ResearchBenchmarksTabular Classification on MLGym TitanicFollow99.8AccuracyARTS*82.32886.86491.495.936Jun 20, 2026Evaluation ResultsMethodMethodLinksAccuracyARTS*Base Model=Qwen 4B, Te...Base Model=Qwen 4B, Test-time trained=true2026.0699.8ARTSBase Model=o3Base Model=o32026.0698.4LinearMethod Type=Prior WorksMethod Type=Prior Works2026.0695.1ARTSBase Model=Qwen 4BBase Model=Qwen 4B2026.0694.9MLEvolveMethod Type=Prior WorksMethod Type=Prior Works2026.0694.6AIRAMethod Type=Prior WorksMethod Type=Prior Works2026.0694.4Human BestMethod Type=Human Base...Method Type=Human Baseline2026.0683