English
Related papers

Related papers: Coverage Improvement and Fast Convergence of On-po…

200 papers

Reinforcement learning with verifiable rewards (RLVR) has become a standard paradigm for post-training large language models. While Group Relative Policy Optimization (GRPO) is widely adopted, its coarse credit assignment uniformly…

Machine Learning · Computer Science 2026-04-03 Gengsheng Li , Tianyu Yang , Junfeng Fang , Mingyang Song , Mao Zheng , Haiyun Guo , Dan Zhang , Jinqiao Wang , Tat-Seng Chua

Most reinforcement learning algorithms seek a single optimal strategy that solves a given task. However, it can often be valuable to learn a diverse set of solutions, for instance, to make an agent's interaction with users more engaging, or…

Machine Learning · Computer Science 2024-01-09 Wentse Chen , Shiyu Huang , Yuan Chiang , Tim Pearce , Wei-Wei Tu , Ting Chen , Jun Zhu

Direct Preference Optimization (DPO) is a widely used RL-free method for aligning language models from pairwise preferences, but it models preferences over full sequences even though generation is driven by per-token decisions. Existing…

Computation and Language · Computer Science 2026-05-15 Truong Nguyen , Tien-Phat Nguyen , Linh Ngo Van , Duy Minh Ho Nguyen , Khoa D. Doan , Trung Le

Reinforcement Learning (RL) from Human Preference-based feedback is a popular paradigm for fine-tuning generative models, which has produced impressive models such as GPT-4 and Claude3 Opus. This framework often consists of two steps:…

Machine Learning · Computer Science 2024-04-17 Jonathan D. Chang , Wenhao Zhan , Owen Oertell , Kianté Brantley , Dipendra Misra , Jason D. Lee , Wen Sun

Direct Preference Optimization (DPO) is a powerful paradigm to align language models with human preferences using pairwise comparisons. However, its binary win-or-loss supervision often proves insufficient for training small models with…

Computation and Language · Computer Science 2025-09-23 Minchan Kwon , Junwon Ko , Kangil Kim , Junmo Kim

With the rapid development and widespread application of Large Language Models (LLMs), their potential safety risks have attracted widespread attention. Reinforcement Learning from Human Feedback (RLHF) has been adopted to enhance the…

Artificial Intelligence · Computer Science 2026-03-25 Shiji Zhao , Mengyang Wang , Shukun Xiong , Fangzhou Chen , Qihui Zhu , Shouwei Ruan , Yisong Xiao , Ranjie Duan , Xun Chen , XingXing Wei

Reinforcement learning for large language models faces a fundamental trade-off between sample efficiency and asymptotic performance: strictly on-policy methods discard trajectories after a single update, while off-policy reuse introduces…

Machine Learning · Computer Science 2026-05-26 Changyu Chen , Xiting Wang , Rui Yan

Self-Distillation Policy Optimization (SDPO) provides dense token-level credit assignment for reinforcement learning with large language models by leveraging the model's own feedback-conditioned predictions as a self-teacher. Unlike GRPO,…

Machine Learning · Computer Science 2026-05-28 Zehao Liu , Yuanpu Cao , Jinghui Chen , Vasant G. Honavar

In real-world decision making tasks, it is critical for data-driven reinforcement learning methods to be both stable and sample efficient. On-policy methods typically generate reliable policy improvement throughout training, while…

Machine Learning · Computer Science 2021-11-02 James Queeney , Ioannis Ch. Paschalidis , Christos G. Cassandras

Offline preference optimization allows fine-tuning large models directly from offline data, and has proved effective in recent alignment practices. We propose generalized preference optimization (GPO), a family of offline losses…

Aligning large language models (LLMs) is a central objective of post-training, often achieved through reward modeling and reinforcement learning methods. Among these, direct preference optimization (DPO) has emerged as a widely adopted…

Computation and Language · Computer Science 2026-03-03 Aladin Djuhera , Farhan Ahmed , Swanand Ravindra Kadhe , Syed Zawad , Heiko Ludwig , Holger Boche

A critical component of the current generation of language models is preference alignment, which aims to precisely control the model's behavior to meet human needs and values. The most notable among such methods is Reinforcement Learning…

Artificial Intelligence · Computer Science 2024-10-22 Oh Joon Kwon , Daiki E. Matsunaga , Kee-Eung Kim

We propose DiFFPO, Diffusion Fast and Furious Policy Optimization, a unified framework for training masked diffusion large language models (dLLMs) to reason not only better (furious), but also faster via reinforcement learning (RL). We…

Machine Learning · Computer Science 2026-01-13 Hanyang Zhao , Dawen Liang , Wenpin Tang , David Yao , Nathan Kallus

Aligning language models with human preferences relies on pairwise preference datasets. While some studies suggest that on-policy data consistently outperforms off -policy data for preference learning, others indicate that the advantages of…

Computation and Language · Computer Science 2025-05-06 Tianjian Li , Daniel Khashabi

Large language models trained with reinforcement learning (RL) for mathematical reasoning face a fundamental challenge: on problems the model cannot solve at all - "cliff" prompts - the RL gradient vanishes entirely, preventing any learning…

Machine Learning · Computer Science 2026-03-26 Ken Ding

On-policy reinforcement learning methods like GRPO suffer from mode collapse: they exhibit reduced solution diversity, concentrating probability mass on a single solution once discovered and ceasing exploration of alternative strategies. We…

Artificial Intelligence · Computer Science 2026-05-20 Xiaozhe Li , Yang Li , Xinyu Fang , Shengyuan Ding , Peiji Li , Yongkang Chen , Yichuan Ma , Tianyi Lyu , Linyang Li , Dahua Lin , Qipeng Guo , Qingwen Liu , Kai Chen

Preference-based alignment objectives have been widely adopted, from RLHF-style pairwise learning in large language models to emerging applications in recommender systems. Yet, existing work rarely examines how Direct Preference…

Information Retrieval · Computer Science 2026-04-01 Hejin Huang , Jusheng Zhang , Kaitong Cai , Jian Wang , Rong Pan

The rapidly increasing capabilities of large language models (LLMs) raise an urgent need to align AI systems with diverse human preferences to simultaneously enhance their usefulness and safety, despite the often conflicting nature of these…

Machine Learning · Computer Science 2024-03-06 Zixuan Liu , Xiaolin Sun , Zizhan Zheng

Large language models (LLMs), despite their extensive pretraining on diverse datasets, require effective alignment to human preferences for practical and reliable deployment. Conventional alignment methods typically employ off-policy…

Computation and Language · Computer Science 2025-07-29 Hyeonji Lee , Daejin Jo , Seohwan Yun , Sungwoong Kim

Without using explicit reward, direct preference optimization (DPO) employs paired human preference data to fine-tune generative models, a method that has garnered considerable attention in large language models (LLMs). However, exploration…

Computer Vision and Pattern Recognition · Computer Science 2025-03-25 Yunhong Lu , Qichao Wang , Hengyuan Cao , Xierui Wang , Xiaoyin Xu , Min Zhang