List your company

Multimodal Safety Evaluation on multimodal safety dataset

0.13ASRClaude 3.5 Sonnet
0.10980.246150.38250.51885Apr 13, 2025

Evaluation results

MethodLinks
Claude 3.5 Sonnet2025.040.130.69440.590.560.68
GPT-4o Mini2025.040.250.64440.550.540.63
Gemini 1.5 Pro2025.040.2550.23330.760.760.75
Claude 3 Haiku2025.040.2650.47220.630.610.67
GPT-4o2025.040.440.12780.720.810.66
Gemini 1.5 Flash2025.040.6350.23890.560.60.46