List your company

Confidence Estimation on SummEval

0.717AUROCVERDI LR ensemble
0.357160.450580.5440.63742May 11, 2026

Evaluation results

MethodLinks
VERDI LR ensembleCalls=1×2026.050.717
Best single signalCalls=1×2026.050.675
CoT-UQCalls=5×2026.050.652
Trace lengthCalls=1×2026.050.561
Logprob confidenceCalls=1×2026.050.371