English
Related papers

Related papers: Hard Negative Sample-Augmented DPO Post-Training f…

200 papers

Despite the efficacy of Direct Preference Optimization (DPO) in aligning Large Language Models (LLMs), reward hacking remains a pivotal challenge. This issue emerges when LLMs excessively reduce the probability of rejected completions to…

Computation and Language · Computer Science 2025-08-26 Chenxu Yang , Ruipeng Jia , Mingyu Zheng , Naibin Gu , Zheng Lin , Siyuan Chen , Weichong Yin , Hua Wu , Weiping Wang

Aligning large language models (LLMs) with human values is an increasingly critical step in post-training. Direct Preference Optimization (DPO) has emerged as a simple, yet effective alternative to reinforcement learning from human feedback…

Artificial Intelligence · Computer Science 2025-07-29 Yifan Wang , Runjin Chen , Bolian Li , David Cho , Yihe Deng , Ruqi Zhang , Tianlong Chen , Zhangyang Wang , Ananth Grama , Junyuan Hong

Language model (LM) post-training relies on two stages of human supervision: task demonstrations for supervised finetuning (SFT), followed by preference comparisons for reinforcement learning from human feedback (RLHF). As LMs become more…

Machine Learning · Computer Science 2025-01-15 Yaowen Ye , Cassidy Laidlaw , Jacob Steinhardt

Alignment of large language models (LLMs) with human values has recently garnered significant attention, with prominent examples including the canonical yet costly Reinforcement Learning from Human Feedback (RLHF) and the simple Direct…

Machine Learning · Computer Science 2025-10-14 Xufei Lv , Kehai Chen , Haoyuan Sun , Xuefeng Bai , Min Zhang , Houde Liu , Kehai Chen

Recent advances in Multimodal Large Language Models (MLLMs) have spurred significant progress in Chain-of-Thought (CoT) reasoning. Building on the success of Deepseek-R1, researchers extended multimodal reasoning to post-training paradigms…

Computer Vision and Pattern Recognition · Computer Science 2025-11-11 Jianyu Qi , Ding Zou , Wenrui Yan , Rui Ma , Jiaxu Li , Zhijie Zheng , Zhiguo Yang , Rongchang Zhao

Long chain-of-thought (CoT) significantly enhances the reasoning capabilities of large language models (LLMs). However, extensive reasoning traces lead to inefficiencies and increased time-to-first-token (TTFT). We propose a training…

Computation and Language · Computer Science 2026-01-08 Roy Xie , David Qiu , Deepak Gopinath , Dong Lin , Yanchao Sun , Chong Wang , Saloni Potdar , Bhuwan Dhingra

Reinforcement Learning from Verifiable Rewards (RLVR) significantly enhances large language models (LLMs) reasoning but severely suffers from calibration degeneration, where models become excessively over-confident in incorrect answers.…

Machine Learning · Computer Science 2026-05-28 Zhengzhao Ma , Xueru Wen , Boxi Cao , Yaojie Lu , Hongyu Lin , Jinglin Yang , Min He , Xianpei Han , Le Sun

Advancements in large language models (LLMs) have demonstrated remarkable capabilities across a diverse range of applications. These models excel in generating text completions that are contextually coherent and cover an extensive array of…

Computation and Language · Computer Science 2024-01-22 Bradley Butcher

Post-training has become essential for adapting large language models (LLMs) to complex downstream behaviors, including instruction following, preference alignment, and multi-step reasoning. Reinforcement learning with verifiable rewards…

Machine Learning · Computer Science 2026-05-20 Chengqian Zhang , Wei Zhu , Kyumin Lee

As creative writing tasks do not have singular correct answers, large language models (LLMs) trained to perform these tasks should be able to generate diverse valid outputs. However, LLM post-training often focuses on improving generation…

Computation and Language · Computer Science 2025-03-24 John Joon Young Chung , Vishakh Padmakumar , Melissa Roemmele , Yuqian Sun , Max Kreminski

Reinforcement learning from human feedback (RLHF) has evolved to be one of the main methods for fine-tuning large language models (LLMs). However, existing RLHF methods are non-robust, and their performance deteriorates if the downstream…

Machine Learning · Computer Science 2025-03-04 Debmalya Mandal , Paulius Sasnauskas , Goran Radanovic

Reward-based alignment methods for large language models (LLMs) face two key limitations: vulnerability to reward hacking, where models exploit flaws in the reward signal; and reliance on brittle, labor-intensive prompt engineering when…

Computation and Language · Computer Science 2025-05-20 Zae Myung Kim , Chanwoo Park , Vipul Raheja , Suin Kim , Dongyeop Kang

Preference optimization (PO) is indispensable for large language models (LLMs), with methods such as direct preference optimization (DPO) and proximal policy optimization (PPO) achieving great success. A common belief is that DPO is…

Machine Learning · Computer Science 2026-05-18 Yue Wang , Qizhou Wang , Zizhuo Zhang , Gang Niu , Bo Han , Masashi Sugiyama

Diffusion large language models (dLLMs) offer a promising route to parallel and efficient text generation, but improving their reasoning ability requires effective post-training. Reinforcement learning with verifiable rewards (RLVR) is a…

Computation and Language · Computer Science 2026-05-12 Zichao Yu , Shengze Xu , Bingqing Jiang , Wenyi Zhang , Difan Zou

Reinforcement Learning (RL) has emerged as a central paradigm for advancing Large Language Models (LLMs), where pre-training and RL post-training share the same log-likelihood formulation. In contrast, recent RL approaches for diffusion…

Machine Learning · Computer Science 2025-09-30 Shuchen Xue , Chongjian Ge , Shilong Zhang , Yichen Li , Zhi-Ming Ma

With the rapid development of Large Language Models (LLMs), numerous Reinforcement Learning from Human Feedback (RLHF) algorithms have been introduced to improve model safety and alignment with human preferences. These algorithms can be…

Machine Learning · Computer Science 2025-02-06 Xuerui Su , Yue Wang , Jinhua Zhu , Mingyang Yi , Feng Xu , Zhiming Ma , Yuting Liu

Reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO) are important techniques to align large language models (LLM) with human preference. However, the quality of RLHF and DPO training is seriously…

Machine Learning · Computer Science 2025-12-10 Ziyi Chen , Junyi Li , Peiran Yu , Heng Huang

Aligning large language models (LLMs) with human preferences usually requires fine-tuning methods such as RLHF and DPO. These methods directly optimize the model parameters, so they cannot be used in test-time to improve model performance,…

Machine Learning · Computer Science 2025-07-04 Xinnan Zhang , Chenliang Li , Siliang Zeng , Jiaxiang Li , Zhongruo Wang , Kaixiang Lin , Songtao Lu , Alfredo Garcia , Mingyi Hong

Mathematical reasoning presents a significant challenge for Large Language Models (LLMs) as it requires ensuring the correctness of each reasoning step. Researchers have been strengthening the mathematical reasoning abilities of LLMs…

Machine Learning · Computer Science 2025-06-23 Yunze Lin

Aligning Large Language Models (LLMs) traditionally relies on costly training and human preference annotations. Self-alignment seeks to reduce these expenses by enabling models to align themselves. To further lower costs and achieve…

Computation and Language · Computer Science 2024-11-15 Somanshu Singla , Zhen Wang , Tianyang Liu , Abdullah Ashfaq , Zhiting Hu , Eric P. Xing