语言大模型 / LLM - arXivDaily 专题

2501.04961 2025-10-23 cs.CL cs.AI cs.CE cs.LG 87%

Demystifying Domain-adaptive Post-training for Financial LLMs

Zixuan Ke, Yifei Ming, Xuan-Phi Nguyen, Caiming Xiong, Shafiq Joty

机构 * Salesforce AI Research（Salesforce人工智能研究）

专题命中后训练与偏好优化：post-training(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments EMNLP 2025 (Oral, ARR best paper nomination)

2510.15020 2025-10-23 stat.ML cs.AI cs.CL cs.LG math.ST stat.TH 85%

The Coverage Principle: How Pre-Training Enables Post-Training

Fan Chen, Audrey Huang, Noah Golowich, Sadhika Malladi, Adam Block, Jordan T. Ash, Akshay Krishnamurthy, Dylan J. Foster

专题命中后训练与偏好优化：post-training(title,abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

2510.15061 2025-10-23 cs.LG cs.CL 84%

Antislop: A Comprehensive Framework for Identifying and Eliminating Repetitive Patterns in Language Models

Samuel Paech, Allen Roush, Judah Goldfeder, Ravid Shwartz-Ziv

专题命中后训练与偏好优化：language model(title)；LLM(abstract)；preference optimization(abstract)；分类 cs.CL、cs.LG

Comments 11 pages + appendices, 16 figures

2510.19325 2025-10-23 cs.CL cs.AI 79%

Balancing Rewards in Text Summarization: Multi-Objective Reinforcement Learning via HyperVolume Optimization

Junjie Song, Yiwen Liu, Dapeng Li, Yin Sun, Shukun Fu, Siqi Chen, Yuji Cao

机构 * AI-4-Business of Li Auto Inc.（利汽车有限公司AI-4-Business）

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

2410.01735 2025-10-23 cs.CL cs.LG 79%

LASeR: Learning to Adaptively Select Reward Models with Multi-Armed Bandits

Duy Nguyen, Archiki Prasad, Elias Stengel-Eskin, Mohit Bansal

机构 * UNC Chapel Hill（杜克大学夏洛特分校）； The University of Texas at Austin（德克萨斯大学奥斯汀分校）

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

Comments NeurIPS 2025 camera-ready. First two authors contributed equally. Code: https://github.com/duykhuongnguyen/LASeR-MAB

2510.19737 2025-10-23 cond-mat.mtrl-sci 78%

Accelerating Moment Tensor Potentials through Post-Training Pruning

Zijian Meng, Karim Zongo, Matthew Thoms, Ryan Eric Grant, Laurent Karim Béland

专题命中后训练与偏好优化：post-training(title,abstract)

Comments 16 pages, 5 figures Software available from https://github.com/RichardZJM/MTP_basis_optimization

2510.18895 2025-10-23 cs.SE cs.AI cs.HC 77%

CosmoCore Affective Dream-Replay Reinforcement Learning for Code Generation

Santhosh Kumar Ravindran

机构 * Microsoft Corporation（微软公司）

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

Comments 12 pages

2510.19479 2025-10-23 cs.LG cs.AI 76%

Graph Unlearning Meets Influence-aware Negative Preference Optimization

Qiang Chen, Zhongze Wu, Ang He, Xi Lin, Shuo Jiang, Shan You, Chang Xu, Yi Chen, Xiu Su

机构 * Central South University（中南大学）； Shanghai Maritime University（上海 Maritime 大学）； Shanghai Jiao Tong University（上海交通大学）； Tongji University（同济大学）； SenseTime Research（商汤科技研究院）； University of Sydney（悉尼大学）； Hong Kong University of Science（香港科学大学）

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

2510.19173 2025-10-23 q-fin.CP cs.AI cs.LG 73%

News-Aware Direct Reinforcement Trading for Financial Markets

Qing-Yu Lan, Zhan-He Wang, Jun-Qian Jiang, Yu-Tong Wang, Yun-Song Piao

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

Comments 9 pages, 4 figures, 3 tables

2510.19050 2025-10-23 cs.AI cs.LG 73%

Rectifying Shortcut Behaviors in Preference-based Reward Learning

Wenqian Ye, Guangtao Zheng, Aidong Zhang

机构 * University of Virginia（弗吉尼亚大学）

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

Comments NeurIPS 2025

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

1. 后训练与偏好优化 11 篇