PANDA: Learning the Hypergraph You Were Never Given via Prototype-Anchored Diffusion

Hypergraph neural networks leverage higher order interactions by propagating node features along an observed incidence. When the observed incidence misses task relevant relations, hypergraph structure learning methods infer or refine hyperedges from node features. When learned and observed structures share a propagation schedule, their diffusion depths cannot be controlled independently. We introduce Prototype Anchored Neural Diffusion for Adaptive Hypergraphs (PANDA), which learns a soft incidence between $N$ nodes and $K$ trainable feature space prototypes. Shallow diffusion over its induced operator limits repeated global coupling, while an observed view filter with independently chosen depth supports either one step smoothing or deeper restart propagation. A learned scalar gate combines the two filtered representations at each layer. The learned incidence requires $O(NK)$ storage, and neither filter needs to materialize a dense node to node operator. Across 12 node classification benchmarks, PANDA achieves the best average accuracy rank of $2.00$, compared with $4.58$ for the runner up TF-HNN. The results support allocating diffusion depth according to relational connectivity, so learning global relations need not entail deep global mixing.