AI Paper Atlas

Research map / Graph learning / GNN architectures

Directional Sheaf Hypergraph Networks: Unifying Learning on Directed and Undirected Hypergraphs

Emanuele Mule, Stefano Fiorini, Antonio Purificato, Federico Siciliano, Stefano Coniglio, Fabrizio Silvestri

ICLR 2026 · Graph learning › GNN architectures · approach: Oversmoothing and heterophily · 4 citations (OpenAlex)

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 the directed sheaf hypergraph Laplacian (DSHNLight) instead of the best baselineDSHNLight · email-EUaccuracy: 68.91 -> 82.67Improved
Use DSHNLight instead of GeDi-HNN on Roman-empireDSHNLight · Roman-empireaccuracy: 83.87 -> 89.24Improved
Add directional sheaf Laplacian versus SheafHyperGNN on TelegramDSHN · Telegramaccuracy: 70.00 -> 88.65Improved
Use DSHN on Cora versus PhenomNNDSHN · Coraaccuracy: 88.12 -> 88.02No clear effect
Use nonzero charge parameter q (directional information) on Cora (ablation)DSHN · CoraaccuracyHurt
Use nonzero charge parameter q on Telegram (ablation)DSHN · TelegramaccuracyImproved
Increase number of layers (ablation)DSHN · node classificationaccuracyImproved
Increase number of layers (ablation)HGNN · node classificationaccuracyHurt
Increase stalk dimension d (ablation)DSHN · node classificationaccuracyImproved
Detach gradients through Laplacian construction (DSHNLight vs DSHN) (ablation)DSHN · Squirrelaccuracy: 43.55 -> 44.09No clear effect
Use DSHNLight instead of DSHN on Telegram (ablation)DSHN · Telegramaccuracy: 88.65 -> 81.15Hurt
Use DSHNLight instead of DSHN on Roman-empire (ablation)DSHN · Roman-empireaccuracy: OOM -> 89.24Improved

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