|国家预印本平台
首页|TANGO: Graph Neural Dynamics via Learned Energy and Tangential Flows

TANGO: Graph Neural Dynamics via Learned Energy and Tangential Flows

TANGO: Graph Neural Dynamics via Learned Energy and Tangential Flows

来源:Arxiv_logoArxiv
英文摘要

We introduce TANGO -- a dynamical systems inspired framework for graph representation learning that governs node feature evolution through a learned energy landscape and its associated descent dynamics. At the core of our approach is a learnable Lyapunov function over node embeddings, whose gradient defines an energy-reducing direction that guarantees convergence and stability. To enhance flexibility while preserving the benefits of energy-based dynamics, we incorporate a novel tangential component, learned via message passing, that evolves features while maintaining the energy value. This decomposition into orthogonal flows of energy gradient descent and tangential evolution yields a flexible form of graph dynamics, and enables effective signal propagation even in flat or ill-conditioned energy regions, that often appear in graph learning. Our method mitigates oversquashing and is compatible with different graph neural network backbones. Empirically, TANGO achieves strong performance across a diverse set of node and graph classification and regression benchmarks, demonstrating the effectiveness of jointly learned energy functions and tangential flows for graph neural networks.

Moshe Eliasof、Eldad Haber、Carola-Bibiane Schönlieb

计算技术、计算机技术

Moshe Eliasof,Eldad Haber,Carola-Bibiane Schönlieb.TANGO: Graph Neural Dynamics via Learned Energy and Tangential Flows[EB/OL].(2025-08-07)[2025-08-18].https://arxiv.org/abs/2508.05070.点此复制

评论