Back

Drug or Pokemon? An analysis of the ability of large language models to discern fabricated medications

2026-01-13 health informatics Title + abstract only
View on medRxiv
Show abstract

BackgroundThe use of large language models (LLMs) is increasing in the medical field; however, LLMs are often subject to "confabulations." Notably, LLMs have vulnerability to adversarial attacks, or fabricated details within prompts, which is concerning given both health misinformation and inadvertent errors in the medical record. This purpose of this study was to determine the effect of adversarial attacks by embedding one fabricated medication into a list of existing medicines. MethodsA total...

Predicted journal destinations

1
Journal of the American Medical Informatics Association
53 training papers
Top 0.5% 17.4%
2
Journal of Medical Internet Research
81 training papers
Top 0.6% 10.8%
3
JAMIA Open
35 training papers
Top 0.3% 8.9%
4
BMC Medical Informatics and Decision Making
36 training papers
Top 0.6% 7.7%
5
Scientific Reports
701 training papers
Top 39% 6.3%
6
npj Digital Medicine
85 training papers
Top 3% 6.3%
7
Journal of Biomedical Informatics
37 training papers
Top 1% 6.3%
8
PLOS Digital Health
88 training papers
Top 4% 5.9%
9
PLOS ONE
1737 training papers
Top 82% 4.9%
10
International Journal of Medical Informatics
25 training papers
Top 0.6% 4.1%
11
JMIR Medical Informatics
16 training papers
Top 3% 1.9%
12
BMC Medical Research Methodology
41 training papers
Top 1% 1.9%
13
Communications Medicine
63 training papers
Top 2% 1.5%
14
JMIR Formative Research
31 training papers
Top 3% 1.3%
15
Computers in Biology and Medicine
39 training papers
Top 8% 0.9%
16
Frontiers in Digital Health
18 training papers
Top 3% 0.9%
17
BMJ Open
553 training papers
Top 63% 0.9%
18
Heliyon
57 training papers
Top 9% 0.9%
19
JMIR Public Health and Surveillance
45 training papers
Top 6% 0.9%
20
Patterns
15 training papers
Top 2% 0.9%
21
JAMA Network Open
125 training papers
Top 31% 0.7%
22
Frontiers in Public Health
135 training papers
Top 34% 0.7%
23
Psychiatry Research
33 training papers
Top 6% 0.7%