MétaCan
Menu
Back to cohort

Making the Case for LLM-Generated Automated Program Repair Benchmarks

2025· article· W4416799210 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

Venuenot available
Typearticle
Language
FieldComputer Science
TopicSoftware Testing and Debugging Techniques
Canadian institutionsAlgoma University
Fundersnot available
KeywordsOverfittingBenchmark (surveying)BenchmarkingFocus (optics)Quality (philosophy)Quality assurance

Abstract

fetched live from OpenAlex

Automated Program Repair (APR) has made significant strides in recent years, particularly with the integration of large language models (LLMs) and deep learning techniques. Yet despite this progress, one fundamental issue continues to hinder advancement: how we evaluate these systems. Many of today’s APR benchmarks suffer from serious limitations—including small dataset sizes, synthetic or unrealistic bug scenarios, overfitting risks, ambiguous evaluation criteria, and a narrow focus on certain programming languages.In this paper, we take a critical look at these challenges by identifying eight core limitations in widely used benchmarks. We then explore how LLM-generated benchmarks can help overcome these obstacles. Finally, we address some potential concerns about LLM-generated benchmarks and propose a quality assurance and validation framework.By combining the strengths of LLMs with thoughtful benchmark design, this work lays the foundation for more robust, diverse, and meaningful evaluation frameworks—paving the way for future breakthroughs in APR research.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.002
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Science and technology studies, Scholarly communication
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.972
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0020.002
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.002
Science and technology studies0.0020.000
Scholarly communication0.0010.000
Open science0.0010.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.049
GPT teacher head0.367
Teacher spread0.318 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it