AdaExploreBench (Stationary) - Calibrated Accuracy: leaderboard
AdaExploreBench's calibrated accuracy with the rule drift removed, a diagnostic setting: the agent still has to discover the hidden decision rule with its own probes, but the rule never changes, so the gap to the main setting measures the cost of undetected drift. Gemini-3.8-Flash and GPT-6-Sol have no runs here.
Metric: CA (%) (0-100). Source: hkust-knowcomp.github.io. Saturation forecast: Around February 2027. 5 models tracked. Updated 19 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash | High | 67.6 | #69 |
| 2 | Qwen 3.8 27B | Medium | 50.57 | #220 |
| 3 | GPT-6 Luna | High | 41.14 | #106 |
| 4 | Gemma 4 12B (IT) | High | 22.78 | #731 |
| 5 | Qwen 3.5 9B | 12.41 | #829 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=adaexplorebench-stationary-calibrated-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-20.