New SOC framework enhances fMRI representation learning

Researchers have developed a new framework for learning representations from functional magnetic resonance imaging (fMRI) data by framing self-supervised learning as a Stochastic Optimal Control (SOC) problem. This approach models brain activity as continuous-time latent dynamics, optimizing a control policy to handle the temporal irregularity and noise inherent in fMRI signals. The method unifies masked autoencoding and joint-embedding prediction techniques, demonstrating state-of-the-art performance on downstream applications without requiring simulation. AI

IMPACT This novel SOC-based approach could improve the robustness and efficiency of AI models processing complex biological data like fMRI.

RANK_REASON The cluster contains an academic paper detailing a new method for representation learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• arXiv
• Byung Hoon Kim
• functional magnetic resonance imaging
• MAE
• self-supervised learning

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →