AI RESEARCH
Information Routing in Atomistic Foundation Models: How Task Alignment and Equivariance Shape Linear Disentanglement
arXiv CS.LG
•
ArXi:2603.03155v2 Announce Type: replace What determines whether a molecular property prediction model organizes its representations so that geometric and compositional information can be cleanly separated? We Across ten models from five architectural families on QM9, we find a \emph{linear accessibility gradient}: models differ by $6.6\times$ in geometric information accessible after composition removal ($R^2_{\mathrm{geom}}$ from 0.081 to 0.533 for HOMO-LUMO gap). Three factors explain this gradient.