng', 'Huajun Chen', 'Ningyu Zhang']
['cs.CL', 'cs.AI', 'cs.IR', 'cs.LG', 'cs.MM']
Large language models (LLMs) have shown remarkable performance in complex reasoning tasks, but their efficiency is hindered by the substantial memory and computational costs associated with generating lengthy tokens. In this paper, we propose LightThinker, a novel method that enables LLMs to dynamically compress interm...
2025-02-21T16:57:22Z
null
null
null
LightThinker: Thinking Step-by-Step Compression
['Jintian Zhang', 'Yuqi Zhu', 'Mengshu Sun', 'Yujie Luo', 'Shuofei Qiao', 'Lun Du', 'Da Zheng', 'Huajun Chen', 'Ningyu Zhang']
2,025
arXiv.org
34
47
['Computer Science']
2,502.1561
A general language model for peptide identification
['Jixiu Zhai', 'Tianchi Lu', 'Haitian Zhong', 'Ziyang Xu', 'Yuhuan Liu', 'Shengrui Xu', 'Jingwan Wang', 'Dan Huang']
['cs.LG', 'cs.AI', '92C40, 68T07', 'I.2.6; J.3']
Accurate identification of bioactive peptides (BPs) and protein post-translational modifications (PTMs) is essential for understanding protein function and advancing therapeutic discovery. However, most computational methods remain limited in their generalizability across diverse peptide functions. Here, we present PDe...
2025-02-21T17:31:22Z
24 pages, 9 figures, 4 tables, submitted to arXiv
null
null
A general language model for peptide identification
['Jixiu Zhai', 'Tianchi Lu', 'Haitian Zhong', 'Ziyang Xu', 'Yuhuan Liu', 'Xueying Wang', 'Dan Huang']
2,025
null
0
66
['Computer Science']
2,502.15637
Mantis: Lightweight Calibrated Foundation Model for User-Friendly Time Series Classification
['Vasilii Feofanov', 'Songkang Wen', 'Marius Alonso', 'Romain Ilbert', 'Hongbo Guo', 'Malik Tiomoko', 'Lujia Pan', 'Jianfeng Zhang', 'Ievgen Redko']
['cs.LG', 'cs.AI', 'stat.ML']
In recent years, there has been increasing interest in developing foundation models for time series data that can generalize across diverse downstream tasks. While numerous forecasting-oriented foundation models have been introduced, there is a notable scarcity of models tailored for time series classification. To addr...
2025-02-21T18:06:09Z
null
null
null
null
null
null
null
null
null
null
2,502.15654
Machine-generated text detection prevents language model collapse
['George Drayson', 'Emine Yilmaz', 'Vasileios Lampos']
['cs.CL', 'cs.LG']
As Large Language Models (LLMs) become increasingly prevalent, their generated outputs are proliferating across the web, risking a future where machine-generated content dilutes human-authored text. Since online data is the primary resource for LLM pre-training, subsequent models could be trained on an unknown portion ...
2025-02-21T18:22:36Z
null
null
null
null
null
null
null
null
null
null
2,502.15798
MaxSup: Overcoming Representation Collapse in Label Smoothing
['Yuxuan Zhou', 'Heng Li', 'Zhi-Qi Cheng', 'Xudong Yan', 'Yifei Dong', 'Mario Fritz', 'Margret Keuper']
['cs.LG', 'cs.AI', 'cs.CV']
Label Smoothing (LS) is widely adopted to reduce overconfidence in neural network predictions and improve generalization. Despite these benefits, recent studies reveal two critical issues with LS. First, LS induces overconfidence in misclassified samples. Second, it compacts feature representations into overly tight cl...
2025-02-18T20:10:34Z
24 pages, 15 tables, 5 figures. Preliminary work under review. Do not distribute
null
null
null
null
null
null
null
null
null
2,502.15814
Slamming: Training a Speech Language Model on One GPU in a Day
['Gallil Maimon', 'Avishai Elmakies', 'Yossi Adi']
['cs.LG', 'cs.AI', 'cs.CL', 'cs.SD', 'eess.AS']
We introduce Slam, a recipe for training high-quality Speech Language Models (SLMs) on a single academic GPU in 24 hours. We do so through empirical analysis of model initialisation and architecture, synthetic training data, preference optimisation with synthetic data and tweaking all other components. We empirically d...
2025-02-19T17:21:15Z
ACL 2025 (Findings)
null
null
Slamming: Training a Speech Language Model on One GPU in a Day
['Gallil Maimon', 'Avishai Elmakies', 'Yossi Adi']
2,025
arXiv.org
3
80
['Computer Science', 'Engineering']
2,502.15894
RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers
['Min Zhao', 'Guande He', 'Yixiao Chen', 'Hongzhou Zhu', 'Chongxuan Li', 'Jun Zhu']
['cs.CV']
Recent advancements in video generation have enabled models to synthesize high-quality, minute-long videos. However, generating even longer videos with temporal coherence remains a major challenge and existing length extrapolation methods lead to temporal repetition or motion deceleration. In this work, we systematical...
2025-02-21T19:28:05Z
ICML 2025
null
null
RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers
['Min Zhao', 'Guande He', 'Yixiao Chen', 'Hongzhou Zhu', 'Chongxuan Li', 'Jun Zhu']
2,025
arXiv.org
11
54
['Computer Science']
2,502.1592
Self-Taught Agentic Long Context Understanding
['Yufan Zhuang', 'Xiaodong Yu', 'Jialian Wu', 'Ximeng Sun', 'Ze Wang', 'Jiang Liu', 'Yusheng Su', 'Jingbo Shang', 'Zicheng Liu', 'Emad Barsoum']
['cs.CL', 'cs.AI']
Answering complex, long-context questions remains a major challenge for large language models (LLMs) as it requires effective question clarifications and context retrieval. We propose Agentic Long-Context Understanding (AgenticLU), a framework designed to enhance an LLM's understanding of such queries by integrating ta...
2025-02-21T20:29:36Z
Published at ACL 2025 Main Conference
null
null
null
null
null
null
null
null
null
2,502.16372
COMPASS: Cross-embodiment Mobility Policy via Residual RL and Skill Synthesis
['Wei Liu', 'Huihua Zhao', 'Chenran Li', 'Joydeep Biswas', 'Soha Pouya', 'Yan Chang']
['cs.RO']
As robots are increasingly deployed in diverse application domains, generalizable cross-embodiment mobility policies are increasingly essential. While classical mobility stacks have proven effective on specific robot platforms, they pose significant challenges when scaling to new embodiments. Learning-based methods, su...
2025-02-22T22:26:30Z
null
null
null
COMPASS: Cross-embodiment Mobility Policy via Residual RL and Skill Synthesis
['Wei Liu', 'Hui Zhao', 'Chenran Li', 'Joydeep Biswas', 'Soha Pouya', 'Yan Chang']
2,025
arXiv.org
0
20
['Computer Science']
2,502.16666
SBSC: Step-By-Step Coding for Improving Mathematical Olympiad Performance
['Kunal Singh', 'Ankan Biswas', 'Sayandeep Bhowmick', 'Pradeep Moturi', 'Siva Kishore Gollapalli']
['cs.AI', 'cs.CL', 'cs.LG']
We propose Step-by-Step Coding (SBSC): a multi-turn math reasoning framework that enables Large Language Models (LLMs) to generate sequence of programs for solving Olympiad level math problems. At each step/turn, by leveraging the code execution outputs and programs of previous steps, the model generates the next sub-t...
2025-02-23T17:51:26Z
Published as a full conference paper at ICLR 2025. Shorter(Early) Version accepted at NeurIPS'24 MATH-AI track
null
null
null
null
null
null
null
null
null
2,502.16779
Unposed Sparse Views Room Layout Reconstruction in the Age of Pretrain Model
['Yaxuan Huang', 'Xili Dai', 'Jianan Wang', 'Xianbiao Qi', 'Yixing Yuan', 'Xiangyu Yue']
['cs.CV', 'cs.AI']
Room layout estimation from multiple-perspective images is poorly investigated due to the complexities that emerge from multi-view geometry, which requires muti-step solutions such as camera intrinsic and extrinsic estimation, image matching, and triangulation. However, in 3D reconstruction, the advancement of recent 3...
2025-02-24T02:14:19Z
Accepted by ICLR 2025. Github page:https://github.com/justacar/Plane-DUSt3R
null
null
null
null
null
null
null
null
null
2,502.16839
"Actionable Help" in Crises: A Novel Dataset and Resource-Efficient Models for Identifying Request and Offer Social Media Posts
['Rabindra Lamsal', 'Maria Rodriguez Read', 'Shanika Karunasekera', 'Muhammad Imran']
['cs.CL']
During crises, social media serves as a crucial coordination tool, but the vast influx of posts--from "actionable" requests and offers to generic content like emotional support, behavioural guidance, or outdated information--complicates effective classification. Although generative LLMs (Large Language Models) can addr...
2025-02-24T04:50:06Z
null
null
null
"Actionable Help" in Crises: A Novel Dataset and Resource-Efficient Models for Identifying Request and Offer Social Media Posts
['Rabindra Lamsal', 'M. Read', 'S. Karunasekera', 'Muhammad Imran']
2,025
arXiv.org
0
46
['Computer Science']
2,502.16943
MAD-AD: Masked Diffusion for Unsupervised Brain Anomaly Detection
['Farzad Beizaee', 'Gregory Lodygensky', 'Christian Desrosiers', 'Jose Dolz']
['cs.CV', 'eess.IV']
Unsupervised anomaly detection in brain images is crucial for identifying injuries and pathologies without access to labels. However, the accurate localization of anomalies in medical images remains challenging due to the inherent complexity and variability of brain structures and the scarcity of annotated abnormal dat...
2025-02-24T08:11:29Z
null
Information Processing in Medical Imaging (IPMI), 2025
null
MAD-AD: Masked Diffusion for Unsupervised Brain Anomaly Detection
['Farzad Beizaee', 'Gregory A. Lodygensky', 'Christian Desrosiers', 'J. Dolz']
2,025
arXiv.org
0
41
['Computer Science', 'Engineering']
2,502.16982
Muon is Scalable for LLM Training
['Jingyuan Liu', 'Jianlin Su', 'Xingcheng Yao', 'Zhejun Jiang', 'Guokun Lai', 'Yulun Du', 'Yidao Qin', 'Weixin Xu', 'Enzhe Lu', 'Junjie Yan', 'Yanru Chen', 'Huabin Zheng', 'Yibo Liu', 'Shaowei Liu', 'Bohong Yin', 'Weiran He', 'Han Zhu', 'Yuzhi Wang', 'Jianzhou Wang', 'Mengnan Dong', 'Zheng Zhang', 'Yongsheng Kang', 'Ha...
['cs.LG', 'cs.AI', 'cs.CL']
Recently, the Muon optimizer based on matrix orthogonalization has demonstrated strong results in training small-scale language models, but the scalability to larger models has not been proven. We identify two crucial techniques for scaling up Muon: (1) adding weight decay and (2) carefully adjusting the per-parameter ...
2025-02-24T09:12:29Z
null
null
null
null
null
null
null
null
null
null