
Brain Inspired BI 226 Tatiana Engel: The High and Low Dimensional Brain
6 snips
Dec 3, 2025 Tatiana Engel is a computational neuroscientist at Princeton, focusing on neural manifolds and brain dynamics. She unveils how latent circuits connect complex cognitive functions with neural connectivity. Discussing the International Brain Laboratory, she explains its mission to standardize data sharing, enhancing collaboration in neuroscience. Tatiana also explores the surprising finding of longer timescales in subcortical neurons compared to the cortex, revealing universal patterns across brain regions. Her insights bridge theoretical models and practical applications in understanding brain activity.
AI Snips
Chapters
Transcript
Episode notes
Joining A Global Whole‑Brain Project
- Tatiana Engel joined the International Brain Laboratory (IBL) in 2021 after they invited theory labs to fill a slot.
- IBL centralized data pipelines and staff engineers so collaborators could access standardized, well-documented whole-brain Neuropixel data quickly.
Manifolds Arise From Circuit Dynamics
- Engel defines a neural manifold as the restricted set of states neural activity explores during behavior.
- Manifolds arise from recurrent circuit dynamics and reflect which states the connectivity makes accessible.
Latent Circuits Reveal Low‑Rank Connectivity
- Latent circuits are low‑dimensional dynamical models inferred to both capture neural activity and reproduce task behavior.
- Fitting latent circuits uncovers a low‑rank connectivity pattern embedded within the larger network that generates task dynamics.
