Sleep Science Podcast S3E5 - Anna Schapiro - Can we model the brain?
Jul 2, 2025
In this engaging discussion, Anna Schapiro, an Associate Professor of Psychology at the University of Pennsylvania, shares insights from her research on computational modeling and memory. She dives into how neural networks help us understand memory representation in the brain and the crucial role of sleep in memory consolidation. Topics include the dynamics of memory reactivation during REM sleep, the complex interplay between the hippocampus and neocortex, and how bilingualism impacts memory processes—all through groundbreaking computational theories.
AI Snips
Chapters
Transcript
Episode notes
Memory Storage via Synaptic Weights
- Memories are stored by changing synaptic weights between neurons across different brain areas.
- Neural network models mimic this process and help explain various types of memory storage and learning.
Complementary Learning Systems Theory
- Complementary Learning Systems theory explains memory by fast hippocampal storage and slow integration into neocortex.
- Replay during sleep may enable the hippocampus to teach cortex and consolidate long-term memories.
Different Sleep Stages Aid Memory
- REM sleep may help revisit and strengthen older memories while non-REM focuses on recent ones.
- Oscillations in sleep enable cleaning and strengthening of memories by modulating neuronal activity.