Understanding Structured Health Data through Interaction-Aware Mixture-of-Experts
Ji H Park ⋅ Ding ⋅ Tianjian Guo
Abstract
We study interaction-aware mixture-of-experts for post-stroke rigidity prediction using multi-level views of structured health records. Despite minimal performance gains, routing attribution reveals systematic importance differences across views, underscoring view construction as key to interpretability.
Chat is not available.
Successful Page Load