Saved in:
Bibliographic Details
Main Authors: Yuan, Xiaowei, Yang, Zhao, Wang, Yequan, Liu, Shengping, Zhao, Jun, Liu, Kang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.11893
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917733720915968
author Yuan, Xiaowei
Yang, Zhao
Wang, Yequan
Liu, Shengping
Zhao, Jun
Liu, Kang
author_facet Yuan, Xiaowei
Yang, Zhao
Wang, Yequan
Liu, Shengping
Zhao, Jun
Liu, Kang
contents Large language models internalize enormous parametric knowledge during pre-training. Concurrently, realistic applications necessitate external contextual knowledge to aid models on the underlying tasks. This raises a crucial dilemma known as knowledge conflicts, where the contextual knowledge clashes with the However, existing decoding works are specialized in resolving knowledge conflicts and could inadvertently deteriorate performance in absence of conflicts. In this paper, we propose an adaptive decoding method, termed as contextual information-entropy constraint decoding (COIECD), to discern whether the knowledge conflicts occur and resolve them. It can improve the model's faithfulness to conflicting context, and simultaneously maintain high performance among non- Our experiments show that COIECD exhibits strong performance and robustness over knowledge conflicts in realistic datasets. Code is available.
format Preprint
id arxiv_https___arxiv_org_abs_2402_11893
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy Constraint
Yuan, Xiaowei
Yang, Zhao
Wang, Yequan
Liu, Shengping
Zhao, Jun
Liu, Kang
Artificial Intelligence
Large language models internalize enormous parametric knowledge during pre-training. Concurrently, realistic applications necessitate external contextual knowledge to aid models on the underlying tasks. This raises a crucial dilemma known as knowledge conflicts, where the contextual knowledge clashes with the However, existing decoding works are specialized in resolving knowledge conflicts and could inadvertently deteriorate performance in absence of conflicts. In this paper, we propose an adaptive decoding method, termed as contextual information-entropy constraint decoding (COIECD), to discern whether the knowledge conflicts occur and resolve them. It can improve the model's faithfulness to conflicting context, and simultaneously maintain high performance among non- Our experiments show that COIECD exhibits strong performance and robustness over knowledge conflicts in realistic datasets. Code is available.
title Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy Constraint
topic Artificial Intelligence
url https://arxiv.org/abs/2402.11893