O3D: Offline Data-driven Discovery and Distillation for Sequential Decision-Making with Large Language Models
This work addresses the challenge of scaling LLM agents for complex, long-horizon tasks in interactive environments, offering an incremental improvement by leveraging offline data without finetuning.
The paper tackles the problem of insufficient few-shot examples and limited context length for LLMs in complex sequential decision-making tasks by proposing O3D, an offline learning framework that uses offline data to discover skills and distill knowledge, resulting in notable performance improvements over baselines on benchmarks like ALFWorld and WebShop.
Recent advancements in large language models (LLMs) have exhibited promising performance in solving sequential decision-making problems. By imitating few-shot examples provided in the prompts (i.e., in-context learning), an LLM agent can interact with an external environment and complete given tasks without additional training. However, such few-shot examples are often insufficient to generate high-quality solutions for complex and long-horizon tasks, while the limited context length cannot consume larger-scale demonstrations with long interaction horizons. To this end, we propose an offline learning framework that utilizes offline data at scale (e.g, logs of human interactions) to improve LLM-powered policies without finetuning. The proposed method O3D (Offline Data-driven Discovery and Distillation) automatically discovers reusable skills and distills generalizable knowledge across multiple tasks based on offline interaction data, advancing the capability of solving downstream tasks. Empirical results under two interactive decision-making benchmarks (ALFWorld and WebShop) verify that O3D can notably enhance the decision-making capabilities of LLMs through the offline discovery and distillation process, and consistently outperform baselines across various LLMs.