3 papers
cs.LG2026
Interactions Between Crosscoder Features: A Compact Proofs Perspective
Dmitry Manning-Coe, Thomas Read, Anna Soligo +4
Dictionary learning methods like Sparse Autoencoders (SAEs) and crosscoders attempt to explain a model by decomposing its activations into independent features. Interactions betwee…
cond-mat.str-el2026
Decomposing momentum scales in the Hubbard Model: From Hatsugai-Kohmoto to Aubry-André
Dmitry Manning-Coe, Barry Bradlyn
The all-to-all momentum coupling of the Hubbard interaction makes interacting lattice models generically unsolvable. In many settings, however, from Peierls instabilities to Moiré…
cs.LG2025
Grokking vs. Learning: Same Features, Different Encodings
Dmitry Manning-Coe, Jacopo Gliozzi, Alexander G. Stapleton +4
Grokking typically achieves similar loss to ordinary, "steady", learning. We ask whether these different learning paths - grokking versus ordinary training - lead to fundamental di…