Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Shi, Wei, Li, Shuang, Yu, Kerun, Chen, Jinglei, Liang, Zujie, Wu, Xinhui, Qian, Yuxi, Wei, Feng, Zheng, Bo, Liang, Jiaqing, Chen, Jiangjie, Xiao, Yanghua
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2410.06519
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929533624516608
author Shi, Wei
Li, Shuang
Yu, Kerun
Chen, Jinglei
Liang, Zujie
Wu, Xinhui
Qian, Yuxi
Wei, Feng
Zheng, Bo
Liang, Jiaqing
Chen, Jiangjie
Xiao, Yanghua
author_facet Shi, Wei
Li, Shuang
Yu, Kerun
Chen, Jinglei
Liang, Zujie
Wu, Xinhui
Qian, Yuxi
Wei, Feng
Zheng, Bo
Liang, Jiaqing
Chen, Jiangjie
Xiao, Yanghua
contents There is a growing interest in expanding the input capacity of language models (LMs) across various domains. However, simply increasing the context window does not guarantee robust performance across diverse long-input processing tasks, such as understanding extensive documents and extracting detailed information from lengthy and noisy data. In response, we introduce SEGMENT+, a general framework that enables LMs to handle extended inputs within limited context windows efficiently. SEGMENT+ utilizes structured notes and a filtering module to manage information flow, resulting in a system that is both controllable and interpretable. Our extensive experiments across various model sizes, focusing on long-document question-answering and Needle-in-a-Haystack tasks, demonstrate the effectiveness of SEGMENT+ in improving performance.
format Preprint
id arxiv_https___arxiv_org_abs_2410_06519
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SEGMENT+: Long Text Processing with Short-Context Language Models
Shi, Wei
Li, Shuang
Yu, Kerun
Chen, Jinglei
Liang, Zujie
Wu, Xinhui
Qian, Yuxi
Wei, Feng
Zheng, Bo
Liang, Jiaqing
Chen, Jiangjie
Xiao, Yanghua
Computation and Language
There is a growing interest in expanding the input capacity of language models (LMs) across various domains. However, simply increasing the context window does not guarantee robust performance across diverse long-input processing tasks, such as understanding extensive documents and extracting detailed information from lengthy and noisy data. In response, we introduce SEGMENT+, a general framework that enables LMs to handle extended inputs within limited context windows efficiently. SEGMENT+ utilizes structured notes and a filtering module to manage information flow, resulting in a system that is both controllable and interpretable. Our extensive experiments across various model sizes, focusing on long-document question-answering and Needle-in-a-Haystack tasks, demonstrate the effectiveness of SEGMENT+ in improving performance.
title SEGMENT+: Long Text Processing with Short-Context Language Models
topic Computation and Language
url https://arxiv.org/abs/2410.06519