GuardBench

Evals

Benchmark / dataset

Benchmark for guardrail models across 40 datasets.

Category
Guardrails & safety classifiers (adjacent)
Insurance exposure
Regulatory/bias
Publisher
Academic
Publisher type
Academic
License
EUPL-1.2
Status
Stale (>12mo)
Last commit
2025-10-09
Stars
39

View source on GitHub →

Wanich Risk Lab, 2026.