Harmful content detection on CADD GPT-4o generations
64.84AccuracyPerspective API
Evaluation results
| Method | Links | |
|---|---|---|
| Perspective API2026.04 | 64.84 | |
| OpenAI Moderation2026.04 | 36.08 | |
| LlamaGuard-12026.04 | 20.28 | |
| LlamaGuard-22026.04 | 0.77 |
| Method | Links | |
|---|---|---|
| Perspective API2026.04 | 64.84 | |
| OpenAI Moderation2026.04 | 36.08 | |
| LlamaGuard-12026.04 | 20.28 | |
| LlamaGuard-22026.04 | 0.77 |