LogosKG Accepted at ACL 2026 (Main Conference)
We are excited to share that our paper LogosKG: Hardware-Optimized Scalable and Interpretable Knowledge Graph Retrieval has been accepted to the main conference at ACL 2026.
This work is led by He Cheng, PhD, postdoc researcher in our lab.
Knowledge graphs (KGs) are a critical substrate for grounding LLM reasoning in structured, verifiable evidence, but existing systems struggle to balance efficiency, scalability, and interpretability when performing multi-hop retrieval at biomedical scale. LogosKG addresses this directly: it is a hardware-aligned framework that executes k-hop traversal as efficient operations over decomposed symbolic representations, integrates degree-aware partitioning and on-demand caching to scale to billion-edge graphs, and achieves substantial efficiency gains over both CPU and GPU baselines without sacrificing retrieval fidelity.
Beyond retrieval, the paper demonstrates a two-round KG-LLM interaction pipeline that uses LogosKG to analyze how KG topology (hop distribution, connectivity patterns) shapes alignment between structured biomedical knowledge and LLM diagnostic reasoning, opening a concrete path toward next-generation KG-LLM integration.
This work is foundational to our broader AI4Science research program on scalable, interpretable biomedical knowledge infrastructure.
