ResearchBenchmarksTable Understanding on OOD Table S1 (test)Follow80.2AccuracyGPT-452.53659.71866.974.082Jan 24, 2025Evaluation ResultsMethodMethodLinksAccuracyGPT-42025.0180.2TAMA2025.0164.93GPT-3.52025.0154.8base2025.0153.6