AdaExploreBench (Evidence Given) - Calibrated Accuracy: leaderboard

AdaExploreBench's calibrated accuracy when the evaluation batches are unmasked, a diagnostic setting: evidence about the hidden rule accumulates passively instead of having to be sought with probes, so the gap to the main setting measures the cost of active exploration.

Metric: CA (%) (0-100). Source: hkust-knowcomp.github.io. Saturation forecast: Around February 2027. 6 models tracked. Updated 19 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1DeepSeek V4.1 FlashHigh70.34#69
2GPT-6 SolHigh66.57#33
3GPT-6 LunaHigh51.53#106
4Gemma 4 12B (IT)High36.52#731
5Qwen 3.8 27BMedium33.78#220
6Qwen 3.5 9B28.18#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-evidence-given-calibrated-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-20.