List your company
ResearchFormerly WizWand

Fine-grained Hate Categorization on H-VLI (Easy)

92.22AccuracyQwen3-VL-Plus
39.928853.504467.0880.6556Mar 22, 2026
Updated 7mo ago

Evaluation results

MethodLinks
Qwen3-VL-PlusTraining=False2026.0392.22
GPT-4oTraining=False2026.0391.94
GPT-5-miniTraining=False, Framework=ARCADE2026.0391.66
GPT-5Training=False2026.0391.51
GPT-5-miniTraining=False2026.0391.25
GPT-4.1-miniTraining=False2026.0391.11
GPT-5Training=False, Framework=ARCADE2026.0390.88
Gemini-2.5-FlashTraining=False2026.0390.28
Qwen-VL-MaxTraining=False2026.0389.44
Gemini-2.5-ProTraining=False2026.0388.75
GPT-4oTraining=False, Framework=ARCADE2026.0387.76
Qwen3-VL-PlusTraining=False, Framework=ARCADE2026.0387.48
Qwen-VL-MaxTraining=False, Framework=ARCADE2026.0386.85
Qwen-VL-PlusTraining=False2026.0385.97
GPT-4.1-miniTraining=False, Framework=ARCADE2026.0384.98
Qwen-VL-PlusTraining=False, Framework=ARCADE2026.0383.5
Gemini-2.5-FlashTraining=False, Framework=ARCADE2026.0383.17
Gemini-2.5-ProTraining=False, Framework=ARCADE2026.0383.03
MMBTTraining=True2026.0368.61
BERT+ViTTraining=True2026.0368.33
PromptHateTraining=True2026.0368.33
BERTTraining=True2026.0365.83
MomentaTraining=True2026.0360.69
ViT-b-16Training=True2026.0341.94