Sparse Autoencoders Can Induce Local Geometry for Mitigating Catastrophic Forgetting
Catastrophic forgetting remains a challenge in supervised continual learning, as adapting to new tasks causes representation drift and loss of previously learned information. This raises the question of which representational changes should be constrained while preserving plasticity. We therefore introduce Sparse Autoencoder Geometric Anchoring (SAGA), which converts a Sparse Autoencoder (SAE) trained at each task boundary into local, low-rank constraints on representation drift. Replay examples specify where the constraints apply, while leading right singular directions of the encoder Jacobian specify which changes are penalized. The resulting geometry is input dependent and can be retai...
| Authors | Kamil Faber, Katarzyna Filus, Łukasz Pawela |
|---|---|
| Primary category | cs.LG |
| Published | 15 October 2026 |
| Last updated at source | 15 October 2026 |
Popularity, last 6 days
0 now. Counts as reported by the source, snapshotted daily by Silytics.
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 24 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.