Does Less Hallucination Mean Less Creativity? An Empirical Investigation in LLMs

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Banerjee, Mohor, Wangsajaya, Nadya Yuki, Alsagoff, Syed Ali Redha, Tan, Min Sen, Chun, Zachary Choy Kit, Wei, Alvin Chan Guo
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911388539027456
author Banerjee, Mohor
Wangsajaya, Nadya Yuki
Alsagoff, Syed Ali Redha
Tan, Min Sen
Chun, Zachary Choy Kit
Wei, Alvin Chan Guo
author_facet Banerjee, Mohor
Wangsajaya, Nadya Yuki
Alsagoff, Syed Ali Redha
Tan, Min Sen
Chun, Zachary Choy Kit
Wei, Alvin Chan Guo
contents Large Language Models (LLMs) exhibit remarkable capabilities in natural language understanding and reasoning, but suffer from hallucination: the generation of factually incorrect content. While numerous methods have been developed to reduce hallucinations, their impact on creative generations remains unexplored. This gap is particularly critical for AI-assisted scientific discovery, which requires both factual accuracy and creative hypothesis generation. We investigate how three hallucination-reduction techniques: Chain of Verification (CoVe), Decoding by Contrasting Layers (DoLa), and Retrieval-Augmented Generation (RAG), affect creativity in LLMs. Evaluating multiple model families (LLaMA, Qwen, Mistral) at varying scales (1B - 70B parameters) on two creativity benchmarks (NeoCoder and CS4), we find that these methods have opposing effects on divergent creativity. CoVe enhances divergent thinking, DoLa suppresses it, and RAG shows minimal impact. Our findings provide guidance for selecting appropriate hallucination-reduction methods in scientific applications, where the balance between factual accuracy and creative exploration is crucial.
format Preprint
id arxiv_https___arxiv_org_abs_2512_11509
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Does Less Hallucination Mean Less Creativity? An Empirical Investigation in LLMs
Banerjee, Mohor
Wangsajaya, Nadya Yuki
Alsagoff, Syed Ali Redha
Tan, Min Sen
Chun, Zachary Choy Kit
Wei, Alvin Chan Guo
Computation and Language
Artificial Intelligence
Large Language Models (LLMs) exhibit remarkable capabilities in natural language understanding and reasoning, but suffer from hallucination: the generation of factually incorrect content. While numerous methods have been developed to reduce hallucinations, their impact on creative generations remains unexplored. This gap is particularly critical for AI-assisted scientific discovery, which requires both factual accuracy and creative hypothesis generation. We investigate how three hallucination-reduction techniques: Chain of Verification (CoVe), Decoding by Contrasting Layers (DoLa), and Retrieval-Augmented Generation (RAG), affect creativity in LLMs. Evaluating multiple model families (LLaMA, Qwen, Mistral) at varying scales (1B - 70B parameters) on two creativity benchmarks (NeoCoder and CS4), we find that these methods have opposing effects on divergent creativity. CoVe enhances divergent thinking, DoLa suppresses it, and RAG shows minimal impact. Our findings provide guidance for selecting appropriate hallucination-reduction methods in scientific applications, where the balance between factual accuracy and creative exploration is crucial.
title Does Less Hallucination Mean Less Creativity? An Empirical Investigation in LLMs
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2512.11509