Reproduce “Causality-Based Fault Localization” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Melane Navaratnarajah, Aidan Dakhama, Tom Yaacov, David A. Kelly, Karine Even-Mendoza, Hana Chockler
SoftReX ranks faulty code locations using sampled code mutations, measuring how they change failing tests while penalizing disruption to passing tests and large semantic changes. On a shared mutant pool, its full configuration achieved MRR 0.349 versus BLMu’s 0.304 and ranked 103 bugs first versus 89. This result reflects combined scoring, not causal scoring alone.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.