brain-dit-universal-multi

---
name: brain-dit-universal-multi
description: "Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining Activation: spiking neural network, neuromorphic computing, neural computation."
---

# Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining

## Overview
Current fMRI foundation models primarily rely on a limited range of brain states and mismatched pretraining tasks, restricting their ability to learn generalized representations across diverse brain states. We present \textit{Brain-DiT}, a universal multi-state fMRI foundation model pretrained on 349,898 sessions from 24 datasets spanning resting, task, naturalistic, disease, and sleep states. Unl...

## Source Paper
- **Title:** Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining
- **Authors:** Junfeng Xia, Wenhao Ye, Xuanye Pan et al.
- **arXiv:** 2604.12683v1
- **Published:** 2026-04-14
- **Categories:** cs.CV, q-bio.NC
- **PDF:** https://arxiv.org/pdf/2604.12683v1

## Key Concepts
- Neural Computation

## Core Contributions
1. Novel approach to neural computation
2. Innovative methodology for brain-inspired AI
3. Practical applications in neuromorphic systems

## Applications
- Neuromorphic hardware implementation
- Energy-efficient neural computation
- Brain-inspired cognitive systems

## References
- Junfeng Xia et al. (2026). "Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining." arXiv:2604.12683v1.

## Activation Keywords
- brain dit universal multi
- spiking neural network
- neuromorphic computing
- neural dynamics
- neural computation

---
*Generated: 2026-04-15*