Search papers, labs, and topics across Lattice.
This paper introduces a GraphRAG stack that enhances clinical LLMs by utilizing a 700K-node medical knowledge graph to constrain responses to verifiable paths, thereby addressing issues of hallucination and opaque reasoning. By employing a directed Pruned Landmark Labeling (PLL) oracle for distance checks and a lightweight AStarNet heuristic for clinically plausible expansions, the system achieves significant improvements in latency and recall. The hybrid approach not only lowers Time to First Token (TTFT) but also reduces clinician-audited hallucinations while maintaining clarity in explanations, making it suitable for real-world applications in medical reasoning.
Clinicians can now rely on a hybrid reasoning model that dramatically reduces hallucinations while enhancing response clarity in multi-hop medical queries.
Hallucinations and opaque reasoning remain unacceptable failure modes for clinical LLMs. We present a production-grade GraphRAG stack that constrains answers to verifiable graph chain-of-thought paths in a heterogeneous, ~700K-node medical knowledge graph powering a fertility assistant. The core idea is targeted navigation: a directed Pruned Landmark Labeling (PLL) oracle provides exact distances for sub-millisecond feasibility checks and simple-path enumeration, while a lightweight AStarNet heuristic operates strictly within the PLL corridor to prioritize clinically plausible expansions. We score and pack a small, diverse set of paths (CUI/semantic-type overlap, length prior, provenance priors) to condition generation, yielding compact prompts and improved Time to First Token (TTFT). On fertility-focused queries, the hybrid (PLL+AStarNet) establishes a better latency/recall Pareto frontier than text-only RAG and single-component baselines, lowers TTFT, and reduces clinician-audited hallucinations while preserving explanation clarity. The result is a practical recipe for explainable, low-hallucination multi-hop medical reasoning ready for real-world deployment.
