AI Paper Atlas

Research map / Graph learning / GNN architectures

MC-HNN: Learning Latent Structural Semantics and High-Rank Representations for Hypergraph Neural Networks

Shuyang Fang, Yuqin Huang, Zelong Yang, Yintao Cai, Xiaoping Min

ICML 2026 · Graph learning › GNN architectures · approach: Expressive equivariant message passing

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use MC-HNN instead of FrameHGNN on featureless heterophilic SenateMC-HNN · Senateaccuracy: 67.61 -> 78.31Improved
use MC-HNN instead of FrameHGNNMC-HNN · Houseaccuracy: 72.82 -> 77.59Improved
use MC-HNN instead of FrameHGNNMC-HNN · CiteSeeraccuracy: 74.72 -> 74.48No clear effect
use MC-HNN instead of Natural-HNNMC-HNN · ModelNet40accuracy: 98.56 -> 98.52No clear effect
add multi-channel message passing (ablation)MC-HNN · Coraaccuracy: 78.73 -> 82.33Improved
add multi-channel message passing (ablation)MC-HNN · NTU2012accuracy: 90.36 -> 90.82No clear effect
add latent hyperedge type encoding (ablation)MC-HNN · Senateaccuracy: 76.62 -> 78.31Improved
add dynamic channel gating (ablation)MC-HNN · Coraaccuracy: 82.02 -> 82.33Improved
add channel orthogonality regularization (ablation)MC-HNN · appendixaccuracyNo clear effect
add MC-HNN multi-channel design to deep models (2 to 64 layers)MC-HNN · Cora, House, Cora-CAaccuracy vs depthImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy