Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders

Carolina Zheng, Nicolas Beltran-Velez, Sweta Karlekar, Claudia Shi, Achille Nazaret, Asif Mallik, Amir Feder, David M. Blei


Abstract
Traditional topic models are effective at uncovering latent themes in large text collections. However, due to their reliance on bag-of-words representations, they struggle to capture semantically abstract features. While some neural variants use richer representations, they are similarly constrained by expressing topics as word lists, which limits their ability to articulate complex topics. We introduce Mechanistic Topic Models (MTMs), a class of topic models that operate on interpretable features learned by sparse autoencoders (SAEs). By defining topics over this semantically rich space, MTMs can reveal deeper conceptual themes with expressive feature descriptions. Moreover, uniquely among topic models, MTMs enable controllable text generation using topic steering vectors. To properly evaluate MTM topics against word list approaches, we propose topic judge, an LLM-based pairwise comparison evaluation framework. Across eight datasets, MTMs match or exceed traditional and neural baselines on coherence metrics, are consistently preferred by topic judge, and enable effective LLM steering.1
Anthology ID:
2026.tacl-1.99
Volume:
Transactions of the Association for Computational Linguistics, Volume 14
Month:
Year:
2026
Address:
Cambridge, MA
Venue:
TACL
SIG:
Publisher:
MIT Press
Note:
Pages:
2187–2212
Language:
URL:
https://aclanthology.org/2026.tacl-1.99/
DOI:
10.1162/tacl.a.797
Bibkey:
Cite (ACL):
Carolina Zheng, Nicolas Beltran-Velez, Sweta Karlekar, Claudia Shi, Achille Nazaret, Asif Mallik, Amir Feder, and David M. Blei. 2026. Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders. Transactions of the Association for Computational Linguistics, 14:2187–2212.
Cite (Informal):
Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders (Zheng et al., TACL 2026)
Copy Citation:
PDF:
https://aclanthology.org/2026.tacl-1.99.pdf