LLM\(\times\)MapReduce: Simplified Long-Sequence Processing using Large Language Models

• Published in: arXiv.org 2024-10
• Main Authors: Zhou, Zihan, Li, Chong, Chen, Xinyi, Wang, Shuo, Yu, Chao, Li, Zhili, Wang, Haoyu, An, Rongqiao, Shi, Qi, Tan, Zhixing, Xu, Han, Shi, Xiaodong, Liu, Zhiyuan, Sun, Maosong
• Format: Article
• Language: English
• Subjects: Context; Documents; Large language models; Texts
• ISSN: 2331-8422

Enlarging the context window of large language models (LLMs) has become a crucial research area, particularly for applications involving extremely long texts. In this work, we propose a novel training-free framework for processing long texts, utilizing a divide-and-conquer strategy to achieve comprehensive document understanding. The proposed LLM\(\times\)MapReduce framework splits the entire document into several chunks for LLMs to read and then aggregates the intermediate answers to produce the final output. The main challenge for divide-and-conquer long text processing frameworks lies in the risk of losing essential long-range information when splitting the document, which can lead the model to produce incomplete or incorrect answers based on the segmented texts. Disrupted long-range information can be classified into two categories: inter-chunk dependency and inter-chunk conflict. We design a structured information protocol to better cope with inter-chunk dependency and an in-context confidence calibration mechanism to resolve inter-chunk conflicts. Experimental results demonstrate that LLM\(\times\)MapReduce can outperform representative open-source and commercial long-context LLMs, and is applicable to several different models.