arXiv：cs.LG· Kyaw Hpone Myint, Nan Jiang, Xiang Li, Zhe Wu, Alexandre G. R. Day, Pranab Mohanty, Giri Iyengar·· 4 小时前AI 评分32
QUARTET：用四分支交叉注意力和随机游走轨迹增强关系图上的 Transformer
QUARTET: Quad-branch cross-Attention and Random-walk Traces for Enhancing Transformers on Relational Graphs
AI 导读
针对关系图上的图 Transformer，研究者提出 QUARTET 架构，在局部子图上做全自注意力，并用四分支交叉注意力从种子特征、种子拓扑、时序动态和协作动态四个视角引入全局上下文。
正文
Abstract:Relational Deep Learning (RDL) models multi-table databases as heterogeneous temporal graphs, and graph transformers currently achieve state-of-the-art performance on benchmarks like RelBench. However, the current leading model, RelGT, suffers from two key limitations: its random local sampler yields loosely connected subgraphs that hinder message passing, and its global attention module relies on a single, seed-feature-based memory that ignores broader macro-level dynamics. To overcome these limitations, we introduce QUARTET, an expressive graph transformer architecture that applies full self-attention on local subgraphs while enriching global context through cross-attention branches. Specifically, QUARTET employs a Causal Random Walk (CRW) sampler based on recency-truncated Personalized PageRank (PPR) to extract compact, hub-robust, and densely connected local subgraphs without temporal leakage. Concurrently, a quad-branch cross-attention module integrates global context from four complementary perspectives: seed feature, seed topology, temporal dynamics, and collaborative dynamics. Across the RelBench v1 classification tasks, QUARTET consistently matches or outperforms the current state-of-the-art graph transformer baselines (HGT and RelGT). Ablation studies confirm that the CRW sampler significantly enriches local neighborhood quality, while the global branches provide essential, task-specific predictive gains.
| Comments: | This work has been accepted for main conference track at Learning on Graphs (LoG) 2026 |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.26855 [cs.LG] |
| (or arXiv:2609.26855v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2609.26855 arXiv-issued DOI via DataCite |
Submission history
From: Kyaw Hpone Myint [view email]
[v1]
Tue, 22 Sep 2026 14:36:19 UTC (313 KB)
[v2]
Thu, 24 Sep 2026 16:47:05 UTC (313 KB)
[v3]
Tue, 6 Oct 2026 21:35:46 UTC (313 KB)
来源：arXiv：cs.LG · arxiv.org