Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2410.06886 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917798473629696 |
|---|---|
| author | Deng, Jingyang Shen, Zhengyang Wang, Boyang Su, Lixin Cheng, Suqi Nie, Ying Wang, Junfeng Yin, Dawei Ma, Jinwen |
| author_facet | Deng, Jingyang Shen, Zhengyang Wang, Boyang Su, Lixin Cheng, Suqi Nie, Ying Wang, Junfeng Yin, Dawei Ma, Jinwen |
| contents | The development of Long-Context Large Language Models (LLMs) has markedly advanced natural language processing by facilitating the process of textual data across long documents and multiple corpora. However, Long-Context LLMs still face two critical challenges: The lost in the middle phenomenon, where crucial middle-context information is likely to be missed, and the distraction issue that the models lose focus due to overly extended contexts. To address these challenges, we propose the Context Filtering Language Model (FltLM), a novel integrated Long-Context LLM which enhances the ability of the model on multi-document question-answering (QA) tasks. Specifically, FltLM innovatively incorporates a context filter with a soft mask mechanism, identifying and dynamically excluding irrelevant content to concentrate on pertinent information for better comprehension and reasoning. Our approach not only mitigates these two challenges, but also enables the model to operate conveniently in a single forward pass. Experimental results demonstrate that FltLM significantly outperforms supervised fine-tuning and retrieval-based methods in complex QA scenarios, suggesting a promising solution for more accurate and reliable long-context natural language understanding applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_06886 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | FltLM: An Intergrated Long-Context Large Language Model for Effective Context Filtering and Understanding Deng, Jingyang Shen, Zhengyang Wang, Boyang Su, Lixin Cheng, Suqi Nie, Ying Wang, Junfeng Yin, Dawei Ma, Jinwen Computation and Language The development of Long-Context Large Language Models (LLMs) has markedly advanced natural language processing by facilitating the process of textual data across long documents and multiple corpora. However, Long-Context LLMs still face two critical challenges: The lost in the middle phenomenon, where crucial middle-context information is likely to be missed, and the distraction issue that the models lose focus due to overly extended contexts. To address these challenges, we propose the Context Filtering Language Model (FltLM), a novel integrated Long-Context LLM which enhances the ability of the model on multi-document question-answering (QA) tasks. Specifically, FltLM innovatively incorporates a context filter with a soft mask mechanism, identifying and dynamically excluding irrelevant content to concentrate on pertinent information for better comprehension and reasoning. Our approach not only mitigates these two challenges, but also enables the model to operate conveniently in a single forward pass. Experimental results demonstrate that FltLM significantly outperforms supervised fine-tuning and retrieval-based methods in complex QA scenarios, suggesting a promising solution for more accurate and reliable long-context natural language understanding applications. |
| title | FltLM: An Intergrated Long-Context Large Language Model for Effective Context Filtering and Understanding |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2410.06886 |