AutoSchemaKG, AutoGraph R1, and multimodal abductive reasoning—three angles on building and using knowledge for language models.
Talk on how we move from symbolic knowledge graphs to neural graph databases, and what this means for data, algorithms, and real-world applications.
🚀 New Paper Alert! Excited to introduce DARK: Unifying Deductive & Abductive Reasoning in Knowledge Graphs with Masked Diffusion Models! 🧠🔗
Making High-Performance Graph RAG Affordable and Scalable
I am joining HKBU's Department of Computer Science as an Assistant Professor in 2026 and am actively recruiting 1-2 PhD students for Fall 2026 in NLP, Knowledge Graphs, and AI Agents.
![Featured image of post [ICLR'26] Taming the Wild Hypothesis: Introducing Controllable Abductive Reasoning in Knowledge Graphs](/p/iclr26-taming-the-wild-hypothesis-introducing-controllable-abductive-reasoning-in-knowledge-graphs/ctrlhgen_hu25833414136091988.png)
![Featured image of post [WWW'26] Unifying Logic Reasoning in Knowledge Graphs with Diffusion Models](/p/www26-unifying-logic-reasoning-in-knowledge-graphs-with-diffusion-models/dark_hu8635004305972634017.png)
![Featured image of post [ICLR'26] AtlasKV: Teaching a Language Model a Billion Facts, Without Breaking the Memory Bank](/p/iclr26-atlaskv-teaching-a-language-model-a-billion-facts-without-breaking-the-memory-bank/framework_hu11097651776597696018.png)