List your company
ResearchFormerly WizWand

Question Answering on NQ 8 additional samples

0.853AUROCClustered Self-Assessment
0.599240.665120.7310.79688Jun 2, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Clustered Self-AssessmentModel=Qwen3-14B2026.060.853
Clustered Self-AssessmentModel=Qwen2.5-32B2026.060.85
Clustered Self-AssessmentModel=Qwen2.5-7B2026.060.846
Clustered Self-AssessmentModel=Gemma-3-12B2026.060.843
Clustered Self-AssessmentModel=Qwen3-30B-A3B2026.060.84
Clustered Self-AssessmentModel=Qwen2.5-14B2026.060.837
EigVModel=Qwen2.5-7B2026.060.825
Clustered Self-AssessmentModel=Gemma-3-27B2026.060.821
EigVModel=Qwen3-14B2026.060.819
Semantic EntropyModel=Qwen3-14B2026.060.812
Semantic EntropyModel=Qwen2.5-7B2026.060.809
EigVModel=Qwen2.5-32B2026.060.806
DegModel=Gemma-3-12B2026.060.805
EigVModel=Qwen3-30B-A3B2026.060.804
P(True)Model=Qwen3-14B2026.060.799
Semantic EntropyModel=Qwen2.5-32B2026.060.799
Semantic EntropyModel=Qwen3-30B-A3B2026.060.793
EigVModel=Qwen2.5-14B2026.060.793
DegModel=Qwen2.5-7B2026.060.793
DegModel=Qwen3-14B2026.060.79
SARModel=Qwen2.5-7B2026.060.789
SARModel=Qwen3-14B2026.060.787
DegModel=Gemma-3-27B2026.060.786
P(True)Model=Qwen2.5-32B2026.060.785
P(True)Model=Qwen3-30B-A3B2026.060.783
Semantic EntropyModel=Qwen2.5-14B2026.060.779
sentenceSARModel=Qwen2.5-7B2026.060.778
DegModel=Qwen2.5-32B2026.060.777
ProbabilityModel=Qwen2.5-7B2026.060.776
ProbabilityModel=Qwen3-14B2026.060.774
SARModel=Qwen3-30B-A3B2026.060.773
Predictive EntropyModel=Qwen2.5-7B2026.060.77
sentenceSARModel=Qwen3-14B2026.060.77
SARModel=Qwen2.5-32B2026.060.769
DegModel=Qwen3-30B-A3B2026.060.769
ProbabilityModel=Qwen2.5-32B2026.060.763
Predictive EntropyModel=Qwen3-14B2026.060.763
DegModel=Qwen2.5-14B2026.060.763
Eigen ScoreModel=Qwen2.5-7B2026.060.76
SARModel=Qwen2.5-14B2026.060.76
sentenceSARModel=Qwen2.5-32B2026.060.759
LN EntropyModel=Qwen2.5-7B2026.060.757
EccModel=Qwen2.5-7B2026.060.757
ProbabilityModel=Gemma-3-27B2026.060.755
EccModel=Qwen3-14B2026.060.755
tokenSARModel=Gemma-3-12B2026.060.754
EigVModel=Gemma-3-12B2026.060.753
Predictive EntropyModel=Qwen2.5-32B2026.060.752
LN EntropyModel=Qwen3-14B2026.060.751
sentenceSARModel=Qwen2.5-14B2026.060.751
tokenSARModel=Gemma-3-27B2026.060.75
sentenceSARModel=Qwen3-30B-A3B2026.060.75
PerplexityModel=Gemma-3-27B2026.060.749
PerplexityModel=Qwen2.5-7B2026.060.748
ProbabilityModel=Qwen3-30B-A3B2026.060.748
Eigen ScoreModel=Qwen3-14B2026.060.748
ProbabilityModel=Qwen2.5-14B2026.060.747
EccModel=Qwen2.5-32B2026.060.747
PerplexityModel=Gemma-3-12B2026.060.745
Predictive EntropyModel=Qwen2.5-14B2026.060.745
EccModel=Gemma-3-27B2026.060.744
Predictive EntropyModel=Qwen3-30B-A3B2026.060.742
EccModel=Gemma-3-12B2026.060.741
LN EntropyModel=Qwen3-30B-A3B2026.060.739
Eigen ScoreModel=Qwen2.5-32B2026.060.739
EigVModel=Gemma-3-27B2026.060.738
PerplexityModel=Qwen3-14B2026.060.737
LN EntropyModel=Qwen2.5-32B2026.060.735
tokenSARModel=Qwen2.5-7B2026.060.732
EccModel=Qwen2.5-14B2026.060.732
P(True)Model=Qwen2.5-14B2026.060.731
EccModel=Qwen3-30B-A3B2026.060.731
PerplexityModel=Qwen3-30B-A3B2026.060.729
ProbabilityModel=Gemma-3-12B2026.060.729
P(True)Model=Qwen2.5-7B2026.060.729
Eigen ScoreModel=Qwen2.5-14B2026.060.729
PerplexityModel=Qwen2.5-32B2026.060.728
Semantic EntropyModel=Gemma-3-12B2026.060.726
tokenSARModel=Qwen3-14B2026.060.725
SARModel=Gemma-3-12B2026.060.724
Eigen ScoreModel=Qwen3-30B-A3B2026.060.723
LN EntropyModel=Qwen2.5-14B2026.060.72
tokenSARModel=Qwen2.5-32B2026.060.714
tokenSARModel=Qwen3-30B-A3B2026.060.714
Semantic EntropyModel=Gemma-3-27B2026.060.708
PerplexityModel=Qwen2.5-14B2026.060.706
sentenceSARModel=Gemma-3-12B2026.060.705
tokenSARModel=Qwen2.5-14B2026.060.701
Predictive EntropyModel=Gemma-3-12B2026.060.697
sentenceSARModel=Gemma-3-27B2026.060.692
Predictive EntropyModel=Gemma-3-27B2026.060.691
SARModel=Gemma-3-27B2026.060.688
LN EntropyModel=Gemma-3-12B2026.060.685
P(True)Model=Gemma-3-27B2026.060.659
LN EntropyModel=Gemma-3-27B2026.060.658
Eigen ScoreModel=Gemma-3-12B2026.060.654
Eigen ScoreModel=Gemma-3-27B2026.060.649
P(True)Model=Gemma-3-12B2026.060.609