Graph-based, Self-Supervised Program Repair from Diagnostic Feedback

2020-05-20 · ICML 2020 1 · Michihiro Yasunaga, Percy Liang

We consider the problem of learning to repair programs from diagnostic feedback (e.g., compiler error messages). Program repair is challenging for two reasons: First, it requires reasoning and tracking symbols across source code and diagnostic feedback. Second, labeled datasets available for program repair are relatively small. In this work, we propose novel solutions to these two challenges. First, we introduce a program-feedback graph, which connects symbols relevant to program repair in source code and diagnostic feedback, and then apply a graph neural network on top to model the reasoning process. Second, we present a self-supervised learning paradigm for program repair that leverages unlabeled programs available online to create a large amount of extra program repair examples, which we use to pre-train our models. We evaluate our proposed approach on two applications: correcting introductory programming assignments (DeepFix dataset) and correcting the outputs of program synthesis (SPoC dataset). Our final system, DrRepair, significantly outperforms prior work, achieving 68.2% full repair rate on DeepFix (+22.9% over the prior best), and 48.4% synthesis success rate on SPoC (+3.7% over the prior best).

📄 PDF Abstract BibTeX arXiv:2005.10636

Code (2)

michiyasunaga/DrRepair 공식 구현 pytorch

https://worksheets.codalab.org/worksheets/0x01838644724a433c932bef4cb5c42fbd 공식 구현

Tasks

Code GenerationDiagnosticGraph Learning Graph Neural Network Program Repair Program Synthesis Self-Supervised Learning

Methods 이 논문이 사용한 방법론

Graph Neural Network 설명 없음

Repair 설명 없음

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.

Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…

Position-Wise Feed-Forward Layer 설명 없음

Multi-Head Attention 설명 없음

Transformer A Transformer is a model architecture that eschews recurrence and instead relies entirely on an [attention…

GAT A Graph Attention Network (GAT) is a neural network architecture that operates on graph-structured data, leveraging masked self-attentional layers to address the shortcomings…

VIBEPASS: Can Vibe Coders Really Pass the Vibe Check?

2026-03-16 · Srijan Bansal, Jiao Fangkai, Yilun Zhou, Austin Xu 외 arxiv

As Large Language Models shift the programming toward human-guided ''vibe coding'', agentic coding tools increasingly rely on models to self-diagnose and repair their own subtle faults -- a capability central to autonomo…

Program Repair

SVRepair: Structured Visual Reasoning for Automated Program Repair

2026-02-05 · Xiaoxuan Tang, Jincheng Wang, Liwei Luo, Jingxuan Xu 외 arxiv

Large language models (LLMs) have recently shown strong potential for Automated Program Repair (APR), yet most existing approaches remain unimodal and fail to leverage the rich diagnostic signals contained in visual arti…

Visual Reasoning Program Repair

Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models

2026-06-30 · Jian Xu, Delu Zeng, John Paisley, Qibin Zhao arxiv

Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \emph{statistically} wrong -- a Gaussian likelihood for he…

SYNFIX: Automatically Fixing Syntax Errors using Compiler Diagnostics

2021-04-29 · Toufique Ahmed, Noah Rose Ledesma, Premkumar Devanbu

Beginning programmers struggle with the complex grammar of modern programming languages like Java, and make lot of syntax errors. The diagnostic syntax error messages from compilers and IDEs are sometimes useful, but oft…

BIG-bench Machine Learning Diagnostic Multi-Label Classification MUlTI-LABEL-ClASSIFICATION+1

Self-Supervised Bug Detection and Repair

2021-05-26 · NeurIPS 2021 12 · Miltiadis Allamanis, Henry Jackson-Flux, Marc Brockschmidt

Machine learning-based program analyses have recently shown the promise of integrating formal and probabilistic reasoning towards aiding software development. However, in the absence of large annotated corpora, training …

Self-Supervised Learning