-
Notifications
You must be signed in to change notification settings - Fork 0
Notebook 26 — Port SAE training to Mamba-2.8B #3
Open
Labels
Description
Mamba architecture lacks SAE coverage in open source. Challenge: state-space activations are structured differently than transformer residuals. Target: train a reference SAE on state_mixer output, report observations. Likely requires hand-rolled hook path — budget 1-2 days.
Metadata
Metadata
Assignees
Labels
Type
Fields
Give feedbackNo fields configured for issues without a type.