Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Concept Identifiability on Sycophancy
Loading...
1
MCC
Linear Probe
0.498408
0.628629
0.75885
0.889071
Feb 14, 2025
MCC
Updated 1mo ago
Evaluation Results
Method
Method
Links
MCC
Linear Probe
Model Backbone=Gemma-2-2B
2025.02
1
SSAE
Model Backbone=Gemma-2-2B
2025.02
0.707
GemmaScope
Model Backbone=Gemma-2-2B
2025.02
0.6344
ReLU-SAE
Model Backbone=Gemma-2-2B
2025.02
0.6005
TopK-SAE
Model Backbone=Gemma-2-2B
2025.02
0.5896
JumpReLU SAE
Model Backbone=Gemma-2-2B
2025.02
0.5177
Feedback
Search any
task
Search any
task