AI Paper Atlas

Research map / Graph learning / GNN architectures

Graph Self-supervised Learning with Accurate Discrepancy Learning

Dongki Kim, Jinheon Baek, Sung Ju Hwang

NeurIPS 2022 · Graph learning › GNN architectures · approach: Graph structure learning · 11 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
add graph discrimination loss L_GD (ablation)D-SLA · ClinToxROC-AUC: 58.00 -> 70.83Improved
add L_edit alone (without L_GD) (ablation)D-SLA · ClinToxROC-AUC: 58.00 -> 57.46No clear effect
add L_edit on top of L_GD (ablation)D-SLA · ClinToxROC-AUC: 70.83 -> 77.48Improved
add L_edit on top of L_GD (ablation)D-SLA · BACEROC-AUC: 81.58 -> 83.53Improved
add triplet margin loss L_margin (ablation)D-SLA · ClinToxROC-AUC: 77.48 -> 80.17Improved
add triplet margin loss L_margin (ablation)D-SLA · BACEROC-AUC: 83.53 -> 83.81Improved
add L_edit on top of L_GD (ablation)D-SLA · COLLAB link predictionaccuracy: 74.23 -> 76.19Improved
increase margin alpha from 5.0 to 10.0 (ablation)D-SLA · BACEROC-AUC: 83.81 -> 78.34Hurt
increase lambda2 from 0.5 to 0.9 (ablation)D-SLA · BACEROC-AUC: 83.81 -> 80.72Hurt
reduce perturbation magnitude from 10% to 0.1% (ablation)D-SLA · COLLAB link predictionaccuracy: 70.67 -> 76.19Improved
add node attribute masking (ablation)D-SLA · COLLAB link predictionaccuracy: 70.42 -> 76.19Improved
D-SLA vs best contrastive/predictive baselinesGIN pretraining · Molecule/PPI averageROC-AUC avg: 72.65 -> 73.85Improved

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