Disentangling Linguistic and Paralinguistic Information with Routed Sparse Autoencoders
arXiv:2610.10865v1 Announce Type: new Abstract: Self-supervised speech encoders contain linguistic and paralinguistic information in a shared, entangled representation space. We combine a TopK sparse autoencoder with route-specific supervision and cross-factor adversaries. Across frozen SPEAR and W…
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-10-09 04:00 · arXiv cs.CL
Disentangling Linguistic and Paralinguistic Information with Routed Sparse Autoencoders