List your company

ResearchFormerly WizWand

Safety Guardrailing

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

Perplexity Filter

0False Positive Rate

32

Apr 3, 2026

Perplexity Filter

0False Positive Rate

26

Apr 3, 2026

ULSPB 350 interaction runs

StateGuard (Base-Ensemble)

0.02HS Rate

24

May 11, 2026

0False Positive Rate

24

Apr 3, 2026

Perplexity Filter

0False Positive Rate

24

Apr 3, 2026

PolicyShiftGuard-7B

61.7Macro F1

9

Jul 8, 2026

PolicyShiftGuard-7B

64.1F1 Score

9

Jul 8, 2026