ResearchBenchmarksScalar probability judgment on UNLIFollow0.057MSECLUE-D0.028160.222830.41750.61217Apr 9, 2026Evaluation ResultsMethodMethodLinksMSECLUE-DParameters=3BParameters=3B2026.040.057TMTOParameters=14BParameters=14B2026.040.075Q3-VLParameters=32BParameters=32B2026.040.075CLUE-TParameters=3BParameters=3B2026.040.081Q3Parameters=32BParameters=32B2026.040.106Q2.5-OParameters=7BParameters=7B2026.040.114Q2.5-VLParameters=32BParameters=32B2026.040.126AFParameters=7BParameters=7B2026.040.272Q2-AParameters=7BParameters=7B2026.040.778