ATBench - Failure Mode Diagnosis: leaderboard
Metric: Fine-grained diagnosis accuracy (%) of the failure-mode label (ATBench taxonomy) over the 497 unsafe ATBench trajectories, with the AgentDoG diagnosis template; higher is better. Source: arxiv.org. Saturation forecast: Around 2035. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 13.5 |
| 2 | Gemini 3.1 Pro (Preview) | 12.6 |
| 3 | GPT-5.2 | 12 |
| 4 | Qwen 3 235B A22B 2507 Instruct | 11.6 |
| 5 | Qwen 3 4B 2507 Instruct | 9.6 |
| 6 | QwQ-32B | 9.4 |
| 7 | Gemini 3 Flash | 8.3 |
| 8 | Qwen 3 4B | 8.2 |
| 9 | Qwen 2.5 7B Instruct | 6 |
| 10 | Llama 3.1 8B Instruct | 5.8 |
| 11 | Qwen 3.5 397B A17B | 3.6 |
| 12 | Qwen 3.5 4B | 3 |
Interactive version: theaggregate.ai/benchmark?slug=atbench-failure-mode-diagnosis · How It Works · Data refreshed daily, snapshot 2026-10-07.