LangProp: A code optimization framework using Large Language Models applied to driving

TL;DR

LangProp is a novel framework that iteratively improves code generated by large language models through automatic evaluation and feedback, applicable to diverse domains including autonomous driving, enhancing code quality and robustness.

Contribution

It introduces LangProp, a new iterative optimization framework for LLM-generated code, integrating evaluation, feedback, and learning techniques for improved performance.

Findings

Effective in Sudoku and CartPole domains

First proof of concept for autonomous driving code optimization

Produces interpretable and verifiable policies

Abstract

We propose LangProp, a framework for iteratively optimizing code generated by large language models (LLMs), in both supervised and reinforcement learning settings. While LLMs can generate sensible coding solutions zero-shot, they are often sub-optimal. Especially for code generation tasks, it is likely that the initial code will fail on certain edge cases. LangProp automatically evaluates the code performance on a dataset of input-output pairs, catches any exceptions, and feeds the results back to the LLM in the training loop, so that the LLM can iteratively improve the code it generates. By adopting a metric- and data-driven training paradigm for this code optimization procedure, one could easily adapt findings from traditional machine learning techniques such as imitation learning, DAgger, and reinforcement learning. We show LangProp's applicability to general domains such as Sudoku and CartPole, as well as demonstrate the first proof of concept of automated code optimization for autonomous driving in CARLA. We show that LangProp can generate interpretable and transparent policies that can be verified and improved in a metric- and data-driven way. Our code is available at https://github.com/shuishida/LangProp.