English
Related papers

Related papers: Scaling Laws for Reward Model Overoptimization

200 papers

Reward models (RM) capture the values and preferences of humans and play a central role in Reinforcement Learning with Human Feedback (RLHF) to align pretrained large language models (LLMs). Traditionally, training these models relies on…

Machine Learning · Computer Science 2024-09-12 Yifei He , Haoxiang Wang , Ziyan Jiang , Alexandros Papangelis , Han Zhao

Reward design in reinforcement learning (RL) is challenging since specifying human notions of desired behavior may be difficult via reward functions or require many expert demonstrations. Can we instead cheaply design rewards using a…

Machine Learning · Computer Science 2023-03-02 Minae Kwon , Sang Michael Xie , Kalesha Bullard , Dorsa Sadigh

The ability of reinforcement learning algorithms to learn effective policies is determined by the rewards available during training. However, for practical problems, obtaining large quantities of reward labels is often infeasible due to…

Machine Learning · Computer Science 2025-10-02 Shreyas Chaudhari , Renhao Zhang , Philip S. Thomas , Bruno Castro da Silva

Best-of-N (BoN) sampling is a widely used inference-time alignment method for language models, whereby N candidate responses are sampled from a reference model and the one with the highest predicted reward according to a learned reward…

Machine Learning · Computer Science 2026-03-09 Ved Sriraman , Adam Block

Reward Models (RMs) are crucial for online alignment of language models (LMs) with human preferences. However, RM-based preference-tuning is vulnerable to reward hacking, whereby LM policies learn undesirable behaviors from flawed RMs. By…

Computation and Language · Computer Science 2026-03-05 Daniel Fein , Max Lamparth , Violet Xiang , Mykel J. Kochenderfer , Nick Haber

Reinforcement learning agents learn from rewards, but humans can uniquely assign value to novel, abstract outcomes in a goal-dependent manner. However, this flexibility is cognitively costly, making learning less efficient. Here, we propose…

Neurons and Cognition · Quantitative Biology 2025-09-11 Gaia Molinaro , Anne G. E. Collins

Reinforcement learning from human feedback (RLHF) is a recent technique to improve the quality of the text generated by a language model, making it closer to what humans would generate. A core ingredient in RLHF's success in aligning and…

Computation and Language · Computer Science 2024-07-08 Miguel Moura Ramos , Patrick Fernandes , António Farinhas , André F. T. Martins

Machine learning relies on the assumption that unseen test instances of a classification problem follow the same distribution as observed training data. However, this principle can break down when machine learning is used to make important…

Machine Learning · Computer Science 2015-11-24 Moritz Hardt , Nimrod Megiddo , Christos Papadimitriou , Mary Wootters

Reinforcement learning with verifiable rewards (RLVR) scales the reasoning ability of large language models (LLMs) but remains bottlenecked by limited labeled samples for continued data scaling. Reinforcement learning with intrinsic rewards…

Machine Learning · Computer Science 2025-10-13 Chuyi Tan , Peiwen Yuan , Xinglin Wang , Yiwei Li , Shaoxiong Feng , Yueqi Zhang , Jiayi Shi , Ji Zhang , Boyuan Pan , Yao Hu , Kan Li

Reinforcement Learning from Human Feedback (RLHF) facilitates the alignment of large language models with human preferences, significantly enhancing the quality of interactions between humans and models. InstructGPT implements RLHF through…

Computation and Language · Computer Science 2023-10-10 Zheng Yuan , Hongyi Yuan , Chuanqi Tan , Wei Wang , Songfang Huang , Fei Huang

Mathematical reasoning is a challenging task for large language models (LLMs), while the scaling relationship of it with respect to LLM capacity is under-explored. In this paper, we investigate how the pre-training loss, supervised data…

Computation and Language · Computer Science 2023-09-14 Zheng Yuan , Hongyi Yuan , Chengpeng Li , Guanting Dong , Keming Lu , Chuanqi Tan , Chang Zhou , Jingren Zhou

Accurately aligning large language models (LLMs) with human preferences is crucial for informing fair, economically sound, and statistically efficient decision-making processes. However, we argue that the predominant approach for aligning…

Machine Learning · Statistics 2025-08-26 Jiancong Xiao , Ziniu Li , Xingyu Xie , Emily Getzen , Cong Fang , Qi Long , Weijie J. Su

One of the fundamental challenges in reinforcement learning (RL) is the one of data efficiency: modern algorithms require a very large number of training samples, especially compared to humans, for solving environments with high-dimensional…

Machine Learning · Computer Science 2021-05-10 Hlynur Davíð Hlynsson , Laurenz Wiskott

Reinforcement Learning from Human Feedback relies on reward models to align large language models with human preferences. However, RLHF often suffers from reward hacking, wherein policy learning exploits flaws in the trained reward model to…

Machine Learning · Computer Science 2025-05-20 Kangwen Zhao , Jianfeng Cai , Jinhua Zhu , Ruopei Sun , Dongyun Xue , Wengang Zhou , Li Li , Houqiang Li

Reward models are used throughout the post-training of language models to capture nuanced signals from preference data and provide a training target for optimization across instruction following, reasoning, safety, and more domains. The…

Computation and Language · Computer Science 2026-04-24 Saumya Malik , Valentina Pyatkin , Sander Land , Jacob Morrison , Noah A. Smith , Hannaneh Hajishirzi , Nathan Lambert

Exposure bias describes the phenomenon that a language model trained under the teacher forcing schema may perform poorly at the inference stage when its predictions are conditioned on its previous predictions unseen from the training…

Computation and Language · Computer Science 2020-04-02 Yifan Xu , Kening Zhang , Haoyu Dong , Yuezhou Sun , Wenlong Zhao , Zhuowen Tu

Reward models (RMs) are central to aligning large language models (LLMs) with human values but have received less attention than pretrained and post-trained LLMs themselves. Because RMs are initialized from LLMs, they inherit…

Recent advancements in large language models (LLMs) have shifted the post-training paradigm from traditional instruction tuning and human preference alignment toward reinforcement learning (RL) focused on reasoning capabilities. However,…

Artificial Intelligence · Computer Science 2025-11-12 Qianxi He , Qingyu Ren , Shanzhe Lei , Xuhong Wang , Yingchun Wang

Neural scaling laws have driven the field's ever-expanding exponential growth in parameters, data and compute. While scaling behaviors for pretraining losses and discriminative benchmarks are well established, generative benchmarks such as…

Machine Learning · Computer Science 2026-02-03 Rylan Schaeffer , Noam Levi , Brando Miranda , Sanmi Koyejo

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
‹ Prev 1 8 9 10 Next ›