读paper16-微调LLM的APR与正确性验证

RepairCAT: Applying Large Language Model to Fix Bugs in AI-Generated Programs

从数据集构造到模型微调。使用微调后的LLM生成数据集进行大模型微调

数据集不进行缺陷定位,而是将整个有问题的程序一并处理,让模型决定修复的位置。

https://github.com/nus-apr/cerberus 。一个研究加速框架,它提供了多种先进程序分析工具(如 Infer 和 Pulse)、模糊测试工具(如 AFL++、Jazzer)以及程序修复工具(如 F1X、SelfAPR 等)的接口

Improving Patch Correctness Analysis via Random Testing and Large Language Models

https://ieeexplore.ieee.org/document/10638611

Accelerating Patch Validation for Program Repair With Interception-Based Execution Scheduling

https://ieeexplore.ieee.org/document/10417068

APPT: Boosting Automated Patch Correctness Prediction via Fine-Tuning Pre-Trained Models

https://ieeexplore.ieee.org/document/10402095