Reproduce “Fault-Free Profiles and Decision-Level Training for Root Cause Analysis via Small Language Models” 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
Kyoohyun Lee, Woohyeok Park, Taeyoon Kim, Sungyong Baik, Kyungyong Lee
The authors develop a microservice root-cause analysis system that uses fault-free anomaly frequencies and decision-level preference training to improve small language model diagnosis. Historical Frequency Profiling contextualizes metric and log anomalies, while Biased-Decision Correction trains choices among competing continuations. On Bank, Qwen3-4B achieves 48.7% Exact versus 18.5% for OpsAgent-14B; results are specific to evaluated systems.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.