We introduce the Principled Reasoning and Acting (PRAct) framework, a novel method for learning and enforcing action principles from trajectory data. Central to our approach is the use of text gradients from a reflection and optimization engine to derive these action principles. To adapt action principles to specific task requirements, we propose a new optimization framework, Reflective Principle Optimization (RPO). After execution, RPO employs a reflector to critique current action principles and an optimizer to update them accordingly. We develop the RPO framework under two scenarios: Reward-RPO, which uses environmental rewards for reflection, and Self-RPO, which conducts self-reflection without external rewards. Additionally, two RPO methods, RPO-Traj and RPO-Batch, is introduced to adapt to different settings. Experimental results across four environments demonstrate that the PRAct agent, leveraging the RPO framework, effectively learns and applies action principles to enhance performance.
@article{arxiv.2410.18528,
title = {PRACT: Optimizing Principled Reasoning and Acting of LLM Agent},
author = {Zhiwei Liu and Weiran Yao and Jianguo Zhang and Rithesh Murthy and Liangwei Yang and Zuxin Liu and Tian Lan and Ming Zhu and Juntao Tan and Shirley Kokane and Thai Hoang and Juan Carlos Niebles and Shelby Heinecke and Huan Wang and Silvio Savarese and Caiming Xiong},
journal= {arXiv preprint arXiv:2410.18528},
year = {2024}
}