Sotabase

Career

· Postdoctoral Researcher, Mila - Quebec Artificial Intelligence Institute2025–2026
· Postdoctoral Researcher, McGill University2024–
· PhD Student, McGill University2017–2024

Publications (55)

Revisiting Heterophily For Graph Neural Networks
Neural Information Processing Systems · 2022
268
cited
Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks
Neural Information Processing Systems · 2019
169
cited
Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?
arXiv.org · 2021
124
cited
When Do Graph Neural Networks Help with Node Classification: Investigating the Homophily Principle on Node Distinguishability
Neural Information Processing Systems · 2023
91
cited
MUDiff: Unified Diffusion for Complete Molecule Generation
LOG IN · 2023
53
cited
A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning
Neural Information Processing Systems · 2021
40
cited
Complete the Missing Half: Augmenting Aggregation Filtering with Diversification for Graph Convolutional Networks
arXiv.org · 2020
39
cited
The Heterophilic Graph Learning Handbook: Benchmarks, Models, Theoretical Analysis, Applications and Challenges
arXiv.org · 2024
32
cited
When Do We Need GNN for Node Classification?
arXiv.org · 2022
17
cited
Reactzyme: A Benchmark for Enzyme-Reaction Prediction
Neural Information Processing Systems · 2024
15
cited
EnzymeFlow: Generating Reaction-specific Enzyme Catalytic Pockets through Flow Matching and Co-Evolutionary Dynamics
arXiv.org · 2024
10
cited
RL Fine-Tuning Heals OOD Forgetting in SFT
arXiv.org · 2025
10
cited
Training Matters: Unlocking Potentials of Deeper Graph Convolutional Neural Networks
International Workshop on Complex Networks & Their Applications · 2020
10
cited
GCEPNet: Graph Convolution-Enhanced Expectation Propagation for Massive MIMO Detection
Global Communications Conference · 2024
9
cited
What Is Missing For Graph Homophily? Disentangling Graph Homophily For Graph Neural Networks
Neural Information Processing Systems · 2024
9
cited
Flexible Diffusion Scopes with Parameterized Laplacian for Heterophilic Graph Learning
LOG IN · 2024
7
cited
Let Your Features Tell The Differences: Understanding Graph Convolution By Feature Splitting
International Conference on Learning Representations · 2025
7
cited
Representation Learning on Heterophilic Graph with Directional Neighborhood Attention
arXiv.org · 2024
7
cited
META-Learning State-based Eligibility Traces for More Sample-Efficient Policy Evaluation
Adaptive Agents and Multi-Agent Systems · 2019
6
cited
What Is Missing In Homophily? Disentangling Graph Homophily For Graph Neural Networks
arXiv.org · 2024
6
cited
Sotabase
Sitao Luan | Researcher Profile | Sotabase | Sotabase