Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Alignment Audit on Jigsaw Toxic Comment
Loading...
0
Average Treatment Effect (ATE)
SAE-based explanation
-0.0016
0.0092
0.02
0.0308
Jun 28, 2025
Average Treatment Effect (ATE)
ATE 95% CI Margin of Error
Updated 3mo ago
Evaluation Results
Method
Method
Links
Average Treatment Effect (ATE)
ATE 95% CI Margin of Error
SAE-based explanation
top_k_features=1
2025.06
0
-
SAE-based explanation
top_k_features=2
2025.06
0
-
SAE-based explanation
top_k_features=3
2025.06
0.02
-
SAE-based explanation
top_k_features=4
2025.06
0.03
-
SAE-based explanation
top_k_features=5
2025.06
0.04
-
Feedback
Search any
task
Search any
task