Search papers, labs, and topics across Lattice.
TopoChunker, a novel agentic framework, addresses semantic fragmentation in RAG by mapping documents to a Structured Intermediate Representation (SIR) that preserves cross-segment dependencies. A dual-agent architecture with an Inspector Agent for cost-optimized extraction and a Refiner Agent for hierarchical lineage reconstruction balances structural fidelity and computational cost. Experiments on GutenQA and GovReport show TopoChunker achieves state-of-the-art performance, improving generation accuracy by 8.0% and Recall@3 to 83.26%, while reducing token overhead by 23.5% compared to LLM-based baselines.
RAG systems can achieve state-of-the-art performance by explicitly preserving document topology, outperforming LLM-based chunking while simultaneously reducing token overhead.
Current document chunking methods for Retrieval-Augmented Generation (RAG) typically linearize text. This forced linearization strips away intrinsic topological hierarchies, creating ``semantic fragmentation''that degrades downstream retrieval quality. In this paper, we propose TopoChunker, an agentic framework that maps heterogeneous documents onto a Structured Intermediate Representation (SIR) to explicitly preserve cross-segment dependencies. To balance structural fidelity with computational cost, TopoChunker employs a dual-agent architecture. An Inspector Agent dynamically routes documents through cost-optimized extraction paths, while a Refiner Agent performs capacity auditing and topological context disambiguation to reconstruct hierarchical lineage. Evaluated on unstructured narratives (GutenQA) and complex reports (GovReport), TopoChunker demonstrates state-of-the-art performance. It outperforms the strongest LLM-based baseline by 8.0% in absolute generation accuracy and achieves an 83.26% Recall@3, while simultaneously reducing token overhead by 23.5%, offering a scalable approach for structure-aware RAG.
