How does feature learning reshape the function space?
João Lobo, Bruno Loureiro, Long Tran-Than, Fanghui Liu · 2026
Read How does feature learning reshape the function space? by João Lobo, Bruno Loureiro, Long Tran-Than, Fanghui Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-17718.txt.
Description
João Lobo and three colleagues show how feature learning reshapes the function space of a two-layer network. After one large gradient step in the proportional regime, the feature distribution matches a target-dependent spiked Gaussian covariance. That induces a data-adaptive kernel reshaping the space and shifting its structure. Learning acts as a distributional transformation in parameter space or input space, equivalently adding a target-dependent kernel. It boosts target-aligned eigenvalues, mixes leading eigenfunctions, and couples the top radial mode with a quadratic harmonic.
Catalog record
| CREATOR | João Lobo, Bruno Loureiro, Long Tran-Than, Fanghui Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 161733 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.17718