Bridging Values and Behavior: A Hierarchical Framework for Proactive Embodied Agents

TL;DR

This paper presents ValuePlanner, a hierarchical framework for embodied agents that integrates high-level value reasoning with low-level action planning, enabling autonomous, long-term, self-directed behavior.

Contribution

It introduces a novel hierarchical architecture using LLMs and classical planners to bridge values and grounded actions, advancing autonomous agent capabilities.

Findings

ValuePlanner generates coherent, long-horizon behaviors in household tasks.

It effectively arbitrates between competing values to produce self-directed actions.

The framework outperforms instruction-following and needs-driven baselines in experiments.

Abstract

Current embodied agents are often limited to passive instruction-following or reactive need-satisfaction, lacking a stable, high-order value framework essential for long-term, self-directed behavior and resolving motivational conflicts. We introduce \textit{ValuePlanner}, a hierarchical cognitive architecture that decouples high-level value scheduling from low-level action execution. \textit{ValuePlanner} employs an LLM-based cognitive module to generate symbolic subgoals by reasoning through abstract value trade-offs, which are then translated into executable action plans by a classical PDDL planner. This process is refined via a closed-loop feedback mechanism. Evaluating such autonomy requires methods beyond task-success rates, and we therefore propose a value-centric evaluation suite measuring cumulative value gain, preference alignment, and behavioral diversity. Experiments in the TongSim household environment demonstrate that \textit{ValuePlanner} arbitrates competing values to generate coherent, long-horizon, self-directed behavior absent from instruction-following and needs-driven baselines. Our work offers a structured approach to bridging intrinsic values and grounded behavior for autonomous agents.