Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Feature Explanation Generation on Llama 3.1-8b SAE llamascope-res-32k Layer 2
Loading...
52.33
Gen Score
SAEExplainer
20.5476
28.7988
37.05
45.3012
Jun 7, 2026
Gen Score
Inp Score
Out Score
Ahigh Score
Alow Score
Delta Activation
Updated 1mo ago
Evaluation Results
Method
Method
Links
Gen Score
Inp Score
Out Score
Ahigh Score
Alow Score
Delta Activation
SAEExplainer
2026.06
52.33
77.67
77
9.08
9
8.99
N-pedia w/ Claude
Generator=Claude Sonne...
2026.06
49.67
75
73.33
8.79
10
8.69
N-pedia w/ GPT-5
Generator=GPT-5
2026.06
46.7
77.33
75.67
8.49
4
8.45
Neuronpedia
Generator=GPT-4o-mini
2026.06
27.87
56
69
5.81
6
5.75
Activation Oracles
2026.06
21.77
50.66
70
5
4
4.96
Feedback
Search any
task
Search any
task