Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Zijun, Wan, Zhennan, Li, Peng, Yan, Ming, Huang, Fei, Liu, Yang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914484071694336
author Liu, Zijun
Wan, Zhennan
Li, Peng
Yan, Ming
Huang, Fei
Liu, Yang
author_facet Liu, Zijun
Wan, Zhennan
Li, Peng
Yan, Ming
Huang, Fei
Liu, Yang
contents With the rapid advancement of post-training techniques for reasoning and information seeking, large language models (LLMs) can incorporate a large quantity of retrieved knowledge to solve complex tasks. However, the limited context window of LLMs obstructs scaling the amount of external knowledge input, prohibiting further improvement. Existing context window extension methods inevitably cause information loss. LLM-based multi-agent methods emerge as a new paradigm to handle massive input in a distributional manner, where we identify two core bottlenecks in existing agent orchestration designs. In this work, we develop a multi-agent framework, \textbf{\ExtAgents}, to overcome the bottlenecks and enable better scalability in inference-time knowledge integration without longer-context training. Benchmarked with our enhanced multi-hop question answering test, \textbf{$\boldsymbol{\infty}$Bench+}, and other public test sets including long survey generation, \ExtAgents significantly enhances the performance over existing non-training methods with the same amount of external knowledge input, regardless of whether it falls \emph{within or exceeds the context window}. Moreover, the method maintains efficiency due to high parallelism. We believe further study in the coordination of LLM agents on increasing external knowledge input could benefit real-world applications.
format Preprint
id arxiv_https___arxiv_org_abs_2505_21471
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration
Liu, Zijun
Wan, Zhennan
Li, Peng
Yan, Ming
Huang, Fei
Liu, Yang
Computation and Language
With the rapid advancement of post-training techniques for reasoning and information seeking, large language models (LLMs) can incorporate a large quantity of retrieved knowledge to solve complex tasks. However, the limited context window of LLMs obstructs scaling the amount of external knowledge input, prohibiting further improvement. Existing context window extension methods inevitably cause information loss. LLM-based multi-agent methods emerge as a new paradigm to handle massive input in a distributional manner, where we identify two core bottlenecks in existing agent orchestration designs. In this work, we develop a multi-agent framework, \textbf{\ExtAgents}, to overcome the bottlenecks and enable better scalability in inference-time knowledge integration without longer-context training. Benchmarked with our enhanced multi-hop question answering test, \textbf{$\boldsymbol{\infty}$Bench+}, and other public test sets including long survey generation, \ExtAgents significantly enhances the performance over existing non-training methods with the same amount of external knowledge input, regardless of whether it falls \emph{within or exceeds the context window}. Moreover, the method maintains efficiency due to high parallelism. We believe further study in the coordination of LLM agents on increasing external knowledge input could benefit real-world applications.
title Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration
topic Computation and Language
url https://arxiv.org/abs/2505.21471