List your company

Over-refusal evaluation on XS (test)

22OR ScoreTHRD-style
-0.885.061116.94Jul 1, 2026

Evaluation results

MethodLinks
THRD-styleDefense Type=Trajectory-aware firewalls, Target Model=mistral, Judge Model=qwen3.5:122b2026.0722
Granite Guardian 8BDefense Type=Per-message guard models, Target Model=mistral, Judge Model=qwen3.5:122b2026.0721
ShieldGemma 9BDefense Type=Per-message guard models, Target Model=mistral, Judge Model=qwen3.5:122b2026.0718
TCA-styleDefense Type=Trajectory-aware firewalls, Target Model=mistral, Judge Model=qwen3.5:122b2026.0714
Cognitive FirewallDefense Type=Trajectory-aware firewalls, Target Model=mistral, Judge Model=qwen3.5:122b2026.078
Llama Guard 3 8BDefense Type=Per-message guard models, Target Model=mistral, Judge Model=qwen3.5:122b2026.073
none (undefended)Defense Type=undefended, Target Model=mistral, Judge Model=qwen3.5:122b2026.070