Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Out-of-Distribution Detection on 20NG (ID) vs MNLI (OOD)
Loading...
0.9989
AUROC
Gemma2-2b LR baseline
0.96874
0.97657
0.9844
0.99223
May 12, 2026
AUROC
AUPR (OOD)
FPR@95% Recall
Updated 21d ago
Evaluation Results
Method
Method
Links
AUROC
AUPR (OOD)
FPR@95% Recall
Gemma2-2b LR baseline
Backbone=Gemma2-2b, Ba...
2026.05
0.9989
0.9988
0.27
Gemma2-9b LR baseline
Backbone=Gemma2-9b, Ba...
2026.05
0.9982
0.997
0.3
Gemma2-9b + 16k SAE + Markov
Backbone=Gemma2-9b, SA...
2026.05
0.9763
0.9768
16.28
Gemma2-2b + 16k SAE + Markov
Backbone=Gemma2-2b, SA...
2026.05
0.9699
0.968
19.84
Feedback
Search any
task
Search any
task