Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Mian, Jin, Lifeng, Song, Linfeng, Mi, Haitao, Yu, Dong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2401.10353
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929215679496192
author Zhang, Mian
Jin, Lifeng
Song, Linfeng
Mi, Haitao
Yu, Dong
author_facet Zhang, Mian
Jin, Lifeng
Song, Linfeng
Mi, Haitao
Yu, Dong
contents One critical issue for chat systems is to stay consistent about preferences, opinions, beliefs and facts of itself, which has been shown a difficult problem. In this work, we study methods to assess and bolster utterance consistency of chat systems. A dataset is first developed for studying the inconsistencies, where inconsistent dialogue responses, explanations of the inconsistencies, and recovery utterances are authored by annotators. This covers the life span of inconsistencies, namely introduction, understanding, and resolution. Building on this, we introduce a set of tasks centered on dialogue consistency, specifically focused on its detection and resolution. Our experimental findings indicate that our dataset significantly helps the progress in identifying and resolving conversational inconsistencies, and current popular large language models like ChatGPT which are good at resolving inconsistencies however still struggle with detection.
format Preprint
id arxiv_https___arxiv_org_abs_2401_10353
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Inconsistent dialogue responses and how to recover from them
Zhang, Mian
Jin, Lifeng
Song, Linfeng
Mi, Haitao
Yu, Dong
Computation and Language
One critical issue for chat systems is to stay consistent about preferences, opinions, beliefs and facts of itself, which has been shown a difficult problem. In this work, we study methods to assess and bolster utterance consistency of chat systems. A dataset is first developed for studying the inconsistencies, where inconsistent dialogue responses, explanations of the inconsistencies, and recovery utterances are authored by annotators. This covers the life span of inconsistencies, namely introduction, understanding, and resolution. Building on this, we introduce a set of tasks centered on dialogue consistency, specifically focused on its detection and resolution. Our experimental findings indicate that our dataset significantly helps the progress in identifying and resolving conversational inconsistencies, and current popular large language models like ChatGPT which are good at resolving inconsistencies however still struggle with detection.
title Inconsistent dialogue responses and how to recover from them
topic Computation and Language
url https://arxiv.org/abs/2401.10353