← back to paper
arxiv: 2605.02215 · 2 revisions
HEJ-Robust: A Robustness Benchmark for LLM-Based Automated Program Repair