{"slug": "quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on", "title": "QUARTET: Quad-branch cross-Attention and Random-walk Traces for Enhancing Transformers on Relational Graphs", "summary": "Researchers introduced QUARTET, a graph transformer architecture that combines a Causal Random Walk (CRW) sampler built on recency-truncated Personalized PageRank (PPR) with a quad-branch cross-attention module drawing on 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, according to the arXiv paper 2609.26855v1. Ablation studies confirm the CRW sampler enriches local neighborhood quality while the global branches deliver task-specific predictive gains.", "body_md": "arXiv:2609.26855v1 Announce Type: new \nAbstract: 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.", "url": "https://wpnews.pro/news/quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on", "canonical_source": "https://arxiv.org/abs/2609.26855", "published_at": "2026-09-24 04:00:00+00:00", "updated_at": "2026-09-24 04:31:27.852895+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "neural-networks"], "entities": ["QUARTET", "RelGT", "HGT", "RelBench", "Causal Random Walk", "Personalized PageRank", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on", "markdown": "https://wpnews.pro/news/quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on.md", "text": "https://wpnews.pro/news/quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on.txt", "jsonld": "https://wpnews.pro/news/quartet-quad-branch-cross-attention-and-random-walk-traces-for-enhancing-on.jsonld"}}