Executable Code Actions Elicit Better LLM Agents

This work proposes to use executable Python code to consolidate LLM agents' actions into a unified action space (CodeAct). (Abstract)

CodeActAgent, finetuned from Llama2 and Mistral, is integrated with Python interpreter and uniquely tailored to perform sophisticated tasks (e.g., model training) using existing libraries and autonomously self-debug.

We implemented a containerized code execution engine based on JupyterKernelGateway.