Toxicity Category Classification on ToxiAlert-Bench (test)
91.18AccuracyToxiAlert
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ToxiAlertType=Para.2026.05 | 91.18 | — | — | — | — | — | — | — | 79.48 | — | 83.3 | 79.34 | 80.21 | |
| ToxiAlertType=Tex.2026.05 | 86.21 | — | — | — | — | — | — | — | — | — | 75.66 | — | — | |
| ToxiAlert2026.05 | 80.04 | 81.1 | 90.94 | 81.85 | 48.78 | 52.76 | 65.95 | 39.3 | 69.69 | 86.33 | — | — | — | |
| GeminiType=Tex.2026.05 | 77.48 | — | — | — | — | — | — | — | — | — | 42.66 | — | — | |
| Qwen2Type=Tex.2026.05 | 77 | — | — | — | — | — | — | — | — | — | 34.77 | — | — | |
| GPT-4oType=Tex.2026.05 | 75.06 | — | — | — | — | — | — | — | — | — | 27.3 | — | — | |
| Qwen2Type=Para.2026.05 | 71.84 | — | — | — | — | — | — | — | 19.28 | — | 3.79 | 20.72 | 55.35 | |
| GPT-4oType=Para.2026.05 | 71.5 | — | — | — | — | — | — | — | 13.81 | — | 0.32 | 15.04 | 53.2 | |
| Gemini-2.5-Flash2026.05 | 70.84 | 53 | 58.89 | 34.32 | 47.15 | 21.61 | 61.64 | 36.19 | 57.55 | 75.38 | — | — | — | |
| GeminiType=Para.2026.05 | 69.48 | — | — | — | — | — | — | — | 31.11 | — | 19.57 | 31.31 | 52.9 | |
| GPT-4o Audio2026.05 | 61.89 | 27.08 | 12.22 | 20.17 | 29.27 | 18.09 | 34.88 | 21.01 | 39.91 | 64.52 | — | — | — | |
| Qwen2-Audio2026.05 | 55.15 | 4.42 | 0 | 12.21 | 0 | 2.51 | 26.83 | 9.73 | 19.24 | 60.41 | — | — | — | |
| DeToxy2026.05 | — | — | — | — | — | — | — | — | — | 85.7 | — | — | — | |
| YIDUN2026.05 | — | — | — | 0.5 | — | 0.5 | 0.65 | — | — | 50.49 | — | — | — |