InjecGuard
Open-source prompt guard with published training data; achieves +30.8% over prior state-of-the-art on the NotInject benchmark, specifically addressing overdefense false positives that break legitimate use cases.
- from
- Prompt Injection
- added
- 2026-10-10
- likes
- 0
Prompt Injection › Tools: “Open-source prompt guard with published training data; achieves +30.8% over prior state-of-the-art on the NotInject benchmark, specifically addressing overdefense false positives that break legitimate use cases.”