Back
Top 0.8%
12.0%
Top 0.5%
12.0%
Top 2%
11.2%
Top 55%
11.2%
Top 20%
9.2%
Top 3%
4.2%
Top 4%
3.9%
Top 6%
2.8%
Top 36%
2.0%
Top 7%
2.0%
Top 2%
2.0%
Top 5%
1.6%
Top 0.2%
1.6%
Top 4%
1.2%
Top 2%
1.2%
Top 3%
1.2%
Top 25%
1.2%
Top 2%
0.7%
Top 10%
0.7%
Top 54%
0.7%
Top 2%
0.7%
The Forgotten Shield: Safety Grafting in Parameter-Space for Medical MLLMs
2025-12-22
health systems and quality improvement
Title + abstract only
View on medRxiv
Show abstract
Medical Multimodal Large Language Models (Medical MLLMs) have achieved remarkable progress in specialized medical tasks; however, research into their safety has lagged, posing potential risks for real-world deployment. In this paper, we first establish a multidimensional evaluation framework to systematically benchmark the safety of current SOTA Medical MLLMs. Our empirical analysis reveals pervasive vulnerabilities across both general and medical-specific safety dimensions in existing models, p...
Predicted journal destinations
1
npj Digital Medicine
85 training papers
2
PLOS Digital Health
88 training papers
3
Journal of the American Medical Informatics Association
53 training papers
4
PLOS ONE
1737 training papers
5
Scientific Reports
701 training papers
6
Journal of Biomedical Informatics
37 training papers
7
BMC Medical Informatics and Decision Making
36 training papers
8
Journal of Medical Internet Research
81 training papers
9
Nature Communications
483 training papers
10
PLOS Computational Biology
141 training papers
11
Royal Society Open Science
49 training papers
12
Computers in Biology and Medicine
39 training papers
13
Patterns
15 training papers
14
BMC Health Services Research
43 training papers
15
JMIRx Med
29 training papers
16
Communications Medicine
63 training papers
17
Frontiers in Public Health
135 training papers
18
Sensors
18 training papers
19
Nature
58 training papers
20
eLife
262 training papers
21
BMJ Open Quality
15 training papers