by allenai · Agent Tool · ★ 108
WildGuard: Open One-stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs Authors: Seungju Han ⭐, Kavel Rao ⭐, Allyson Ettinger ☀️, Liwei Jiang ☀️, Yuchen Lin, Nathan Lambert, Yejin Choi, Nouha Dziri ⭐ Co-first authors, ☀️ co-second authors 🌟 WildGuard will appear at NeurIPS 2024 Datasets & Benchmarks! 🌟 WildGuard is a safety classification model for user-model chat exchanges. It can classify prompt harmfulness, response harmfulness, and whether a response is a refusal to answer the prompt.
| Stars | 108 |
| Forks | 11 |
| Language | Python |
| Category | Agent Tool |
| Quality Score | 67.4848422587015/100 |
| Open Issues | 3 |
| Last Updated | 2024-12-02 |
| Created | 2024-06-13 |
| Platforms | python |
| Est. Tokens | ~7k |
Explore other popular agent tool tools:
wildguard is Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs. It is categorized as a Agent Tool with 108 GitHub stars.
wildguard is primarily written in Python.
You can find installation instructions and usage details in the wildguard GitHub repository at github.com/allenai/wildguard. The project has 108 stars and 11 forks, indicating an active community.