Pith. sign in

REVIEW 18 cited by

RepairLLaMA: Efficient Representations and Fine-Tuned Adapters for Program Repair

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2312.15698 v6 pith:MUGOHFLV submitted 2023-12-25 cs.SE cs.LG

André Silva , Sen Fang , Martin Monperrus This is my paper · ORCID
classification cs.SEcs.LG
keywords repairprogramfine-tuningrepairllamarepresentationsbugscodellms
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Automated Program Repair (APR) has evolved significantly with the advent of Large Language Models (LLMs). Fine-tuning LLMs for program repair is a recent avenue of research, with many dimensions which have not been explored. Existing work mostly fine-tune LLMs with naive code representations and does not scale to frontier models. To address this problem, we propose RepairLLaMA, a novel program repair approach that 1) identifies optimal code representations for APR with fine-tuned models, and 2) pioneers state-of-the-art parameter-efficient fine-tuning technique (PEFT) for program repair. This results in RepairLLaMA producing a highly effective `program repair adapter' for fixing bugs with AI. Our experiments demonstrate the validity of both concepts. First, fine-tuning adapters with program repair specific code representations enables the model to use meaningful repair signals and produce better patches. Second, parameter-efficient fine-tuning helps fine-tuning to converge and clearly contributes to the effectiveness of RepairLLaMA in fixing bugs outside the fine-tuning data distribution. Overall, RepairLLaMA correctly fixes 144 Defects4J v2, 109 HumanEval-Java, and 20 GitBug-Java bugs, outperforming all baselines.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 18 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Balancing Usefulness and Naturalness: An LLM-based Curation Pipeline for Code Review Comments

    cs.SE 2026-07 conditional novelty 6.0 of 10

    Selective LLM reformulation guided by high-quality exemplars yields cleaner, more diverse code-review datasets that improve downstream comment generation and code refinement.

  2. Empirical Study of Code Large Language Models for Binary Security Patch Detection

    cs.SE 2025-09 conditional novelty 6.0 of 10

    Fine-tuned code LLMs detect binary security patches accurately on decompiler pseudo-code, and adding source-code patches to the training data improves accuracy and F1 further.

  3. Guiding AI to Fix Its Own Flaws: An Empirical Study on LLM-Driven Secure Code Generation

    cs.SE 2025-06 conditional novelty 6.0 of 10

    Across two Python security benchmarks, all eight LLMs produced vulnerable code at rates from 9.8% to 42.1%, and self-generated hints plus explained feedback reduced vulnerabilities mainly for the strongest models.

  4. Seeing is Fixing: Cross-Modal Reasoning with Multimodal LLMs for Visual Software Issue Fixing

    cs.SE 2025-06 conditional novelty 6.0 of 10

    GUIRepair, a cross-modal LLM pipeline that converts issue screenshots into reproduction code and rendered patch screenshots into validation feedback, resolves 157/517 SWE-bench M instances with GPT-4o and 175 with o4-mini.

  5. From Empirical Evaluation to Context-Aware Enhancement: Repairing Regression Errors with LLMs

    cs.SE 2025-06 conditional novelty 6.0 of 10

    Adding bug-inducing commit context to LLM prompts raises correct regression repairs from 9 to 16 of 99 Java bugs (1.8x).

  6. Larger Is Not Always Better: Exploring Small Open-source Language Models in Logging Statement Generation

    cs.SE 2025-05 conditional novelty 6.0 of 10

    A fine-tuned 14B small open-source model with LoRA and RAG outperforms larger proprietary LLMs on automated Java logging statement generation in AL-Bench point estimates.

  7. PATCH: Empowering Large Language Model with Programmer-Intent Guidance and Collaborative-Behavior Simulation for Automatic Bug Fixing

    cs.SE 2025-01 conditional novelty 6.0 of 10

    A role-playing, multi-agent LLM repair pipeline with commit-message intent guidance reaches 33.97% Fix@1 on BFP, beating GPT-4's 19.96%.

  8. When Fine-Tuning LLMs Meets Data Privacy: An Empirical Study of Federated Learning in LLM-Based Program Repair

    cs.SE 2024-12 conditional novelty 6.0 of 10

    Federated fine-tuning of six code LLMs on private bug-fix data improves program repair to near-centralized levels, with negligible impact from heterogeneous code.

  9. GP-Tree: An in-memory spatial index combining adaptive grid cells with a prefix tree for efficient spatial querying

    cs.DB 2026-03 unverdicted novelty 5.0 of 10

    GP-Tree indexes spatial objects via adaptive grid-cell approximations in a prefix tree and claims up to 10× faster range, distance, and k-NN queries than STR-Tree and Quad-Tree.

  10. The Impact of Fine-tuning Large Language Models on Automated Program Repair

    cs.SE 2025-07 conditional novelty 5.0 of 10

    On three Java APR benchmarks, LoRA and IA3 adapters match or beat full-model fine-tuning for most tested code LLMs while training less than one percent of parameters.

  11. Empirical Evaluation of Large Language Models in Automated Program Repair

    cs.SE 2025-06 conditional novelty 5.0 of 10

    An empirical study of four open-source LLMs across six benchmarks shows code-specialized models often outperform larger general models, and most correct repairs appear early in generation.

  12. The Art of Repair: Optimizing Iterative Program Repair with Instruction-Tuned Models

    cs.SE 2025-05 conditional novelty 5.0 of 10

    Using only 1,000 fine-tuning examples with full fine-tuning improves the number of plausible repairs by up to 78%, and iterative generation strategies help base models most on complex benchmarks like Defects4J.

  13. Distilling Desired Comments for Enhanced Code Review with Large Language Models

    cs.SE 2024-12 conditional novelty 5.0 of 10

    Desiview identifies desired review comments from code review datasets using the perplexity difference of the actual fix with and without each comment, and the distilled data improves LLaMA-based code review models.

  14. Are Large Language Models Memorizing Bug Benchmarks?

    cs.SE 2024-11 conditional novelty 5.0 of 10

    Some base LLMs, particularly codegen-multi, show strong memorization of the Defects4J bug benchmark, while newer models like LLaMa 3.1 show weaker leakage signals.

  15. Accelerating Automatic Program Repair with Dual Retrieval-Augmented Fine-Tuning and Patch Generation on Large Language Models

    cs.SE 2025-07 conditional novelty 4.0 of 10

    A dual-retrieval, threshold-gated, full-parameter fine-tuned code LLM (SelRepair) reports state-of-the-art exact-match patch generation on Tufano and VulRepair benchmarks.

  16. Llama-3.1-FoundationAI-SecurityLLM-Base-8B Technical Report

    cs.CR 2025-04 reject novelty 4.0 of 10

    A continued-pretrained 8B cybersecurity LLM claims to match GPT-4o-mini and Llama 3.1-70B on certain cyber threat intelligence benchmarks, but the decisive benchmark overlaps with its training corpus.

  17. Harnessing Large Language Models for Curated Code Reviews

    cs.SE 2025-02 conditional novelty 4.0 of 10

    An LLM-based curation pipeline improves code review dataset quality and boosts downstream comment generation and code refinement performance.

  18. LLM4CVE: Enabling Iterative Automated Vulnerability Repair with Large Language Models

    cs.SE 2025-01 conditional novelty 4.0 of 10

    An iterative LLM pipeline with LoRA fine-tuning and CodeBLEU-based feedback improves automated repair of real-world CVE functions, reaching a human-validated correctness score of 8.51/10 with Llama 3 70B.

Pith tools