English
Related papers

Related papers: The Perils of Optimizing Learned Reward Functions:…

200 papers

We consider the predict-then-optimize paradigm for decision-making in which a practitioner (1) trains a supervised learning model on historical data of decisions, contexts, and rewards, and then (2) uses the resulting model to make future…

Machine Learning · Computer Science 2024-06-13 Samuel Tan , Peter I. Frazier

Safe generalization in reinforcement learning requires not only that a learned policy acts capably in new situations, but also that it uses its capabilities towards the pursuit of the designer's intended goal. The latter requirement may…

Model-based Reinforcement Learning (MBRL) has been widely adapted due to its sample efficiency. However, existing worst-case regret analysis typically requires optimistic planning, which is not realistic in general. In contrast, motivated…

Machine Learning · Statistics 2023-01-10 Lingxiao Wang , Ping Li

The goal of inverse reinforcement learning (IRL) is to infer a reward function that explains the behavior of an agent performing a task. The assumption that most approaches make is that the demonstrated behavior is near-optimal. In many…

Machine Learning · Computer Science 2020-11-20 Luis Haug , Ivan Ovinnikov , Eugene Bykovets

We present an approach for the quantification of the usefulness of transfer in reinforcement learning via regret bounds for a multi-agent setting. Considering a number of $\aleph$ agents operating in the same Markov decision process,…

Machine Learning · Computer Science 2025-11-14 Adrienne Tuynman , Ronald Ortner

Optimization models used to make discrete decisions often contain uncertain parameters that are context-dependent and estimated through prediction. To account for the quality of the decision made based on the prediction, decision-focused…

Machine Learning · Computer Science 2024-07-30 Noah Schutte , Krzysztof Postek , Neil Yorke-Smith

State-of-the-art model-based reinforcement learning methods train policies on imagined rollouts. These rollouts are trajectories generated by a learned dynamics model and are scored by a learned reward model, but without querying the true…

Machine Learning · Computer Science 2026-05-13 Nadav Timor , Ravid Shwartz-Ziv , Micah Goldblum , Yann LeCun , David Harel

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

In safety-critical applications of reinforcement learning such as healthcare and robotics, it is often desirable to optimize risk-sensitive objectives that account for tail outcomes rather than expected reward. We prove the first regret…

Machine Learning · Computer Science 2022-10-12 O. Bastani , Y. J. Ma , E. Shen , W. Xu

Reinforcement learning (RL) systems typically optimize scalar reward functions that assume precise and reliable evaluation of outcomes. However, real-world objectives--especially those derived from human preferences--are often uncertain,…

Machine Learning · Computer Science 2026-04-30 Disha Singha

Learning policies via preference-based reward learning is an increasingly popular method for customizing agent behavior, but has been shown anecdotally to be prone to spurious correlations and reward hacking behaviors. While much prior work…

Machine Learning · Computer Science 2023-03-21 Jeremy Tien , Jerry Zhi-Yang He , Zackory Erickson , Anca D. Dragan , Daniel S. Brown

We consider undiscounted reinforcement learning in Markov decision processes (MDPs) where both the reward functions and the state-transition probabilities may vary (gradually or abruptly) over time. For this problem setting, we propose an…

Machine Learning · Computer Science 2019-09-11 Pratik Gajane , Ronald Ortner , Peter Auer

The theory of reinforcement learning currently suffers from a mismatch between its empirical performance and the theoretical characterization of its performance, with consequences for, e.g., the understanding of sample efficiency, safety,…

Machine Learning · Computer Science 2022-02-14 Feicheng Wang , Lucas Janson

Reward design is a fundamental problem in reinforcement learning (RL). A misspecified or poorly designed reward can result in low sample efficiency and undesired behaviors. In this paper, we propose the idea of programmatic reward design,…

Machine Learning · Computer Science 2022-01-10 Weichao Zhou , Wenchao Li

To design rewards that align with human goals, Reinforcement Learning from Human Feedback (RLHF) has emerged as a prominent technique for learning reward functions from human preferences and optimizing policies via reinforcement learning…

Machine Learning · Computer Science 2025-05-14 Taehyun Cho , Seokhun Ju , Seungyub Han , Dohyeong Kim , Kyungjae Lee , Jungwoo Lee

The correct specification of reward models is a well-known challenge in reinforcement learning. Hand-crafted reward functions often lead to inefficient or suboptimal policies and may not be aligned with user values. Reinforcement learning…

Artificial Intelligence · Computer Science 2024-10-24 Muhan Lin , Shuyang Shi , Yue Guo , Behdad Chalaki , Vaishnav Tadiparthi , Ehsan Moradi Pari , Simon Stepputtis , Joseph Campbell , Katia Sycara

Existing studies on constrained reinforcement learning (RL) may obtain a well-performing policy in the training environment. However, when deployed in a real environment, it may easily violate constraints that were originally satisfied…

Machine Learning · Computer Science 2024-05-06 Zhongchang Sun , Sihong He , Fei Miao , Shaofeng Zou

Inverse reinforcement learning (IRL) aims to infer an agent's preferences (represented as a reward function $R$) from their behaviour (represented as a policy $\pi$). To do this, we need a behavioural model of how $\pi$ relates to $R$. In…

Machine Learning · Computer Science 2024-03-12 Joar Skalse , Alessandro Abate

Reinforcement learning from human feedback (RLHF) is a widely used method for aligning large language models with human preferences. However, RLHF often suffers from reward model overoptimisation, in which models overfit to the reward…

Machine Learning · Computer Science 2025-09-30 Lorenz Wolf , Robert Kirk , Mirco Musolesi

Inferring reward functions from human behavior is at the center of value alignment - aligning AI objectives with what we, humans, actually want. But doing so relies on models of how humans behave given their objectives. After decades of…

Machine Learning · Computer Science 2023-10-31 Joey Hong , Kush Bhatia , Anca Dragan