ResearchBenchmarksFeasible Certification on AstrophysicsFollow9Certified Success ScorePHACT8.558.77599.225Jun 29, 2026Evaluation ResultsMethodMethodLinksCertified Success ScoreLoop Fail CountWrong Verdict CountTimeout CountPHACTModel=Gemini 2.5 Flash...Model=Gemini 2.5 Flash, Temperature=02026.069100