AI Paper Atlas

Research map / Graph learning / GNN architectures

Expressive and Invariant Graph Learning via Canonical Tree Cover Neural Networks

Michael Ito, Danai Koutra, Jenna Wiens

ICLR 2026 · Graph learning › GNN architectures · approach: Weisfeiler-Leman expressivity hierarchies

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 a spanning tree cover instead of a single tree (ablation)CTNN · SCOPaccuracy: 68.7 -> 72.1Improved
use a spanning tree cover instead of a single tree (ablation)CTNN · GO BIOAUC: 69.4 -> 82.0Improved
use a spanning tree cover instead of a single tree (ablation)CTNN · GO MOLAUC: 79.9 -> 86.6Improved
use a spanning tree cover instead of a single tree (ablation)CTNN · BBBPAUC: 87.4 -> 86.1No clear effect
use TreeMPNN instead of plain MPNN on trees (ablation)CTNN · SCOPaccuracy: 57.1 -> 72.1Improved
use TreeMPNN instead of plain MPNN on trees (ablation)CTNN · HIVAUC: 73.8 -> 75.2Improved
use TreeMPNN instead of plain MPNN on trees (ablation)CTNN · BBBPAUC: 81.3 -> 86.1Improved
use TreeMPNN (residual-edge message passing) instead of TreeRNN (ablation)CTNN · SCOPaccuracy: 64.8 -> 72.1Improved
use TreeMPNN instead of TreeRNN (ablation)CTNN · BACEAUC: 83.5 -> 79.3Hurt
use TreeMPNN instead of TreeRNN (ablation)CTNN · HIVAUC: 77.9 -> 75.2Hurt
use CTNN tree cover instead of RCM sequence orderingCTNN · ClinToxAUC: 70.7 -> 84.7Improved
use CTNN instead of hand-crafted FingerprintCTNN · BACEAUC: 82.9 -> 79.3Hurt

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