README.md

---
title: RefusalBench
emoji: 🧬
colorFrom: red
colorTo: indigo
sdk: gradio
sdk_version: 5.50.0
app_file: app.py
pinned: true
license: mit
tags:
  - benchmark
  - llm-evaluation
  - ai-safety
  - biosecurity
  - refusal
  - leaderboard
datasets:
  - appliedscientific/refusalbench
---

# RefusalBench

Interactive leaderboard for the RefusalBench benchmark — a reproducible, evergreen evaluation of frontier LLM refusal on biological research prompts.

**Paper:** [arXiv:2605.21545](https://arxiv.org/abs/2605.21545)  
**GitHub:** [AppliedScientific/refusalbench](https://github.com/AppliedScientific/refusalbench)