Research map / Label-efficient and robust learning / Self-supervised and contrastive
Self-Supervised Learning of Brain Dynamics from Broad Neuroimaging Data
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use causal sequence modeling (CSM) pre-training instead of autoencoding | pre-trained transformer · HCP N=1 | accuracy: 60.2 -> 64.1 | Helped |
| use CSM pre-training instead of autoencoding | pre-trained transformer · HCP N=48 | accuracy: 91.5 -> 94.8 | Helped |
| use CSM pre-training instead of autoencoding | pre-trained transformer · MDTB N=11 | accuracy: 86.7 -> 90.0 | Helped |
| pre-training vs linear baseline trained from scratch | CSM · HCP N=48 | accuracy: 51.9 -> 94.8 | Helped |
| pre-training vs decoding head p(X) trained from scratch (small data) | Autoencoding · HCP N=6 | accuracy | No clear effect |
| pre-training vs decoding head p(X) trained from scratch (large data) | CSM · HCP N=24 | accuracy: 34.9 -> 92.7 | Helped |
| larger embedding dimension (192 to 768) (ablation) | all frameworks · upstream validation | validation loss | Helped |
| increase model depth (ablation) | all frameworks · upstream validation | validation loss | No clear effect |
| initialize from pre-trained language models (GPT-2, BERT) (ablation) | CSM / Seq-BERT · upstream validation | validation loss: 0.73 (GPT-2), 0.80 (BERT) | No clear effect |
Closest papers by meaning
- BrainLM: A foundation model for brain activity recordings (ICLR 2024 · similarity 0.89)
- The Brain's Bitter Lesson: Scaling Speech Decoding With Self-Supervised Learning (ICML 2025 · similarity 0.88)
- CSBrain: A Cross-scale Spatiotemporal Brain Foundation Model for EEG Decoding (NeurIPS 2025 · similarity 0.86)
- Leveraging unlabelled data for generalizable neural population decoding (NeurIPS 2026 · similarity 0.86)
- A Scalable, Causal, and Energy Efficient Framework for Neural Decoding with Spiking Neural Networks (NeurIPS 2025 · similarity 0.86)
- BrainWorld: A Structural-Prior-Conditioned Generative Model for Whole-Brain 4D fMRI Dynamics (NeurIPS 2026 · 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.