Towards Autonomous Agents: Adaptive-planning, Reasoning, and Acting in Language Models

TL;DR

This paper introduces a novel in-context learning algorithm that enables language models to act autonomously by self-correcting in task execution, demonstrated through success in a text-based game environment.

Contribution

The paper presents a new self-correcting in-context learning method that improves autonomous decision-making in language models for task solving.

Findings

Gemma-2-9b-it model successfully completed 2 out of 6 previously failed tasks.

Self-correction significantly enhances problem-solving capabilities of language models.

The approach demonstrates potential for developing more advanced autonomous language agents.

Abstract

We propose a novel in-context learning algorithm for building autonomous decision-making language agents. The language agent continuously attempts to solve the same task by self-correcting each time the task fails. Our selected language agent demonstrates the ability to solve tasks in a text-based game environment. Our results show that the gemma-2-9b-it language model, using our proposed method, can successfully complete two of six tasks that failed in the first attempt. This highlights the effectiveness of our approach in enhancing the problem-solving capabilities of a single language model through self-correction, paving the way for more advanced autonomous agents. The code is publicly available at https://github.com/YenCheHsiao/AutonomousLLMAgentwithAdaptingPlanning.