Towards Effectively Leveraging Execution Traces for Program Repair with Code LLMs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Haque, Mirazul, Babkin, Petr, Farmahinifarahani, Farima, Veloso, Manuela
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915278333411328
author Haque, Mirazul
Babkin, Petr
Farmahinifarahani, Farima
Veloso, Manuela
author_facet Haque, Mirazul
Babkin, Petr
Farmahinifarahani, Farima
Veloso, Manuela
contents Large Language Models (LLMs) show promising performance on various programming tasks, including Automatic Program Repair (APR). However, most approaches to LLM-based APR are limited to the static analysis of the programs, while disregarding their runtime behavior. Inspired by knowledge-augmented NLP, in this work, we aim to remedy this potential blind spot by augmenting standard APR prompts with program execution traces. We evaluate our approach using the GPT family of models on three popular APR datasets. Our findings suggest that simply incorporating execution traces into the prompt provides a limited performance improvement over trace-free baselines, in only 2 out of 6 tested dataset / model configurations. We further find that the effectiveness of execution traces for APR diminishes as their complexity increases. We explore several strategies for leveraging traces in prompts and demonstrate that LLM-optimized prompts help outperform trace-free prompts more consistently. Additionally, we show trace-based prompting to be superior to finetuning a smaller LLM on a small-scale dataset; and conduct probing studies reinforcing the notion that execution traces can complement the reasoning abilities of the LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2505_04441
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Effectively Leveraging Execution Traces for Program Repair with Code LLMs
Haque, Mirazul
Babkin, Petr
Farmahinifarahani, Farima
Veloso, Manuela
Machine Learning
Software Engineering
Large Language Models (LLMs) show promising performance on various programming tasks, including Automatic Program Repair (APR). However, most approaches to LLM-based APR are limited to the static analysis of the programs, while disregarding their runtime behavior. Inspired by knowledge-augmented NLP, in this work, we aim to remedy this potential blind spot by augmenting standard APR prompts with program execution traces. We evaluate our approach using the GPT family of models on three popular APR datasets. Our findings suggest that simply incorporating execution traces into the prompt provides a limited performance improvement over trace-free baselines, in only 2 out of 6 tested dataset / model configurations. We further find that the effectiveness of execution traces for APR diminishes as their complexity increases. We explore several strategies for leveraging traces in prompts and demonstrate that LLM-optimized prompts help outperform trace-free prompts more consistently. Additionally, we show trace-based prompting to be superior to finetuning a smaller LLM on a small-scale dataset; and conduct probing studies reinforcing the notion that execution traces can complement the reasoning abilities of the LLMs.
title Towards Effectively Leveraging Execution Traces for Program Repair with Code LLMs
topic Machine Learning
Software Engineering
url https://arxiv.org/abs/2505.04441