HELM AIR-Bench 2024 - #39-40.34: Characterization of identity - Predicted Personality: leaderboard
Metric: Refusal Rate (%). Source: crfm.stanford.edu. 87 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 100 |
| 2 | GPT-5 | 100 |
| 3 | GPT-4.1 Nano | 100 |
| 4 | Claude Sonnet 4.5 | 100 |
| 5 | GPT-5 Nano | 100 |
| 6 | GPT-OSS-120B | 100 |
| 7 | Llama 4 Scout Instruct | 100 |
| 8 | GPT-4o Mini (2024-07-18) | 100 |
| 9 | GPT-4o (2024-08-06) | 100 |
| 10 | GPT-OSS-20B | 100 |
| 11 | Llama 3 8B Instruct | 100 |
| 12 | DeepSeek R1 | 100 |
| 13 | DeepSeek V3 | 100 |
| 14 | Claude Haiku 4.5 (20251001) | 100 |
| 15 | Llama 4 Maverick Instruct FP8 | 100 |
Interactive version: theaggregate.ai/benchmark?slug=helm-air-bench-2024-39-40-34-characterization-of-identity-predicted-personality · How It Works · Data refreshed daily, snapshot 2026-09-19.