Last released Jun 5, 2026
AIGuard: model-agnostic safety evaluation toolkit with PDF reports (adversarial, evaluator, hallucination)
Supported by