Salvato in:
Dettagli Bibliografici
Autori principali: Koopman, Bevan, Zuccon, Guido
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2507.12296
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916846498742272
author Koopman, Bevan
Zuccon, Guido
author_facet Koopman, Bevan
Zuccon, Guido
contents Despite widespread debunking, many psychological myths remain deeply entrenched. This paper investigates whether Large Language Models (LLMs) mimic human behaviour of myth belief and explores methods to mitigate such tendencies. Using 50 popular psychological myths, we evaluate myth belief across multiple LLMs under different prompting strategies, including retrieval-augmented generation and swaying prompts. Results show that LLMs exhibit significantly lower myth belief rates than humans, though user prompting can influence responses. RAG proves effective in reducing myth belief and reveals latent debiasing potential within LLMs. Our findings contribute to the emerging field of Machine Psychology and highlight how cognitive science methods can inform the evaluation and development of LLM-based systems.
format Preprint
id arxiv_https___arxiv_org_abs_2507_12296
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Humans are more gullible than LLMs in believing common psychological myths
Koopman, Bevan
Zuccon, Guido
Human-Computer Interaction
Despite widespread debunking, many psychological myths remain deeply entrenched. This paper investigates whether Large Language Models (LLMs) mimic human behaviour of myth belief and explores methods to mitigate such tendencies. Using 50 popular psychological myths, we evaluate myth belief across multiple LLMs under different prompting strategies, including retrieval-augmented generation and swaying prompts. Results show that LLMs exhibit significantly lower myth belief rates than humans, though user prompting can influence responses. RAG proves effective in reducing myth belief and reveals latent debiasing potential within LLMs. Our findings contribute to the emerging field of Machine Psychology and highlight how cognitive science methods can inform the evaluation and development of LLM-based systems.
title Humans are more gullible than LLMs in believing common psychological myths
topic Human-Computer Interaction
url https://arxiv.org/abs/2507.12296