ResearchBenchmarksRegex Induction on Simple Regex SD=1 (test)Follow100Run-Level Success RateAgentic48.83262.11675.488.684Jun 9, 2026Evaluation ResultsMethodMethodLinksRun-Level Success RateAgenticSetup=Full proposed fr...Setup=Full proposed framework2026.06100SingleSetup=Clustered counte...Setup=Clustered counterexamples2026.0692.1StandardSetup=No counterexamplesSetup=No counterexamples2026.0650.8