Comparison of different regularization methods for training SAE models on the layer 1 MLP of TinyStories 2L 33M.
Lovis Heindrich
lovish
AI & ML interests
None yet
Organizations
Collections
2
models
30
lovish/SAE-pythia-70m-L3-6
Updated
•
2
lovish/SAE-pythia-70m-L2-6
Updated
•
1
lovish/SAE-pythia-70m-L4-5
Updated
•
1
lovish/SAE-pythia-70m-L1-28
Updated
•
1
lovish/SAE-pythia-70m-L0-23
Updated
•
1
lovish/SAE-pythia-70m-L4-22
Updated
•
1
lovish/SAE-pythia-70m-L1-19
Updated
•
1
lovish/SAE-pythia-70m-L3-16
Updated
•
1
lovish/SAE-pythia-70m-L0-15
Updated
•
2
lovish/SAE-pythia-70m-L2-11
Updated
•
1
datasets
None public yet