AI Paper Atlas

Research map / Graph learning / GNN architectures

AdS-GNN - a Conformally Equivariant Graph Neural Network

Maksim Zhdanov, Nabil Iqbal, Erik J Bekkers, Patrick Forré

ICLR 2026 · Graph learning › GNN architectures · approach: Equivariant networks for molecules · 1 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
replace Euclidean distance with AdS proper distance in EGNN messages (conformal equivariance)EGNN · PascalVOC-SP (LRGB)test F1: 27.80 -> 28.07No clear effect
use AdS-GNN instead of EGNN on SuperPixel MNIST with rotated and scaled test dataEGNN · SuperPixel MNIST (augmented test)test errorImproved
use AdS-GNN instead of EGNN on shape segmentation, in distributionEGNN · Shape segmentationaccuracyImproved
use AdS-GNN instead of MPNN (translation-only) on shape segmentationMPNN · Shape segmentationaccuracyHurt
use AdS-GNN instead of EGNN for Ising N-point correlation predictionEGNN · 2d Ising N-point functionsrelative L2Improved
reduce graph connectivity k_con from fully connected to 2 (ablation)AdS-GNN and EGNN · Ising generalization across Ngeneralization and in-distribution performanceMixed
compare AdS-GNN to P-Theta-NITA with orientation information on SuperPixel MNIST in-distributionAdS-GNN · SuperPixel MNISTclassification errorHurt

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