List your company

Question Answering on TruthfulQA (Brier Score)

13.57Brier ScoreCAGE-CAL
12.236821.235930.23539.2341May 28, 2026

Evaluation results

MethodLinks
CAGE-CALCalibration category=Trained calibrators2026.0513.57
DiscoUQ-LLMCalibration category=Trained calibrators2026.0515.31
+ Scaling-bin.Calibration category=Post-hoc plurality calibrators2026.0516.06
+ PlattCalibration category=Post-hoc plurality calibrators2026.0516.36
+ IsotonicCalibration category=Post-hoc plurality calibrators2026.0516.45
Scalar + GBTCalibration category=Trained calibrators2026.0517.49
Plurality shareCalibration category=Post-hoc plurality calibrators2026.0524.52
GraphCalCalibration category=Trained calibrators2026.0528.32
Collab. Cal.Calibration category=LLM-elicited confidence estimators2026.0544.21
LLM-Cal (+topo)Calibration category=LLM-elicited confidence estimators2026.0546.51
LLM-Cal (no topo)Calibration category=LLM-elicited confidence estimators2026.0546.9