Leveraging Logical Rules in Knowledge Editing: A Cherry on the Top

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Cheng, Keyuan, Ali, Muhammad Asif, Yang, Shu, Lin, Gang, Zhai, Yuxuan, Fei, Haoyang, Xu, Ke, Yu, Lu, Hu, Lijie, Wang, Di
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916261046255616
author Cheng, Keyuan
Ali, Muhammad Asif
Yang, Shu
Lin, Gang
Zhai, Yuxuan
Fei, Haoyang
Xu, Ke
Yu, Lu
Hu, Lijie
Wang, Di
author_facet Cheng, Keyuan
Ali, Muhammad Asif
Yang, Shu
Lin, Gang
Zhai, Yuxuan
Fei, Haoyang
Xu, Ke
Yu, Lu
Hu, Lijie
Wang, Di
contents Multi-hop Question Answering (MQA) under knowledge editing (KE) is a key challenge in Large Language Models (LLMs). While best-performing solutions in this domain use a plan and solve paradigm to split a question into sub-questions followed by response generation, we claim that this approach is sub-optimal as it fails for hard to decompose questions, and it does not explicitly cater to correlated knowledge updates resulting as a consequence of knowledge edits. This has a detrimental impact on the overall consistency of the updated knowledge. To address these issues, in this paper, we propose a novel framework named RULE-KE, i.e., RULE based Knowledge Editing, which is a cherry on the top for augmenting the performance of all existing MQA methods under KE. Specifically, RULE-KE leverages rule discovery to discover a set of logical rules. Then, it uses these discovered rules to update knowledge about facts highly correlated with the edit. Experimental evaluation using existing and newly curated datasets (i.e., RKE-EVAL) shows that RULE-KE helps augment both performances of parameter-based and memory-based solutions up to 92% and 112.9%, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2405_15452
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Leveraging Logical Rules in Knowledge Editing: A Cherry on the Top
Cheng, Keyuan
Ali, Muhammad Asif
Yang, Shu
Lin, Gang
Zhai, Yuxuan
Fei, Haoyang
Xu, Ke
Yu, Lu
Hu, Lijie
Wang, Di
Computation and Language
Artificial Intelligence
Machine Learning
Multi-hop Question Answering (MQA) under knowledge editing (KE) is a key challenge in Large Language Models (LLMs). While best-performing solutions in this domain use a plan and solve paradigm to split a question into sub-questions followed by response generation, we claim that this approach is sub-optimal as it fails for hard to decompose questions, and it does not explicitly cater to correlated knowledge updates resulting as a consequence of knowledge edits. This has a detrimental impact on the overall consistency of the updated knowledge. To address these issues, in this paper, we propose a novel framework named RULE-KE, i.e., RULE based Knowledge Editing, which is a cherry on the top for augmenting the performance of all existing MQA methods under KE. Specifically, RULE-KE leverages rule discovery to discover a set of logical rules. Then, it uses these discovered rules to update knowledge about facts highly correlated with the edit. Experimental evaluation using existing and newly curated datasets (i.e., RKE-EVAL) shows that RULE-KE helps augment both performances of parameter-based and memory-based solutions up to 92% and 112.9%, respectively.
title Leveraging Logical Rules in Knowledge Editing: A Cherry on the Top
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2405.15452