{"slug": "generative-representation-learning-on-hyper-relational-knowledge-graphs-via", "title": "Generative Representation Learning on Hyper-relational Knowledge Graphs via Masked Discrete Diffusion", "summary": "Researchers have introduced KREPE, the first generative representation learning method for hyper-relational knowledge graphs (HKGs) that can generate valid facts from arbitrarily masked queries, including completing partially observed facts or creating entirely new facts from scratch. The method uses masked discrete diffusion to model probability distributions of missing components, addressing real-world scenarios where multiple or all parts of a fact may be unknown. KREPE achieves state-of-the-art performance on standard HKG link prediction benchmarks and outperforms LLM-based baselines in generating novel and correct facts.", "body_md": "arXiv:2605.24064v1 Announce Type: new\nAbstract: Hyper-relational knowledge graphs (HKGs) effectively represent complex facts. While inferring new knowledge in HKGs is a critical problem, current methods cast it as a simple link prediction, assuming that nearly all entities and relations within a fact are known, leaving only a single blank to be filled. However, this restricted assumption may not hold in real-world scenarios in which multiple, or even all, constituent components of a fact may be missing simultaneously. To bridge this gap, we introduce a task called fact generation: generating a valid hyper-relational fact from an arbitrarily masked query, i.e., completing a partially observed fact or generating a fact from scratch. We propose KREPE, the first generative representation learning method for HKGs that learns to model the probability distributions of missing components conditioned on the local fact components and global structure of HKGs via a masked discrete diffusion. KREPE models both the intra-fact dependencies by contextual message passing and inter-fact correlations by aggregating stochastically sampled contexts. KREPE seamlessly unifies link prediction and fact generation within a single training framework, achieving state-of-the-art performance on standard HKG link prediction benchmarks and outperforming LLM-based baselines in generating novel and correct facts.", "url": "https://wpnews.pro/news/generative-representation-learning-on-hyper-relational-knowledge-graphs-via", "canonical_source": "https://arxiv.org/abs/2605.24064", "published_at": "2026-05-26 04:00:00+00:00", "updated_at": "2026-05-26 04:07:38.186152+00:00", "lang": "en", "topics": ["machine-learning", "generative-ai", "artificial-intelligence", "neural-networks", "ai-research"], "entities": ["KREPE"], "alternates": {"html": "https://wpnews.pro/news/generative-representation-learning-on-hyper-relational-knowledge-graphs-via", "markdown": "https://wpnews.pro/news/generative-representation-learning-on-hyper-relational-knowledge-graphs-via.md", "text": "https://wpnews.pro/news/generative-representation-learning-on-hyper-relational-knowledge-graphs-via.txt", "jsonld": "https://wpnews.pro/news/generative-representation-learning-on-hyper-relational-knowledge-graphs-via.jsonld"}}