Geometry-Aware Offline-to-Online Learning in Linear Contextual Bandits

ArXi:2604.24016v1 Announce Type: new We study offline-to-online learning in linear contextual bandits with biased offline regression data: the offline parameter need not match the online one, so history should not be treated as a single warm start. We model directional transfer with a shift certificate $(M_{\mathrm{shift}},\rho)$ and offline ridge estimation, yielding a geometry-aware confidence region for the online parameter rather than an isotropic radius.