English
Related papers

Related papers: On the Robustness of Reward Models for Language Mo…

200 papers

Reinforcement learning from human feedback (RLHF) is a standard approach for fine-tuning large language models to follow instructions. As part of this process, learned reward models are used to approximately model human preferences.…

Machine Learning · Computer Science 2024-03-12 Thomas Coste , Usman Anwar , Robert Kirk , David Krueger

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

Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for aligning large language models (LLMs) with human preferences, yet it is susceptible to reward overoptimization, in which policy models overfit to the reward model,…

Reward models in RLHF are trained to score only the final token of a response - a choice that discards rich signal from every intermediate position and produces models whose token-level outputs are noise. We argue this is a missed…

Machine Learning · Computer Science 2026-04-28 Alex Nikulkov

Aligning the behavior of Large language models (LLMs) with human intentions and values remains a critical challenge. Reinforcement learning from human feedback (RLHF) aligns LLMs by training a reward model (RM) on human preferences and…

Computation and Language · Computer Science 2025-12-25 Jiayi Zhou , Jiaming Ji , Juntao Dai , Dong Li , Yaodong Yang

Reward models (RMs) are essential for aligning Large Language Models (LLMs) with human preferences. However, they often struggle with capturing complex human preferences and generalizing to unseen data. To address these challenges, we…

Computation and Language · Computer Science 2025-08-06 Anamika Lochab , Ruqi Zhang

Reinforcement Learning from Human Feedback (RLHF) has emerged as a important paradigm for aligning large language models (LLMs) with human preferences during post-training. This framework typically involves two stages: first, training a…

Machine Learning · Computer Science 2025-04-08 Wenyuan Xu , Xiaochen Zuo , Chao Xin , Yu Yue , Lin Yan , Yonghui Wu

Reinforcement Learning with Human Feedback (RLHF) and Reinforcement Learning with Verifiable Rewards (RLVR) are the main RL paradigms used in LLM post-training, each offering distinct advantages. However, RLHF struggles with…

Computation and Language · Computer Science 2026-05-19 Zhilin Wang , Jiaqi Zeng , Olivier Delalleau , Ellie Evans , Daniel Egert , Hoo-Chang Shin , Felipe Soares , Yi Dong , Oleksii Kuchaiev

Reward models have been increasingly critical for improving the reasoning capability of LLMs. Existing research has shown that a well-trained reward model can substantially improve model performances at inference time via search. However,…

Machine Learning · Computer Science 2024-11-28 Jiaxuan Gao , Shusheng Xu , Wenjie Ye , Weilin Liu , Chuyi He , Wei Fu , Zhiyu Mei , Guangju Wang , Yi Wu

Frontier models have demonstrated exceptional capabilities following the integration of task-reward-based reinforcement learning (RL) into their training pipelines, enabling systems to evolve from pure reasoning models into sophisticated…

Machine Learning · Computer Science 2026-04-20 Sarthak Mittal , Leo Gagnon , Guillaume Lajoie

Reward models are central to aligning language models with human preferences via reinforcement learning (RL). As RL is increasingly applied to settings such as verifiable rewards and multi-objective alignment, RMs are expected to encode…

Machine Learning · Computer Science 2026-05-21 Jiwoo Hong , Shao Tang , Zhipeng Wang

Despite the success of Reinforcement Learning from Human Feedback (RLHF) in aligning language models with human values, reward hacking-or reward over-optimization-remains a major challenge. We identify two key obstacles to its mitigation:…

Machine Learning · Computer Science 2025-10-16 Yuchun Miao , Liang Ding , Sen Zhang , Rong Bao , Lefei Zhang , Dacheng Tao

To ensure that large language model (LLM) responses are helpful and non-toxic, a reward model trained on human preference data is usually used. LLM responses with high rewards are then selected through best-of-$n$ (BoN) sampling or the LLM…

Machine Learning · Computer Science 2024-07-04 Adam X. Yang , Maxime Robeyns , Thomas Coste , Zhengyan Shi , Jun Wang , Haitham Bou-Ammar , Laurence Aitchison

The effectiveness of reinforcement learning (RL) agents in continuous control robotics tasks is mainly dependent on the design of the underlying reward function, which is highly prone to reward hacking. A misalignment between the reward…

This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many algorithms are highly sensitive to misspecifications in the…

Machine Learning · Computer Science 2025-10-30 Erhan Xu , Kai Ye , Hongyi Zhou , Luhan Zhu , Francesco Quinzan , Chengchun Shi

Despite the success of reinforcement learning from human feedback (RLHF) in aligning language models with human values, reward hacking, also termed reward overoptimization, remains a critical challenge. This issue primarily arises from…

Machine Learning · Computer Science 2024-11-04 Yuchun Miao , Sen Zhang , Liang Ding , Rong Bao , Lefei Zhang , Dacheng Tao

While large language models demonstrate remarkable capabilities, they often present challenges in terms of safety, alignment with human values, and stability during training. Here, we focus on two prevalent methods used to align these…

Computation and Language · Computer Science 2023-10-26 Gabriel Mukobi , Peter Chatain , Su Fong , Robert Windesheim , Gitta Kutyniok , Kush Bhatia , Silas Alberti

This paper explores multiple optimization methods to improve the performance of rating-based reinforcement learning (RbRL). RbRL, a method based on the idea of human ratings, has been developed to infer reward functions in reward-free…

Machine Learning · Computer Science 2025-01-15 Evelyn Rose , Devin White , Mingkang Wu , Vernon Lawhern , Nicholas R. Waytowich , Yongcan Cao

Reinforcement learning from human feedback (RLHF) emerges as a promising paradigm for aligning large language models (LLMs). However, a notable challenge in RLHF is overoptimization, where beyond a certain threshold, the pursuit of higher…

Machine Learning · Computer Science 2024-01-02 Yuanzhao Zhai , Han Zhang , Yu Lei , Yue Yu , Kele Xu , Dawei Feng , Bo Ding , Huaimin Wang

Reward models have become a staple in modern NLP, serving as not only a scalable text evaluator, but also an indispensable component in many alignment recipes and inference-time algorithms. However, while recent reward models increase…

Computation and Language · Computer Science 2025-09-22 Zhaofeng Wu , Michihiro Yasunaga , Andrew Cohen , Yoon Kim , Asli Celikyilmaz , Marjan Ghazvininejad