Defects-4J V-2.0

'Defects-4J V-2.0' is a benchmark dataset that contains a collection of real-world Java programs with known defects, used to evaluate automated program repair techniques.

Papers using Defects-4J V-2.0 (8)

PracRepair: LLM-Empowered Automated Program Repair Inspired by Human-Like Debugging Practices2026Specification Vibing for Automated Program Repair2026Repair Ingredients Are All You Need: Improving Large Language Model-Based Program Repair via Repair Ingredients Search2025Integrating Various Software Artifacts for Better LLM-based Bug Localization and Program Repair2024A Syntax-Guided Edit Decoder for Neural Program Repair2021 · 196 citesKNOD: Domain Knowledge Distilled Tree Decoder for Automated Program Repair2023 · 55 citesGAMMA: Revisiting Template-based Automated Program Repair via Mask Prediction2023 · 47 citesHybrid Automated Program Repair by Combining Large Language Models and Program Analysis2024