1 citations · 1 across the 8 of their papers we have counts for
Showing cs.LGShow all
3 papers · 1 filter
cs.LG2026
Weights Read and Write Features: Scalable Parameter Decomposition Grounded in Activation Space
Tue M. Cao, Lisiane Pruinelli, My T. Thai
Activation space and parameter space provide complementary views of model computation. Activations represent information, while weights read, transform, and write that information.…
cs.LG2026
Semantic Optimal Transport for Sparse Autoencoder Feature Matching and Circuit Compression
Tue M. Cao, Nguyen Do, My T. Thai
Sparse autoencoders (SAEs) have become a central tool for interpreting language models. However, two key SAE analyses that remain difficult to scale are (1) matching semantically s…
cs.LG2026
Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders
Tue M. Cao, Hoang X. Nhat, Raed Alharbi +2
Learning hierarchical features in Sparse Autoencoders (SAEs) is essential for capturing the structured nature of real-world data and mitigating issues like feature absorption or sp…