Sparse Autoencoders Find Partially Interpretable Features in Italian Small Language Models

Alessandro Bondielli, Lucia Passaro, Alessandro Lenci


Anthology ID:
2025.clicit-1.11
Volume:
Proceedings of the Eleventh Italian Conference on Computational Linguistics (CLiC-it 2025)
Month:
September
Year:
2025
Address:
Cagliari, Italy
Editors:
Cristina Bosco, Elisabetta Jezek, Marco Polignano, Manuela Sanguinetti
Venue:
CLiC-it
SIG:
Publisher:
CEUR Workshop Proceedings
Note:
Pages:
95–104
Language:
URL:
https://preview.aclanthology.org/ingest-latest-mitpress-cl-tacl/2025.clicit-1.11/
DOI:
Bibkey:
Cite (ACL):
Alessandro Bondielli, Lucia Passaro, and Alessandro Lenci. 2025. Sparse Autoencoders Find Partially Interpretable Features in Italian Small Language Models. In Proceedings of the Eleventh Italian Conference on Computational Linguistics (CLiC-it 2025), pages 95–104, Cagliari, Italy. CEUR Workshop Proceedings.
Cite (Informal):
Sparse Autoencoders Find Partially Interpretable Features in Italian Small Language Models (Bondielli et al., CLiC-it 2025)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingest-latest-mitpress-cl-tacl/2025.clicit-1.11.pdf