语言大模型 / LLM - arXivDaily 专题

2511.15038 2025-11-20 cs.SD cs.AI eess.AS 57%

Aligning Generative Music AI with Human Preferences: Methods and Challenges

Dorien Herremans, Abhinaba Roy

专题命中后训练与偏好优化：preference optimization(abstract)；分类 cs.AI

Comments Accepted at the AAAI-2026 Senior Member Track

2509.12179 2025-11-19 cs.AI cs.MA 57%

Co-Alignment: Rethinking Alignment as Bidirectional Human-AI Cognitive Adaptation

Yubo Li, Weiyi Song

机构 * Carnegie Mellon University（卡内基梅隆大学）

专题命中后训练与偏好优化：RLHF(abstract)；分类 cs.AI

2510.20411 2025-10-24 cs.CL 57%

Teacher Demonstrations in a BabyLM's Zone of Proximal Development for Contingent Multi-Turn Interaction

Suchir Salhan, Hongyi Gu, Donya Rooein, Diana Galvan-Sosa, Gabrielle Gaudeau, Andrew Caines, Zheng Yuan, Paula Buttery

机构 * ALTA Institute（ALTA研究所）； Dept. of Computer Science & Technology, Cambridge University（计算机科学与技术系，剑桥大学）； Bocconi University（博科尼大学）； Sheffield University（谢菲尔德大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.CL

Comments Outstanding Paper Award, EMNLP 2025 BabyLM Workshop - Oral presentation, Suzhou, China

2510.13656 2025-10-23 cs.LG 57%

Rebalancing with Calibrated Sub-classes (RCS): A Statistical Fusion-based Framework for Robust Imbalanced Classification across Modalities

Priyobrata Mondal, Faizanuddin Ansari, Swagatam Das

机构 * Electronics and Communication Sciences Unit（电子与通信科学单元）； Indian Statistical Institute（印度统计研究所）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

2510.14253 2025-10-22 cs.AI 57%

Towards Agentic Self-Learning LLMs in Search Environment

Wangtao Sun, Xiang Cheng, Jialin Fan, Yao Xu, Xing Yu, Shizhu He, Jun Zhao, Kang Liu

机构 * The Laboratory of Cognition and Decision Intelligence for Complex Systems, Institute of Automation, Chinese Academy of Sciences（认知与决策智能复杂系统实验室，自动化研究所，中国科学院）； School of Artificial Intelligence, University of Chinese Academy of Sciences（人工智能学院，中国科学院大学）； Xiaohongshu Inc（小红书公司）； Meituan Inc（美团公司）； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）； Tsinghua University（清华大学）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.AI

2503.09561 2025-10-17 cs.LG 57%

Strategyproof Reinforcement Learning from Human Feedback

Thomas Kleine Buening, Jiarui Gan, Debmalya Mandal, Marta Kwiatkowska

机构 * ETH Zurich（苏黎世联邦理工学院）； ETH AI Center（ETH人工智能中心）； Department of Computer Science University of Oxford（牛津大学计算机科学系）； Department of Computer Science University of Warwick（Warwick大学计算机科学系）

专题命中后训练与偏好优化：RLHF(abstract)；分类 cs.LG

Comments To appear at NeurIPS 2025

2510.14583 2025-10-17 cs.CV cs.CL 57%

Talking Points: Describing and Localizing Pixels

Matan Rusanovsky, Shimon Malnick, Shai Avidan

机构 * Tel Aviv University（特拉维夫大学）

专题命中后训练与偏好优化：language model(abstract)；分类 cs.CL

2510.06629 2025-10-09 cs.CR cs.CV cs.LG 57%

Unsupervised Backdoor Detection and Mitigation for Spiking Neural Networks

Jiachen Li, Bang Wu, Xiaoyu Xia, Xiaoning Liu, Xun Yi, Xiuzhen Zhang

机构 * RMIT University（皇家墨尔本理工大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

Comments To appear in The 28th International Symposium on Research in Attacks, Intrusions and Defenses (RAID 2025)

2510.04096 2025-10-07 cs.IR cs.GT cs.LG 57%

RLRF: Competitive Search Agent Design via Reinforcement Learning from Ranker Feedback

Tommy Mordo, Sagie Dekel, Omer Madmon, Moshe Tennenholtz, Oren Kurland

机构 * Faculty of Data and Decision Sciences（数据与决策科学学院）； Technion – Israel Institute of Technology（技术学院–以色列理工学院）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.LG

2510.03679 2025-10-07 cs.LG stat.ML 57%

Group Policy Gradient

Junhua Chen, Zixi Zhang, Hantao Zhong, Rika Antonova

机构 * University of Cambridge（剑桥大学）

专题命中后训练与偏好优化：RLHF(abstract)；分类 cs.LG

2509.18314 2025-10-06 cs.CL 57%

Exploiting Tree Structure for Credit Assignment in RL Training of LLMs

Hieu Tran, Zonghai Yao, Hong Yu

机构 * Center for Healthcare Organization and Implementation Research, VA Bedford Health Care（VA Bedford Health Care 机构）； Manning College of Information and Computer Sciences, University of Massachusetts Amherst（马萨诸塞大学阿默斯特分校信息与计算机科学学院）； Miner School of Computer and Information Sciences, University of Massachusetts Lowell（马萨诸塞大学洛厄尔分校计算机与信息科学学院）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.CL

Comments 15 pages

2509.18849 2025-09-26 cs.AI 57%

MAPO: Mixed Advantage Policy Optimization

Wenke Huang, Quan Zhang, Yiyang Fang, Jian Liang, Xuankun Rong, Huanjin Yao, Guancheng Wan, Ke Liang, Wenwen He, Mingjun Li, Leszek Rutkowski, Mang Ye, Bo Du, Dacheng Tao

机构 * Wuhan University（武汉大学）； ByteDance（字节跳动）； National University of Defense Technology（国防科技大学）； Nanyang Technological University（南洋理工大学）； The AGH University of Krakow（克拉科夫AGH大学）

专题命中后训练与偏好优化：foundation model(abstract)；分类 cs.AI

2509.17917 2025-09-23 cs.AI 57%

Orcust: Stepwise-Feedback Reinforcement Learning for GUI Agent

Junyu Lu, Songxin Zhang, Zejian Xie, Zhuoyang Song, Jiaxing Zhang

机构 * Lionrock AI Lab（狮岩人工智能实验室）； China Merchants Research Institute of Advanced Technology（中国商人先进科技研究院）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.AI

2407.04615 2025-09-23 cs.CL 57%

On the Low-Rank Parametrization of Reward Models for Controlled Language Generation

Sergey Troshin, Vlad Niculae, Antske Fokkens

专题命中后训练与偏好优化：language model(abstract)；分类 cs.CL

Comments TMLR 2025

2504.05154 2025-09-19 cs.CL 57%

CARE: Multilingual Human Preference Learning for Cultural Awareness

Geyang Guo, Tarek Naous, Hiromi Wakaki, Yukiko Nishimura, Yuki Mitsufuji, Alan Ritter, Wei Xu

机构 * Georgia Institute of Technology（佐治亚理工学院）； Sony Group Corporation（索尼集团）； Sony AI（索尼人工智能）

专题命中后训练与偏好优化：language model(abstract)；分类 cs.CL

Comments Accepted at EMNLP 2025

2509.12618 2025-09-17 cs.RO cs.AI cs.CV 57%

ActiveVLN: Towards Active Exploration via Multi-Turn RL in Vision-and-Language Navigation

Zekai Zhang, Weiye Zhu, Hewei Pan, Xiangchen Wang, Rongtao Xu, Xing Sun, Feng Zheng

机构 * Southern University of Science and Technology, China（南方科技大学，中国）； Spatialtemporal AI, China（时空AI，中国）； MBZUAI, UAE（MBZUAI，阿联酋）； Tencent Youtu Lab（腾讯优图实验室）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.AI

2502.19279 2025-09-12 cs.CL 57%

CritiQ: Mining Data Quality Criteria from Human Preferences

Honglin Guo, Kai Lv, Qipeng Guo, Tianyi Liang, Zhiheng Xi, Demin Song, Qiuyinzhe Zhang, Yu Sun, Kai Chen, Xipeng Qiu, Tao Gui

专题命中后训练与偏好优化：language model(abstract)；分类 cs.CL

Comments to be published in ACL 2025, Code is available at https://github.com/KYLN24/CritiQ

2501.06058 2025-09-12 cs.MA cs.LG 57%

Capability-Aware Shared Hypernetworks for Flexible Heterogeneous Multi-Robot Coordination

Kevin Fu, Shalin Anand Jain, Pierce Howell, Harish Ravichandar

机构 * Georgia Institute of Technology（佐治亚理工学院）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

Comments 22 pages, 8 figures, equal authorship between Kevin Fu and Shalin Anand Jain Manuscript accepted for publication at the 9th Conference on Robot Learning (CoRL 2025), Seoul, Korea

2509.07108 2025-09-10 stat.ML cs.LG 57%

ADHAM: Additive Deep Hazard Analysis Mixtures for Interpretable Survival Regression

Mert Ketenci, Vincent Jeanselme, Harry Reyes Nieva, Shalmali Joshi, Noémie Elhadad

机构 * Columbia University（哥伦比亚大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

2509.04372 2025-09-05 stat.ML cs.GL cs.LG math.ST stat.TH 57%

Connections between reinforcement learning with feedback,test-time scaling, and diffusion guidance: An anthology

Yuchen Jiao, Yuxin Chen, Gen Li

机构 * Department of Statistics and Data Science, Chinese University of Hong Kong（统计与数据科学系，香港中文大学）； Department of Statistics and Data Science, the Wharton School, University of Pennsylvania（统计与数据科学系，沃顿商学院，宾夕法尼亚大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

2508.21240 2025-09-01 cs.LG 57%

Class Incremental Continual Learning with Self-Organizing Maps and Variational Autoencoders Using Synthetic Replay

Pujan Thapa, Alexander Ororbia, Travis Desell

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

2508.16656 2025-08-26 cs.LG 57%

OASIS: Open-world Adaptive Self-supervised and Imbalanced-aware System

Miru Kim, Mugon Joe, Minhae Kwon

机构 * Soongsil University（松蔚大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

Comments Accepted at the 34th ACM International Conference on Information and Knowledge Management (CIKM 2025)

2508.16230 2025-08-25 cs.CV cs.AI 57%

FlexMUSE: Multimodal Unification and Semantics Enhancement Framework with Flexible interaction for Creative Writing

Jiahao Chen, Zhiyong Ma, Wenbiao Du, Qingyuan Chuai

机构 * Cao Tu Li(Guangzhou) Technology Co., Ltd, China（广东曹图利技术有限公司）； South China University of Technology, China（华南理工大学）

专题命中后训练与偏好优化：preference optimization(abstract)；分类 cs.AI

2407.16800 2025-08-22 cs.LG math.OC 57%

Wasserstein Distributionally Robust Shallow Convex Neural Networks

Julien Pallage, Antoine Lesage-Landry

机构 * Department of Electrical Engineering（电气工程系）； Polytechnique Montréal（蒙特利尔理工学院）； GERAD（GERAD研究中心）； Mila（Mila研究中心）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

2508.11363 2025-08-18 cs.LG 57%

Fusing Rewards and Preferences in Reinforcement Learning

Sadegh Khorasani, Saber Salehkaleybar, Negar Kiyavash, Matthias Grossglauser

专题命中后训练与偏好优化：RLHF(abstract)；分类 cs.LG

2508.04524 2025-08-07 cs.CV cs.AI 57%

RAIDX: A Retrieval-Augmented Generation and GRPO Reinforcement Learning Framework for Explainable Deepfake Detection

Tianxiao Li, Zhenglin Huang, Haiquan Wen, Yiwei He, Shuchang Lyu, Baoyuan Wu, Guangliang Cheng

机构 * University of Liverpool（利物浦大学）； Beihang University（北航大学）； The Chinese University of Hong Kong（香港中文大学）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.AI

2507.20880 2025-07-29 cs.SD cs.AI 57%

JAM: A Tiny Flow-based Song Generator with Fine-grained Controllability and Aesthetic Alignment

Renhang Liu, Chia-Yu Hung, Navonil Majumder, Taylor Gautreaux, Amir Ali Bagherzadeh, Chuan Li, Dorien Herremans, Soujanya Poria

专题命中后训练与偏好优化：preference optimization(abstract)；分类 cs.AI

Comments https://github.com/declare-lab/jamify

2507.04695 2025-07-22 cs.LG 57%

Interpretable Reward Modeling with Active Concept Bottlenecks

Sonia Laguna, Katarzyna Kobalczyk, Julia E. Vogt, Mihaela Van der Schaar

机构 * Department of Computer Science, ETH Zurich, Zurich, Switzerland（苏黎世联邦理工学院计算机科学系）； Department of Applied Mathematics and Theoretical Physics, University of Cambridge, United Kingdom（剑桥大学应用数学与理论物理系）

专题命中后训练与偏好优化：RLHF(abstract)；分类 cs.LG

Journal ref ICML 2025 Workshop on Programmatic Representations for Agent Learning

2504.12016 2025-07-18 cs.LG 57%

Active Human Feedback Collection via Neural Contextual Dueling Bandits

Arun Verma, Xiaoqiang Lin, Zhongxiang Dai, Daniela Rus, Bryan Kian Hsiang Low

机构 * Singapore-MIT Alliance for Research and Technology（新加坡-麻省理工联合研究技术联盟）； Department of Computer Science, National University of Singapore（国立新加坡大学计算机科学系）； The Chinese University of Hong Kong, Shenzhen, China（香港中文大学（深圳））； CSAIL, MIT, USA（MIT计算机科学与人工智能实验室）

专题命中后训练与偏好优化：LLM(abstract)；分类 cs.LG

Comments 19 pages

2502.04850 2025-07-15 cs.LG cs.DC 57%

Aequa: Fair Model Rewards in Collaborative Learning via Slimmable Networks

Nurbek Tastan, Samuel Horvath, Karthik Nandakumar

机构 * Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)（Mohamed bin Zayed大学人工智能学院）； Michigan State University (MSU)（密歇根州立大学）

专题命中后训练与偏好优化：post-training(abstract)；分类 cs.LG

1. 后训练与偏好优化 4541 篇

Aligning Generative Music AI with Human Preferences: Methods and Challenges

Co-Alignment: Rethinking Alignment as Bidirectional Human-AI Cognitive Adaptation

Teacher Demonstrations in a BabyLM's Zone of Proximal Development for Contingent Multi-Turn Interaction

Rebalancing with Calibrated Sub-classes (RCS): A Statistical Fusion-based Framework for Robust Imbalanced Classification across Modalities

Towards Agentic Self-Learning LLMs in Search Environment

Strategyproof Reinforcement Learning from Human Feedback

Talking Points: Describing and Localizing Pixels

Unsupervised Backdoor Detection and Mitigation for Spiking Neural Networks

RLRF: Competitive Search Agent Design via Reinforcement Learning from Ranker Feedback

Group Policy Gradient

Exploiting Tree Structure for Credit Assignment in RL Training of LLMs

MAPO: Mixed Advantage Policy Optimization

Orcust: Stepwise-Feedback Reinforcement Learning for GUI Agent

On the Low-Rank Parametrization of Reward Models for Controlled Language Generation

CARE: Multilingual Human Preference Learning for Cultural Awareness

ActiveVLN: Towards Active Exploration via Multi-Turn RL in Vision-and-Language Navigation

CritiQ: Mining Data Quality Criteria from Human Preferences

Capability-Aware Shared Hypernetworks for Flexible Heterogeneous Multi-Robot Coordination

ADHAM: Additive Deep Hazard Analysis Mixtures for Interpretable Survival Regression

Connections between reinforcement learning with feedback,test-time scaling, and diffusion guidance: An anthology

Class Incremental Continual Learning with Self-Organizing Maps and Variational Autoencoders Using Synthetic Replay

OASIS: Open-world Adaptive Self-supervised and Imbalanced-aware System

FlexMUSE: Multimodal Unification and Semantics Enhancement Framework with Flexible interaction for Creative Writing

Wasserstein Distributionally Robust Shallow Convex Neural Networks

Fusing Rewards and Preferences in Reinforcement Learning

RAIDX: A Retrieval-Augmented Generation and GRPO Reinforcement Learning Framework for Explainable Deepfake Detection

JAM: A Tiny Flow-based Song Generator with Fine-grained Controllability and Aesthetic Alignment

Interpretable Reward Modeling with Active Concept Bottlenecks

Active Human Feedback Collection via Neural Contextual Dueling Bandits

Aequa: Fair Model Rewards in Collaborative Learning via Slimmable Networks