.png){kind=icon}
.png){kind=icon}
Return
Brain-inspired large model mindreading
.png){kind=icon}
OA
AI
DOI：10.1016/j.neuroimage.2026.122108.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Human responses to visual ToM questions are more concise and certain. • MLLMI showed stronger bilateral precuneus and middle temporal gyrus activation. • Enhanced functional connectivity in task-coordination and attention networks in MLLMI. • Neural-based Transformer model achieved 78.6% classification accuracy. • We propose a Knowledge-Thinking-Adaptation (KTA) roadmap for AI visual ToM.
Keywords:
Multimodal large language models
Theory of mind
Visual question answering
Mentalization
Brain-inspired intelligence
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.5
Papers:
1.3K
Citations:
10.6W