Pref-GUIDE: Continual Policy Learning from Real-Time Human Feedback via Preference-Based Learning
This addresses the challenge of online reinforcement learning where human feedback is needed but difficult to specify, offering a scalable approach for continual policy adaptation, though it is incremental as it builds on prior preference-based methods.
The paper tackled the problem of training reinforcement learning agents with real-time human scalar feedback, which is often noisy and inconsistent, by proposing Pref-GUIDE to transform such feedback into preference-based data for improved reward model learning, resulting in significant outperformance over baselines across three challenging environments, with the voting variant exceeding expert-designed dense rewards.
Training reinforcement learning agents with human feedback is crucial when task objectives are difficult to specify through dense reward functions. While prior methods rely on offline trajectory comparisons to elicit human preferences, such data is unavailable in online learning scenarios where agents must adapt on the fly. Recent approaches address this by collecting real-time scalar feedback to guide agent behavior and train reward models for continued learning after human feedback becomes unavailable. However, scalar feedback is often noisy and inconsistent, limiting the accuracy and generalization of learned rewards. We propose Pref-GUIDE, a framework that transforms real-time scalar feedback into preference-based data to improve reward model learning for continual policy training. Pref-GUIDE Individual mitigates temporal inconsistency by comparing agent behaviors within short windows and filtering ambiguous feedback. Pref-GUIDE Voting further enhances robustness by aggregating reward models across a population of users to form consensus preferences. Across three challenging environments, Pref-GUIDE significantly outperforms scalar-feedback baselines, with the voting variant exceeding even expert-designed dense rewards. By reframing scalar feedback as structured preferences with population feedback, Pref-GUIDE offers a scalable and principled approach for harnessing human input in online reinforcement learning.