Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Interpretability Evaluation on Top-activated Texts
Loading...
0.74
Embedding Similarity
CLIP+SAE
-0.1024
0.1163
0.335
0.5537
Feb 16, 2025
Embedding Similarity
Win Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Embedding Similarity
Win Rate
CLIP+SAE
2025.02
0.74
61
CLIP
2025.02
0.45
59
CLIP+NCL
2025.02
0.45
60
DeCLIP
2025.02
-0.07
46
Feedback
Search any
task
Search any
task