Research map / Probabilistic and scientific modeling / Dynamical systems
MnemoDyn: Learning Resting State Dynamics from $40$K FMRI sequences
Sourav Pal, Viet Luong, Hoseok Lee, Tingting Dan, Guorong Wu, Richard Davidson, Won Hwa Kim, Vikas Singh
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| scale pretraining data from 1000 to 20000 UKB subjects (ablation) | MnemoDyn · HCP-Aging sex | accuracy: 0.706 -> 0.793 | Improved |
| scale pretraining data from 1000 to 20000 UKB subjects (ablation) | MnemoDyn · HCP-Aging age | MSE: 1.023 -> 1.012 | Improved |
| increase operator blocks from 1 to 2 (ablation) | MnemoDyn · HCP-Aging age | MSE: 0.914 -> 0.903 | Improved |
| increase operator blocks from 1 to 2 (ablation) | MnemoDyn · HCP-Aging sex | accuracy: 0.768 -> 0.8010 | Improved |
| Mask pretraining instead of Denoise pretraining (ablation) | MnemoDyn · HCP-Aging sex | accuracy (%): 80.2 -> 83.10 | Improved |
| Brain-JEPA-style masking instead of plain Mask pretraining (ablation) | MnemoDyn · HCP-Aging | age MSE: 0.90 -> 0.90 | No clear effect |
| different wavelet families (ablation) | MnemoDyn · downstream tasks | performance | No clear effect |
| stack blocks with varying latent dimensions and add residual connections | MnemoDyn · representation | performance | Improved |
| wavelet-operator model instead of Brain-JEPA | ADNI NC/MCI · ADNI NC vs MCI | accuracy (%): 76.84 -> 96.12 | Improved |
| MnemoDyn instead of Brain-JEPA | UK Biobank sex · UK Biobank sex | accuracy (%): 88.17 -> 88.40 | No clear effect |
Closest papers by meaning
- A scalable generative model for dynamical system reconstruction from neuroimaging data (NeurIPS 2024 · similarity 0.88)
- FAST-Brain: A Flow-Aligned Spatio-Temporal Surrogate Brain Model (NeurIPS 2026 · similarity 0.88)
- BrainWorld: A Structural-Prior-Conditioned Generative Model for Whole-Brain 4D fMRI Dynamics (NeurIPS 2026 · similarity 0.87)
- BrainVista: Modeling Naturalistic Brain Dynamics as Multimodal Next-Token Prediction (NeurIPS 2026 · similarity 0.87)
- Dynamical Modeling of Behaviorally Relevant Spatiotemporal Patterns in Neural Imaging Data (ICML 2025 · similarity 0.86)
- Brain Harmony: A Multimodal Foundation Model Unifying Morphology and Function into 1D Tokens (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
