Harmful content detection on Trolling-oriented generations Llama-3.1 70B
26.04AccuracyOpenAI Moderation
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI Moderation2026.04 | 26.04 | |
| Perspective API2026.04 | 24.23 | |
| LlamaGuard-22026.04 | 11.92 | |
| LlamaGuard-12026.04 | 10.51 |
Loading the SOTA2 catalog…