04:00
2026-05-29
arxiv.org
large-language-models
Micro-Macro Retrieval: Reducing Long-Form Hallucination in Large Language Models
Researchers have introduced Micro-Macro Retrieval (M2R), a new framework designed to reduce hallucination in large language models during long-form text generation. The system addresses the problem ofโฆ