LoGU: Long-form Generation with Uncertainty Expressions

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Yang, Ruihan, Zhang, Caiqi, Zhang, Zhisong, Huang, Xinting, Yang, Sen, Collier, Nigel, Yu, Dong, Yang, Deqing
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910984870821888
author Yang, Ruihan
Zhang, Caiqi
Zhang, Zhisong
Huang, Xinting
Yang, Sen
Collier, Nigel
Yu, Dong
Yang, Deqing
author_facet Yang, Ruihan
Zhang, Caiqi
Zhang, Zhisong
Huang, Xinting
Yang, Sen
Collier, Nigel
Yu, Dong
Yang, Deqing
contents While Large Language Models (LLMs) demonstrate impressive capabilities, they still struggle with generating factually incorrect content (i.e., hallucinations). A promising approach to mitigate this issue is enabling models to express uncertainty when unsure. Previous research on uncertainty modeling has primarily focused on short-form QA, but realworld applications often require much longer responses. In this work, we introduce the task of Long-form Generation with Uncertainty(LoGU). We identify two key challenges: Uncertainty Suppression, where models hesitate to express uncertainty, and Uncertainty Misalignment, where models convey uncertainty inaccurately. To tackle these challenges, we propose a refinement-based data collection framework and a two-stage training pipeline. Our framework adopts a divide-and-conquer strategy, refining uncertainty based on atomic claims. The collected data are then used in training through supervised fine-tuning (SFT) and direct preference optimization (DPO) to enhance uncertainty expression. Extensive experiments on three long-form instruction following datasets show that our method significantly improves accuracy, reduces hallucinations, and maintains the comprehensiveness of responses.
format Preprint
id arxiv_https___arxiv_org_abs_2410_14309
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LoGU: Long-form Generation with Uncertainty Expressions
Yang, Ruihan
Zhang, Caiqi
Zhang, Zhisong
Huang, Xinting
Yang, Sen
Collier, Nigel
Yu, Dong
Yang, Deqing
Computation and Language
Artificial Intelligence
While Large Language Models (LLMs) demonstrate impressive capabilities, they still struggle with generating factually incorrect content (i.e., hallucinations). A promising approach to mitigate this issue is enabling models to express uncertainty when unsure. Previous research on uncertainty modeling has primarily focused on short-form QA, but realworld applications often require much longer responses. In this work, we introduce the task of Long-form Generation with Uncertainty(LoGU). We identify two key challenges: Uncertainty Suppression, where models hesitate to express uncertainty, and Uncertainty Misalignment, where models convey uncertainty inaccurately. To tackle these challenges, we propose a refinement-based data collection framework and a two-stage training pipeline. Our framework adopts a divide-and-conquer strategy, refining uncertainty based on atomic claims. The collected data are then used in training through supervised fine-tuning (SFT) and direct preference optimization (DPO) to enhance uncertainty expression. Extensive experiments on three long-form instruction following datasets show that our method significantly improves accuracy, reduces hallucinations, and maintains the comprehensiveness of responses.
title LoGU: Long-form Generation with Uncertainty Expressions
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2410.14309