Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Idioms on EPIE
Loading...
99.2
AUROC
Full Representation Classifier
91.4
93.425
95.45
97.475
Apr 20, 2026
AUROC
Updated 1mo ago
Evaluation Results
Method
Method
Links
AUROC
Full Representation Classifier
Backbone=Llama-3.1-8B,...
2026.04
99.2
Full Rep.
Model=Gemma2-9B, Repre...
2026.04
98.6
Full Rep.
Model=GPT-OSS-20B, Rep...
2026.04
98.5
Full Rep.
Model=Qwen3-8B, Repres...
2026.04
97.9
Subspace
Model=Gemma2-9B, Repre...
2026.04
97.6
one-directional concreteness axis
Backbone=Llama-3.1-8B,...
2026.04
95.3
Subspace
Model=Qwen3-8B, Repres...
2026.04
93.1
Subspace
Model=GPT-OSS-20B, Rep...
2026.04
91.7
Feedback
Search any
task
Search any
task