AI Paper Atlas

Research map / Graph learning / Knowledge and temporal graphs

Entangled No More: Multi-Domain Decoupling for Robust Dynamic Graph Neural Networks

Youda Mo, Chaobo He, Junwei Cheng, Peng Mei, Quanlong Guan

ICML 2026 · Graph learning › Knowledge and temporal graphs · approach: Temporal graph neural architectures

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 MP-K2alman (multi-particle kernel Kalman module) (ablation)DeR-Mamba · COLLAB link prediction, originalAUC (%): 91.79 -> 94.83Improved
add MP-K2alman (ablation)DeR-Mamba · COLLAB NETTACK evasionaverage AUC drop (%): 31.8 -> 18.0Improved
add AFDM (adversarial-aware frequency decoupling) (ablation)DeR-Mamba · COLLAB link prediction, originalAUC (%): 94.06 -> 94.83Improved
add AFDM (ablation)DeR-Mamba · COLLAB NETTACK evasionaverage AUC drop (%): 21.9 -> 18.0Improved
add AFDM (ablation)DeR-Mamba · COLLAB feature attack gamma=0.5AUC (%): 69.35 -> 70.16No clear effect
add both MP-K2alman and AFDM (ablation)DeR-Mamba · COLLAB structure attackAUC (%): 90.10 -> 93.74Improved
add both modules (ablation)DeR-Mamba · Yelp structure attackAUC (%): 74.07 -> 79.28Improved
add both modules (ablation)DeR-Mamba · ACT structure attackAUC (%): 94.13 -> 96.60Improved
use DeR-Mamba instead of DG-Mambarobust DGNN · COLLAB structure attackAUC (%): 92.60 -> 93.74Improved
use DeR-Mamba instead of DG-Mambarobust DGNN · Yelp NETTACK evasionaverage AUC drop (%): 22.1 -> 13.9Improved
use DeR-Mamba instead of DG-Mambarobust DGNN · COLLAB NETTACK evasion np=1AUC drop (%): 12.6 -> 12.7No clear effect
use DeR-Mamba instead of DG-Mambarobust DGNN · Yelp feature attack gamma=0.5AUC (%): 61.82 -> 61.60Hurt

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