An Offline Adaptation Framework for Constrained Multi-Objective Reinforcement Learning
This work addresses the need for automated preference inference in multi-objective RL, reducing reliance on human prior knowledge, though it is incremental as it builds on existing offline and multi-objective RL methods.
The paper tackles the problem of multi-objective reinforcement learning requiring handcrafted preferences by proposing an offline adaptation framework that infers preferences from demonstrations, and it shows the framework can handle safety constraints with unknown thresholds, achieving policies that align with real preferences in empirical tests.
In recent years, significant progress has been made in multi-objective reinforcement learning (RL) research, which aims to balance multiple objectives by incorporating preferences for each objective. In most existing studies, specific preferences must be provided during deployment to indicate the desired policies explicitly. However, designing these preferences depends heavily on human prior knowledge, which is typically obtained through extensive observation of high-performing demonstrations with expected behaviors. In this work, we propose a simple yet effective offline adaptation framework for multi-objective RL problems without assuming handcrafted target preferences, but only given several demonstrations to implicitly indicate the preferences of expected policies. Additionally, we demonstrate that our framework can naturally be extended to meet constraints on safety-critical objectives by utilizing safe demonstrations, even when the safety thresholds are unknown. Empirical results on offline multi-objective and safe tasks demonstrate the capability of our framework to infer policies that align with real preferences while meeting the constraints implied by the provided demonstrations.