3 papers
cs.LG2026
Support-Compiled Feature Folding: More Evidence at Lower Memory Across Tabular Foundation Models
Tian Zhou, Beverly Jin, Xue Wang +6
Wide tables offer tabular foundation models more evidence, but accessing it can exhaust their memory: full-width pairwise mixing grows quadratically with the number of columns, whi…
cs.LG2026
Transferable Evidence Reconstruction for Longitudinal Glucose Representations
Tian Zhou, Bingqing Peng, Linxiao Yang +6
Long physiological recordings contain many routine measurements, while predictive information often lies in rare events, sustained burden, and recurring patterns. These properties…
cs.LG2026
What Do Tabular Foundation Models Compute In Context? In-Situ Representation Refinement through Attention-Gated Updates
Tian Zhou, Beverly Jin, Linxiao Yang +6
A tabular foundation model must discover which distinctions matter for each new table without updating its parameters. We develop in-situ representation refinement: support labels…