List your company

Response Safety Assessment on TTI Attack Dataset Responses

49Safe Response Countanthropic-claude-3.5-haiku
29.2434.3739.544.63Apr 23, 2026

Evaluation results

MethodLinks
anthropic-claude-3.5-haiku2026.04491982
openai-gpt-4.1-mini2026.04464928
openai-gpt-4o-2024-08-062026.04464928
openai-gpt-4o-mini2026.04464928
meta-llama-llama-4-maverick2026.044559010
deepseek-deepseek-r1-05282026.044559010
mistral-medium-32026.044468812
mistral-small-3.1-24b-instruct2026.0440108020
mistral-devstral-small2026.0438127624
gemini-2.0-flash2026.0433176634
gemini-1.5-flash-8b2026.0432186436
gemini-2.0-flash-lite2026.0430206040
gemini-1.5-flash2026.0430206040