Low-rank sparse autoencoders: Unifying efficiency and geometric regularization for large language model interpretability

Jiajia Mu, Benying Tan, Jie Ren, Mingyu Guo, Chenchen Luo, Ruibin Bai. Low-rank sparse autoencoders: Unifying efficiency and geometric regularization for large language model interpretability. Inf. Sci., 755:123768, 2026. [doi]

Abstract

Abstract is missing.