Toxicity Detection on OpenAI Moderation API Evaluation (test)
88.15AccuracyOMod
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| OModDescription=OpenAI Moderation API2024.05 | 88.15 | 87.03 | 82.38 | 31.99 | 15.13 | 11.69 | |
| MULITraining dataset=ToxicChat2024.05 | 86.85 | 85.84 | 78.54 | 37.16 | 24.9 | 22.8 | |
| MULITraining dataset=LMSYS-Chat-1M2024.05 | 86.61 | 87.52 | 77.78 | 41.38 | 25.86 | 18.01 | |
| LlamaGuard2024.05 | 85.95 | 84.74 | 75.86 | 34.87 | 14.56 | 12.64 |