List your company
ResearchFormerly WizWand

Binary Hate Detection on H-VLI (Hard)

79.39AccuracyGemini-2.5-Pro
9.751627.830845.9163.9892Mar 22, 2026
Updated 7mo ago

Evaluation results

MethodLinks
Gemini-2.5-ProTraining=False, Framework=ARCADE2026.0379.39
GPT-4oTraining=False, Framework=ARCADE2026.0378.79
Gemini-2.5-ProTraining=False2026.0378.11
Gemini-2.5-FlashTraining=False, Framework=ARCADE2026.0375.15
Gemini-2.5-FlashTraining=False2026.0369.23
GPT-5Training=False, Framework=ARCADE2026.0369.13
GPT-4.1-miniTraining=False, Framework=ARCADE2026.0366.06
GPT-4.1-miniTraining=False2026.0363.91
GPT-5-miniTraining=False, Framework=ARCADE2026.0361.82
GPT-4oTraining=False2026.0361.54
ViT-b-16Training=True2026.0355.62
Qwen3-VL-PlusTraining=False, Framework=ARCADE2026.0351.52
Qwen-VL-MaxTraining=False2026.0345.45
BERTTraining=True2026.0339.64
Qwen-VL-MaxTraining=False, Framework=ARCADE2026.0338.79
BERT+ViTTraining=True2026.0338.46
PromptHateTraining=True2026.0337.87
MMBTTraining=True2026.0336.69
MomentaTraining=True2026.0333.14
Qwen3-VL-PlusTraining=False2026.0331.52
Qwen-VL-PlusTraining=False, Framework=ARCADE2026.0329.88
GPT-5Training=False2026.0324.79
Qwen-VL-PlusTraining=False2026.0314.55
GPT-5-miniTraining=False2026.0312.43