Query-Policy Misalignment in Preference-Based Reinforcement Learning

• URL: http://arxiv.org/abs/2305.17400v3
• Date: Fri, 5 Jul 2024 14:26:21 GMT
• Title: Query-Policy Misalignment in Preference-Based Reinforcement Learning
• Authors: Xiao Hu, Jianxiong Li, Xianyuan Zhan, Qing-Shan Jia, Ya-Qin Zhang,
• Abstract summary: We show that the seemingly informative queries selected to improve the overall quality of reward model actually may not align with RL agents' interests. We show that this issue can be effectively addressed via near on-policy query and a specially designed hybrid experience replay. Our method achieves substantial gains in both human feedback and RL sample efficiency.
• Score: 21.212703100030478
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Preference-based reinforcement learning (PbRL) provides a natural way to align RL agents' behavior with human desired outcomes, but is often restrained by costly human feedback. To improve feedback efficiency, most existing PbRL methods focus on selecting queries to maximally improve the overall quality of the reward model, but counter-intuitively, we find that this may not necessarily lead to improved performance. To unravel this mystery, we identify a long-neglected issue in the query selection schemes of existing PbRL studies: Query-Policy Misalignment. We show that the seemingly informative queries selected to improve the overall quality of reward model actually may not align with RL agents' interests, thus offering little help on policy learning and eventually resulting in poor feedback efficiency. We show that this issue can be effectively addressed via near on-policy query and a specially designed hybrid experience replay, which together enforce the bidirectional query-policy alignment. Simple yet elegant, our method can be easily incorporated into existing approaches by changing only a few lines of code. We showcase in comprehensive experiments that our method achieves substantial gains in both human feedback and RL sample efficiency, demonstrating the importance of addressing query-policy misalignment in PbRL tasks.

Related papers

• Reward-Augmented Data Enhances Direct Preference Alignment of LLMs [56.24431208419858]
  We introduce reward-conditioned Large Language Models (LLMs) that learn from the entire spectrum of response quality within the dataset. We propose an effective yet simple data relabeling method that conditions the preference pairs on quality scores to construct a reward-augmented dataset.
  arXiv  Detail & Related papers  (2024-10-10T16:01:51Z)
• Preference-Guided Reinforcement Learning for Efficient Exploration [7.83845308102632]
  We introduce LOPE: Learning Online with trajectory Preference guidancE, an end-to-end preference-guided RL framework. Our intuition is that LOPE directly adjusts the focus of online exploration by considering human feedback as guidance. LOPE outperforms several state-of-the-art methods regarding convergence rate and overall performance.
  arXiv  Detail & Related papers  (2024-07-09T02:11:12Z)
• Preference Elicitation for Offline Reinforcement Learning [59.136381500967744]
  We propose Sim-OPRL, an offline preference-based reinforcement learning algorithm. Our algorithm employs a pessimistic approach for out-of-distribution data, and an optimistic approach for acquiring informative preferences about the optimal policy.
  arXiv  Detail & Related papers  (2024-06-26T15:59:13Z)
• Joint Demonstration and Preference Learning Improves Policy Alignment with Human Feedback [58.049113055986375]
  We develop a single stage approach named Alignment with Integrated Human Feedback (AIHF) to train reward models and the policy. The proposed approach admits a suite of efficient algorithms, which can easily reduce to, and leverage, popular alignment algorithms. We demonstrate the efficiency of the proposed solutions with extensive experiments involving alignment problems in LLMs and robotic control problems in MuJoCo.
  arXiv  Detail & Related papers  (2024-06-11T01:20:53Z)
• LIRE: listwise reward enhancement for preference alignment [27.50204023448716]
  We propose a gradient-based reward optimization approach that incorporates the offline rewards of multiple responses into a streamlined listwise framework. LIRE is straightforward to implement, requiring minimal parameter tuning, and seamlessly aligns with the pairwise paradigm. Our experiments demonstrate that LIRE consistently outperforms existing methods across several benchmarks on dialogue and summarization tasks.
  arXiv  Detail & Related papers  (2024-05-22T10:21:50Z)
• PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback [106.63518036538163]
  We present a novel unified bilevel optimization-based framework, textsfPARL, formulated to address the recently highlighted critical issue of policy alignment in reinforcement learning. Our framework addressed these concerns by explicitly parameterizing the distribution of the upper alignment objective (reward design) by the lower optimal variable. Our empirical results substantiate that the proposed textsfPARL can address the alignment concerns in RL by showing significant improvements.
  arXiv  Detail & Related papers  (2023-08-03T18:03:44Z)
• Sample-Efficient, Exploration-Based Policy Optimisation for Routing Problems [2.6782615615913348]
  This paper presents a new reinforcement learning approach that is based on entropy. In addition, we design an off-policy-based reinforcement learning technique that maximises the expected return. We show that our model can generalise to various route problems.
  arXiv  Detail & Related papers  (2022-05-31T09:51:48Z)
• Reward Uncertainty for Exploration in Preference-based Reinforcement Learning [88.34958680436552]
  We present an exploration method specifically for preference-based reinforcement learning algorithms. Our main idea is to design an intrinsic reward by measuring the novelty based on learned reward. Our experiments show that exploration bonus from uncertainty in learned reward improves both feedback- and sample-efficiency of preference-based RL algorithms.
  arXiv  Detail & Related papers  (2022-05-24T23:22:10Z)
• Information Directed Reward Learning for Reinforcement Learning [64.33774245655401]
  We learn a model of the reward function that allows standard RL algorithms to achieve high expected return with as few expert queries as possible. In contrast to prior active reward learning methods designed for specific types of queries, IDRL naturally accommodates different query types. We support our findings with extensive evaluations in multiple environments and with different types of queries.
  arXiv  Detail & Related papers  (2021-02-24T18:46:42Z)
• On the Search for Feedback in Reinforcement Learning [6.29295842374861]
  We advocate searching over a local feedback representation consisting of an open-loop sequence, and an associated optimal linear feedback law completely determined by the open-loop. We show that this alternate approach results in highly efficient training, the answers obtained are repeatable and reliable, and the resulting closed performance is superior to global state-of-the-art RL techniques.
  arXiv  Detail & Related papers  (2020-02-21T02:44:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.