AuraTracer智迹闻
中文

EVENT DOSSIER

Embedded Graph Flows for Categorical Graph Generation

2026-09-07 12:00 Science across 2 days 🔥 47.2 heat score
2sources
2days unfolding
47.2heat score
4mentions
SummaryAI generated

The research team proposed a generative model called Embedded Graph Flow (EGF) aimed at solving the problem of classifying graph structures. This model uses a permutation-isomorphic graph transformer to transfer Gaussian noise to the continuous embedded endpoints of nodes and unordered edge categories, and maps back to discrete graph categories through a terminal reader. In molecular generation benchmarks, EGF performed excellently: on the QM9 dataset, its Fréchet ChemNet Distance (FCD) was 0.150, significantly better than the classification diffusion baseline DiGress (0.717) and the bridge-based baseline GruM (0.812), achieving the best results in all four metrics. Additionally, when applied to the ZINC250k dataset, EGF maintained the lowest maximum mean difference (MMD) with the neighborhood subgraph pairwise distance kernel (NSPDK), indicating its high consistency with the local substructures of reference molecules. The relevant code has been publicly released on the GitHub repository.

Related eventsRELATED EVENTS
Key entitiesKEY ENTITIES
EGFQM9Trusted-System-LabZINC250k

Coverage · reports per dayLANGUAGE SPLIT

Entity relations
EGF × QM91EGF × Trusted-System-Lab1EGF × ZINC250k1QM9 × Trusted-System-Lab1QM9 × ZINC250k1Trusted-System-Lab × ZI…1

Integrated timelineUNIFIED TIMELINE

  1. 2026-09-04

    Embedded Graph Flows for Categorical Gr…

    提出一种名为嵌入式图流(EGF)的生成模型,该模型利用置换等变图变换器将高斯噪声传输至节点与无序边类别的连续嵌入端点,并通过终端读取器映射回离散图类别。在分子基准测试中,EGF 表现具有竞争力;在 QM9 数据集上,其在四项指标中均优于其…

  2. 2026-09-07

    Embedded Graph Flows for Categorical Gr…

    提出一种名为嵌入式图流(EGF)的生成模型,该模型利用置换等变图变换器将高斯噪声传输至节点及无序边类别的连续嵌入端点,并通过终端读取映射回离散图类别。在分子基准测试中,EGF 在 QM9 数据集上于四项指标中均取得最佳结果,其 Fréch…

SignalsSIGNALS

Keyword heat
  • Trusted-System-Lab2
  • EGF1
  • QM91
  • ZINC250k1

All reports (2)SOURCES

A arXiv cs.LG en 2026-09-05 00:26

Embedded Graph Flows for Categorical Graph Generation

提出一种名为嵌入式图流(EGF)的生成模型,该模型利用置换等变图变换器将高斯噪声传输至节点与无序边类别的连续嵌入端点,并通过终端读取器映射回离散图类别。在分子基准测试中,EGF 表现具有竞争力;在 QM9 数据集上,其在四项指标中均优于其他两种方法,其中 Fréchet ChemNet Distance(FCD)达到 0.150,显著低于分类扩散基线 DiGress 的 0.717 和基于桥梁的基线 GruM 的 0.812。当应用于 ZINC250k 中的大分子时,EGF 在采用邻域子图成对距离核(NSPDK)时仍保持最低的最大均值差异(MMD),表明其与参考分子的局部子结构高度一致。相关代码已发布于 GitHub 仓库 https://github.com/Trusted-System-Lab/EGF。

A arXiv cs.LG en 2026-09-07 12:00

Embedded Graph Flows for Categorical Graph Generation

提出一种名为嵌入式图流(EGF)的生成模型,该模型利用置换等变图变换器将高斯噪声传输至节点及无序边类别的连续嵌入端点,并通过终端读取映射回离散图类别。在分子基准测试中,EGF 在 QM9 数据集上于四项指标中均取得最佳结果,其 Fréchet ChemNet Distance(FCD)为 0.150,优于分类扩散基线 DiGress(0.717)和基于桥梁的基线 GruM(0.812)。在 ZINC250k 数据集上应用时,EGF 使用邻域子图成对距离核保持最低的最大均值差异(MMD),表明其与参考分子的局部子结构高度一致。相关代码已发布于 GitHub 仓库。