Bowen Deng

PhD student@SUN YAT-SEN UNIVERSITY·中国·OpenReview

研究方向

Graph Machine Learning · Self-supervised Learning · Robust Graph Neural Network · Domain Adaptation · Open-world Machine Learning · Large Language Model · LLM Post-training

6.8

Rethinking the Gold Standard: Why Discrete Curvature Fails to Fully Capture Over-squashing in GNNs?

ICLR 2026Poster

二作

8.2

Self-Assembling Graph Perceptrons

NeurIPS 2025Spotlight

三作

7.3

Soft-consensual Federated Learning for Data Heterogeneity via Multiple Paths

NeurIPS 2025Poster

7.2

Less is More: Federated Graph Learning with Alleviating Topology Heterogeneity from A Causal Perspective

合作者 (15)

Bowen Deng

Rethinking the Gold Standard: Why Discrete Curvature Fails to Fully Capture Over-squashing in GNNs?

Self-Assembling Graph Perceptrons

Soft-consensual Federated Learning for Data Heterogeneity via Multiple Paths

Less is More: Federated Graph Learning with Alleviating Topology Heterogeneity from A Causal Perspective

Unsupervised Federated Graph Learning

GLNCD: Graph-Level Novel Category Discovery

Towards Understanding Parametric Generalized Category Discovery on Graphs

Graph Neural Ricci Flow: Evolving Feature from a Curvature Perspective