这篇论文把知识图谱模型泛化的底层逻辑讲清楚了——原来只要看到半个链接就能猜出整条，还分出了四种情况。搞图神经网络或KG推理的人别错过。
这篇论文研究了知识图谱基础模型（KGFM）的零样本泛化机制。作者提出“半链接”概念，即三元组(h,r,t)中的(h,r)或(r,t)部分。实验表明，只需观察到测试三元组的一个半链接在推理图中，就足以准确预测完整链接。基于此，论文划分了四种半链接观测组合场景，并在这些场景下对现有KGFM进行了分层分析。分析发现，当前最先进的KGFM依赖已见半链接进行预测，而未见半链接带来不同挑战，这为未来模型改进提供了诊断方向。
Half a Link can Be Enough to Predict a Whole Link: Understanding Generalization in Knowledge Graph Foundation Models
Knowledge graph (KG) foundation models (KGFMs) are zero-shot generalizers: trained once, they can predict links on unseen graphs without retraining. However, understanding when and how they can robustly generalize across KGs is still an open question. In this paper, we shed some light on their generalization mechanisms highlighting how their performance on unseen KGs is not uniform when it comes to partially seen links, which we call half-links. In fact, we show that to predict a test triple $(h,r,t)$ it might suffice in practice to have observed the half-link $(h,r)$ or $(r,t)$ in the inference graph. This yields a taxonomy of four scenarios when combinations of these half-links are observed or not. In a rigorous stratified analysis over these scenarios, we reveal that SoTA KGFMs use seen half links for predictions, while unseen half-links pose different challenges. As such, our finer-grained taxonomy can be a diagnostic protocol for robust KGFM generalization and highlights where novel KGFMs can improve.