Abstract
Multi-entity compositional questions pose significant challenges to existing retrieval-augmented language models. Conventional methods fall into a dilemma: standard RAG lacks dynamic reasoning, traditional Graph-RAG is limited by structural sparsity, and LLM-constructed Graph-RAG incurs prohibitive costs. We propose HyGRL, a unified framework that embeds unstructured text into structured knowledge graphs, creating a heterogeneous network for flexible evidence retrieval.
Reasoning is formulated as adaptive structure induction, learned via a robust two-stage process: (1) imitation learning distills heuristic expert signals, and (2) reinforcement learning refines the policy using LLM-driven preference rewards. Experiments demonstrate that HyGRL effectively merges textual richness with structural knowledge, outperforming SOTA baselines in answer accuracy and reasoning fidelity while maintaining extremely low token costs and near real-time inference.
Code available at: GitHub - HyGRL
Blogger's Review: HyGRL showcases immense potential in multi-entity reasoning by leveraging the strengths of both textual and graph structures. Its innovative learning mechanisms are particularly noteworthy, and I look forward to seeing more advancements based on this framework in practical applications!