Malicious Prompt Detection on AdvBench (subset)
30Fires Ratev2 rubric (deterministic synthesis function)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| v2 rubric (deterministic synthesis function)Prompts=50, Rules eval=32026.06 | 30 | 60 |
| Method | Links | ||
|---|---|---|---|
| v2 rubric (deterministic synthesis function)Prompts=50, Rules eval=32026.06 | 30 | 60 |