Author name: Bhavik Mangla

MDKeyChunker: Single-Call LLM Enrichment with Rolling Keys and Key-Based Restructuring for High-Accuracy RAG

Bhavik Mangla / March 30, 2026

arXiv:2603.23533v2 Announce Type: replace
Abstract: RAG pipelines typically rely on fixed-size chunking, which ignores document structure, fragments semantic units across boundaries, and requires multiple LLM calls per chunk for metadata extraction. W…

cs.AI, cs.CL, cs.IR, cs.LG

MDKeyChunker: Single-Call LLM Enrichment with Rolling Keys and Key-Based Restructuring for High-Accuracy RAG

Bhavik Mangla / March 26, 2026

arXiv:2603.23533v1 Announce Type: cross
Abstract: RAG pipelines typically rely on fixed-size chunking, which ignores document structure, fragments semantic units across boundaries, and requires multiple LLM calls per chunk for metadata extraction. We …