← Google · Gemini Individual model view Google · Gemini

gemini-2.5-pro

Appears in 6 benchmarksMean lean (confidence − pass rate): +0.064/6 benchmarks lean overconfident (prospective probe)

Positioning spread: every benchmark, one model

0.000.250.500.751.00SQuAD (factual recall)MMLU-Pro (knowledge)LegalBench (legal reasoning)MathBench (competition math)OmniMath (advanced math)SciCode (scientific code)performanceconfidence (red gap = overconfident)
Performance vs. confidence for gemini-2.5-pro, per benchmark (prospective probe).
BenchmarkTask accConfidenceF₁Leans
SQuAD (factual recall)0.580.680.76+0.10 overconfident
MMLU-Pro (knowledge)0.670.870.82+0.20 overconfident
LegalBench (legal reasoning)0.870.790.82-0.09 cautious
MathBench (competition math)0.911.000.95+0.09 overconfident
OmniMath (advanced math)0.700.990.83+0.29 overconfident
SciCode (scientific code)0.610.380.54-0.23 cautious

In the full cloud

-3-3-2-2-1-100112233Performance z-score within benchmark/probe →Confidence z-score →
gemini-2.5-pro conditions all other model/condition points equal relative confidence and pass rate

Pairwise signal: pairs involving gemini-2.5-pro

Match accuracy controls for the performance base-rate gap
gemini-2.5-pro pairs
18/ 171
gemini-2.5-pro mean tau
+0.053
All-pairs mean
+0.037
gemini-2.5-pro p<0.05
12(66.7%)
-1.0-0.50.00.51.0Pair signal: do confidence gaps rank performance gaps? (Kendall tau-b)
all model pairs (observed) base-rate-matched null calibration-preserving null gemini-2.5-pro pair (filled = p<0.05) gemini-2.5-pro mean all-pairs mean

The four metacognitive outcomes

No curated cases for this selection yet — outcome-matrix extraction currently covers a sample of MMLU-Pro trials.
vs gemini-2.0-flash-001 →vs gemini-2.5-flash →vs gemini-3-flash-preview →vs gemini-3-pro-preview →vs gemini-3.1-pro-preview → Compare with anything →