牛哥精选 · 所有

1

📝 深度技术 arXiv 机器学习 2026-06-15

Generalizing GNNs with Tokenized Mixture of Experts

KDD 2026论文提出Tokenized MoE框架，用专家混合机制突破GNN泛化瓶颈

arXiv:2602.09258v2 Announce Type: replace Abstract: Deployed graph neural networks (GNNs) are frozen at deployment yet must fit clean data, generalize…

gnn mixture of 图神经网络泛化 tokenized

2

📝 深度技术 arXiv 机器学习 2026-06-11

Beyond the Golden Teacher: Enhancing Graph Learning through LLM-GNN Co-teaching

LLM与GNN协同教学攻克少样本图学习难题，双模型互补超越单打独斗。

arXiv:2606.11583v1 Announce Type: new Abstract: Text-attributed graphs (TAGs) underlie real-world applications such as citation networks, social media…

图学习少样本学习 llm gnn 协同教学

3

📝 深度技术 arXiv AI 2026-06-10

ERAlign: Energy-based Representation Alignment of GNNs and LLMs on Text-attributed Graphs

聚焦能量对齐方法，为GNN与LLM在文本图上的协同建模拓展新思路。

arXiv:2606.10461v1 Announce Type: cross Abstract: Text-attributed Graphs (TAGs) incorporate textual node attributes with graph structures to describe …

gnn llm 文本属性图表示对齐基于能量模型

4

📝 深度技术 arXiv AI 2026-05-28

Let Relations Speak: An End-to-End LLM-GNN Soft Prompt Framework for Fraud Detection

LLM与GNN巧妙联姻，用软提示融合关系图信息，为欺诈检测提供新范式。

arXiv:2605.28524v1 Announce Type: new Abstract: In recent years, Large Language Models (LLMs) have shown great capability in processing graph tasks su…

llm gnn 欺诈检测软提示图神经网络

5

🤖 AI·大模型 arXiv 机器学习 2026-05-27

Morphling: Fast, Fused, and Flexible GNN Training at Scale

大规模GNN训练中硬件挑战的突破：Morphling通过融合不规则图遍历与密集矩阵计算实现快速灵活训练

arXiv:2512.01678v5 Announce Type: replace Abstract: Graph Neural Networks (GNNs) present a fundamental hardware challenge by fusing irregular, memory-…

graph neur gnn训练硬件优化融合架构大规模

6

📝 深度技术 arXiv AI 2026-05-23

Ex-GraphRAG: Interpretable Evidence Routing for Graph-Augmented LLMs

新型图增强LLM框架Ex-GraphRAG，通过可解释证据路由解决GNN编码器节点贡献纠缠问题。

arXiv:2605.21994v1 Announce Type: cross Abstract: GraphRAG conditions language models on subgraphs retrieved from knowledge graphs, encoded via messag…

graphrag 可解释性图增强大模型证据路由 gnn

7

📝 深度技术 arXiv AI 2026-05-20

Trapping Attacker in Dilemma: Examining Internal Correlations and External Influences of Trigger for Defending GNN Backdoors

提出通过分析触发器内部相关性和外部影响来防御GNN后门攻击的新方法，使攻击者陷入两难困境。

arXiv:2605.08278v2 Announce Type: replace-cross Abstract: GNNs have become a standard tool for learning on relational data, yet they remain highly vul…

gnn 后门攻击防御触发器内外相关性

8

📝 深度技术 arXiv 机器学习 2026-05-20

B-cos GNNs: Faithful Explanations through Dynamic Linearity

基于动态线性度的忠诚可解释GNN方法，为图神经网络的复杂决策提供清晰可信的推理路径。

arXiv:2605.19778v1 Announce Type: new Abstract: We introduce B-cos GNNs, an inherently explainable class of graph neural networks whose predictions de…

b-cos gnns 图神经网络可解释ai 动态线性忠实解释

🐂 牛哥精选