ResearchBenchmarksRegex Induction on Extended Regex SD=2 (test)Follow74.1Run-Level Success RateAgentic37.49246.99656.566.004Jun 9, 2026Evaluation ResultsMethodMethodLinksRun-Level Success RateAgenticSetup=Full proposed fr...Setup=Full proposed framework2026.0674.1SingleSetup=Clustered counte...Setup=Clustered counterexamples2026.0668.5StandardSetup=No counterexamplesSetup=No counterexamples2026.0638.9