A Metamorphic Testing Approach to Diagnosing Memorization in LLM-Based Program Repair

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: De Koning, Milan, Asgari, Ali, Derakhshanfar, Pouria, Panichella, Annibale
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911618426732544
author De Koning, Milan
Asgari, Ali
Derakhshanfar, Pouria
Panichella, Annibale
author_facet De Koning, Milan
Asgari, Ali
Derakhshanfar, Pouria
Panichella, Annibale
contents LLM-based automated program repair (APR) techniques have shown promising results in reducing debugging costs. However, prior results can be affected by data leakage: large language models (LLMs) may memorize bug fixes when evaluation benchmarks overlap with their pretraining data, leading to inflated performance estimates. In this paper, we investigate whether we can better reveal data leakage by combining metamorphic testing (MT) with negative log-likelihood (NLL), which has been used in prior work as a proxy for memorization. We construct variant benchmarks by applying semantics-preserving transformations to two widely used datasets, Defects4J and GitBug-Java. Using these benchmarks, we evaluate the repair success rates of seven LLMs on both original and transformed versions, and analyze the relationship between performance degradation and NLL. Our results show that all evaluated state-of-the-art LLMs exhibit substantial drops in patch generation success rates on transformed benchmarks, ranging from -4.1% for GPT-4o to -15.98% for Llama-3.1. Furthermore, we find that this degradation strongly correlates with NLL on the original benchmarks, suggesting that models perform better on instances they are more likely to have memorized. These findings show that combining MT with NLL provides stronger and more reliable evidence of data leakage, while metamorphic testing alone can help mitigate its effects in LLM-based APR evaluations.
format Preprint
id arxiv_https___arxiv_org_abs_2604_21579
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle A Metamorphic Testing Approach to Diagnosing Memorization in LLM-Based Program Repair
De Koning, Milan
Asgari, Ali
Derakhshanfar, Pouria
Panichella, Annibale
Software Engineering
Artificial Intelligence
LLM-based automated program repair (APR) techniques have shown promising results in reducing debugging costs. However, prior results can be affected by data leakage: large language models (LLMs) may memorize bug fixes when evaluation benchmarks overlap with their pretraining data, leading to inflated performance estimates. In this paper, we investigate whether we can better reveal data leakage by combining metamorphic testing (MT) with negative log-likelihood (NLL), which has been used in prior work as a proxy for memorization. We construct variant benchmarks by applying semantics-preserving transformations to two widely used datasets, Defects4J and GitBug-Java. Using these benchmarks, we evaluate the repair success rates of seven LLMs on both original and transformed versions, and analyze the relationship between performance degradation and NLL. Our results show that all evaluated state-of-the-art LLMs exhibit substantial drops in patch generation success rates on transformed benchmarks, ranging from -4.1% for GPT-4o to -15.98% for Llama-3.1. Furthermore, we find that this degradation strongly correlates with NLL on the original benchmarks, suggesting that models perform better on instances they are more likely to have memorized. These findings show that combining MT with NLL provides stronger and more reliable evidence of data leakage, while metamorphic testing alone can help mitigate its effects in LLM-based APR evaluations.
title A Metamorphic Testing Approach to Diagnosing Memorization in LLM-Based Program Repair
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2604.21579