YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Topk-Staircase-Tied-All: A Top-K SAE where the entire encoder and decoder matrices (Wenc, Wdec) are shared across all layers (no slicing). The shared hidden dimension was 40 × dmodel.

Downloads last month

-

Downloads are not tracked for this model. How to track
Safetensors
Model size
216k params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Collection including davidquarel/topk-staircase-share-all.shakespeare_64x4