List your company

ResearchFormerly WizWand

Over-refusal Compliance

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

PH Seemingly Toxic

87Compliance Rate (Keyword Filter)

5

Mar 5, 2026

OK Seemingly Toxic

86Compliance Rate (Keyword Filter)

5

Mar 5, 2026

OR Seemingly Toxic

87Compliance Rate (Keyword Filter)

5

Mar 5, 2026

CoCo Seemingly Toxic

98Compliance Rate (Keyword Filter)

5

Mar 5, 2026

98Compliance Rate (Keyword Filter)

5

Mar 5, 2026