List your company

ResearchFormerly WizWand

Offline OOD Detection

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

Mathematical Reasoning Far-shift OOD

TV Score w/ DiSmo

96.54AUROC

14

Feb 26, 2026

Mathematical Reasoning Near-shift OOD

98.76AUROC

14

Feb 26, 2026