Share your thoughts, 1 month free Claude Pro on usSee more

Selective Prediction on TruthfulQA

39.4PRR

Evo-Gpt-9

Updated 12d ago

Evaluation Results

Method	Links
Evo-Gpt-9 2026.04		39.4
Evo-Gpt-7 2026.04		39.4
SequenceProbability 2026.04		39.4
AttentionScore 2026.04		39.4
Evo-Anth-30 2026.04		39.3
Evo-Anth-22 2026.04		39.3
Evo-Anth-15 2026.04		39.3
SP×ExpectedRank 2026.04		39.1
Evo-Anth-1 2026.04		39.1
Evo-Anth-8 2026.04		39.1
SP×MeanEntropy 2026.04		39
RAUQ 2026.04		37.2
Perplexity 2026.04		36.2
MeanTokenEntropy 2026.04		36.1
Evo-Gpt-5 2026.04		35.8
Evo-Gpt-1 2026.04		35.8
Evo-Gpt-13 2026.04		35.4
Evo-Gpt-11 2026.04		34.1
Evo-Gpt-15 2026.04		33.7
Evo-Gpt-3 2026.04		33.5