AI Paper Atlas

Research map / Graph learning / GNN architectures

Torsion Graph Neural Networks

Cong Shen, Xiang Liu, Jiawei Luo, Kelin Xia

TPAMI 2025 · Graph learning › GNN architectures · approach: Physics-inspired GNNs · 6 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
weight message passing with analytic torsion (TorGNN vs GCN)GCN · link prediction average of 16 networksAUC: 0.9144 -> 0.9372Improved
weight message passing with analytic torsion (TorGNN vs GCN)GCN · link prediction average of 16 networksAUPR: 0.8649 -> 0.9319Improved
analytic torsion weight versus best baselineKGIN · link prediction average of 16 networksAUC: 0.9011 -> 0.9372Improved
use 2-hop neighborhood for local complex (l_sub=2) (ablation)TorGNN(1,1) · Drugbank_DTI link predictionAUC: 0.9380 -> 0.9651Improved
use 2-hop neighborhood for local complex (l_sub=2) (ablation)TorGNN(1,1) · ChG-Miner link predictionAUC: 0.9533 -> 0.9583Improved
use 2-hop neighborhood for local complex (l_sub=2) (ablation)TorGNN(1,1) · facebook_comp link predictionAUC: 0.8974 -> 0.8573Hurt
use 2-hop neighborhood for local complex (l_sub=2) (ablation)TorGNN(1,1) · p2p-Gnutella06 link predictionAUC: 0.8464 -> 0.9051Improved
add triangles (n=2) to local complex (ablation)TorGNN(1,1) · Drugbank_DTI link predictionAUC: 0.9380 -> 0.9306Hurt
add triangles (n=2) to local complex (ablation)TorGNN(1,1) · CA-GrQc link predictionAUC: 0.9573 -> 0.9328Hurt
add triangles (n=2) to local complex (ablation)TorGNN(1,1) · Pubmed node classificationaccuracy: 0.865 -> 0.872Improved
add triangles (n=2) to local complex (ablation)TorGNN(1,1) · Cora node classificationaccuracy: 0.832 -> 0.823Hurt
use 2-hop neighborhood for local complex (l_sub=2) (ablation)TorGNN(1,1) · Citeseer node classificationaccuracy: 0.748 -> 0.749No clear effect

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