Binary Hate Detection on H-VLI (Normal)
88.58AccuracyMMBT
Evaluation results
| Method | Links | |
|---|---|---|
| MMBTTraining=True2026.03 | 88.58 | |
| GPT-5Training=False, Framework=ARCADE2026.03 | 87.23 | |
| BERTTraining=True2026.03 | 85.47 | |
| BERT+ViTTraining=True2026.03 | 84.78 | |
| GPT-4.1-miniTraining=False, Framework=ARCADE2026.03 | 78.47 | |
| Gemini-2.5-FlashTraining=False, Framework=ARCADE2026.03 | 77.78 | |
| GPT-4oTraining=False, Framework=ARCADE2026.03 | 77.08 | |
| MomentaTraining=True2026.03 | 76.47 | |
| Gemini-2.5-ProTraining=False, Framework=ARCADE2026.03 | 73.96 | |
| Qwen-VL-PlusTraining=False, Framework=ARCADE2026.03 | 73.17 | |
| GPT-5-miniTraining=False, Framework=ARCADE2026.03 | 72.57 | |
| PromptHateTraining=True2026.03 | 69.2 | |
| Qwen-VL-MaxTraining=False, Framework=ARCADE2026.03 | 68.64 | |
| Qwen3-VL-PlusTraining=False, Framework=ARCADE2026.03 | 67.83 | |
| ViT-b-16Training=True2026.03 | 61.94 | |
| Gemini-2.5-ProTraining=False2026.03 | 61.25 | |
| Gemini-2.5-FlashTraining=False2026.03 | 53.29 | |
| GPT-5Training=False2026.03 | 50.67 | |
| GPT-4.1-miniTraining=False2026.03 | 50.52 | |
| Qwen3-VL-PlusTraining=False2026.03 | 47.04 | |
| GPT-4oTraining=False2026.03 | 39.45 | |
| GPT-5-miniTraining=False2026.03 | 31.83 | |
| Qwen-VL-PlusTraining=False2026.03 | 23.34 | |
| Qwen-VL-MaxTraining=False2026.03 | 17.77 |