This paper is in the following e-collection/theme issue:

Artificial Intelligence (1723) Generative Language Models Including ChatGPT (492) Epublishing and Open Access (96) Reviews in Medical Education (71) New Methods and Approaches in Medical Education (310) Methods and Instruments in Medical Informatics (164) Chatbots and Conversational Agents (627)

Published on in Vol 26 (2024)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/53164, first published .
Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis

Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis

Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis

Authors of this article:

Mikaël Chelli1 Author Orcid Image ;   Jules Descamps2 Author Orcid Image ;   Vincent Lavoué1 Author Orcid Image ;   Christophe Trojani1 Author Orcid Image ;   Michel Azar1 Author Orcid Image ;   Marcel Deckert3 Author Orcid Image ;   Jean-Luc Raynier1 Author Orcid Image ;   Gilles Clowez1 Author Orcid Image ;   Pascal Boileau1 Author Orcid Image ;   Caroline Ruetsch-Chelli3 Author Orcid Image
Article Authors Cited by (39) Tweetations (11) Metrics
All Time | Last Month | Last Six Months | Last Year
JMIR Publications JMIR Publications @jmirpub

New in JMIR: Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis https://t.co/rLPlQ7L7tP https://t.co/BF6TDijTiB

3:51 PM · May 22, 2024

3
3
Journal of Medical Internet Research Journal of Medical Internet Research @JMIR_

RT @jmirpub: New in JMIR: Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis https…

3:51 PM · May 22, 2024

3
3
Susan J Bondy Susan J Bondy @SueBondy_TO

Nota bene on performance of ChatGPT and comparable... https://t.co/ZkyUQ5Lbaj

4:16 PM · May 22, 2024

0
0
Tom Bellis Tom Bellis @tommybellis

Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis https://t.co/0wbHCS7mWU https://t.co/kwZumP4PYt

4:53 PM · May 22, 2024

1
0
YMED-X YMED-X @YmedX269

RT @jmirpub: New in JMIR: Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis https…

1:50 AM · May 23, 2024

3
3
Dechao Feng Dechao Feng @DechaoF

RT @jmirpub: New in JMIR: Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis https…

1:51 AM · May 23, 2024

3
3
Sara Abu Khudair Sara Abu Khudair @SaraAbuKhudair1

Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis Link: https://t.co/HOEkqbKK9z The hallucination rates (misleading references) were 39.6% (55/139) for GPT-3.5 , 28.6% (34/119) for GPT-4, and 91.4% (95/104) for Bard.

10:11 AM · May 24, 2024

1
0
The vOICe vision BCI ???????? The vOICe vision BCI ???????? @seeingwithsound

GPT 4 hallucination rate is 28.6% on a simple task: citing title, author, and year of publication https://t.co/pSgiazxxKl Hallucination rates and reference accuracy of ChatGPT and Bard for systematic reviews: Comparative analysis https://t.co/bZVHFCkuCS #AI #LLM

3:58 PM · Jul 04, 2024

0
0
i am i am @iam833

@Ralphblunts @ThatGuyM88 @AntiDisinfo86 Why would I save an incorrect output ? This is common knowledge btw you’re pretty much asking me “does it rain sometimes?” “Take a picture of it raining” rn Here’s a pubmed article on LLM hallucinations https://t.co/B2yUj6DF6x https://t.co/oRJfEMQgfo

6:43 PM · Aug 11, 2024

0
0
T T @T42247325

@blejdea_alex @Ziapaws @adokogasou What's the point in a research tool that provides inaccurate, unsourced information? 13% precision and 28% hallucination rate on GPT 4 is pretty awful as a research tool. ( From a more recent study) https://t.co/IRVUeNRDp2

10:58 AM · Oct 13, 2024

3
0
MentalHealthAI MentalHealthAI @MentalHealthAI

Are you sure you want AI giving you therapeutic feedback? https://t.co/UOk50zWS7q

3:23 AM · Nov 24, 2024

0
0
кадмій кадмій @kadmii1

@GeekProgrammer @Erickschultz11 @GreyGhostMN @DrNeilStone here, apparently it's too hard to open the relevant link in the first paragraph in an article that summarizes the key finding in the study cited they weren't kidding when they say your brain is like a muscle https://t.co/jq4ZyiveWK https://t.co/jKkJa1BZsN

12:32 PM · Jan 07, 2025

0
0