Reproduce “Discovering Performance Archetypes: Critical-Path-Aware Pattern Analysis and Regression Detection” 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
Kaveh Shahedi, Heng Li, Maxime Lamothe, Foutse Khomh
The authors analyze dynamic call paths and resource behavior to identify recurring performance patterns and detect regressions. They cluster 79,789 paths from six C/C++ applications into 13 archetypes, five appearing in at least five programs, and combine path, resource, archetype, and performance-bound signals. The detector reaches F1 0.867 on synthetic regressions; real-world regressions and production use remain untested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.