Refusal Self-Prediction - Lean Harmful: leaderboard
Refusal self-prediction accuracy on requests refused more often than not but not always. Higher is better. Source: arxiv.org/abs/2604.00228.
Metric: Prediction Accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 78.8 | #148 |
| 2 | Claude Sonnet 4 | 75 | #207 |
| 3 | GPT-5.2 | 69.1 | #115 |
| 4 | Llama 3.1 405B | 66.7 | #476 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=refusal-self-prediction-lean-harmful · How It Works · Data refreshed daily, snapshot 2026-11-05.