In this seminar, the representation and differentiation of multiple tasks in neural systems will be examined, drawing on both theoretical models and experimental data. The presentation will begin with a brief overview of previous work on recurrent neural networks and structural representations of computation, followed by recent analyses of population calcium imaging data. The focus is on two complementary themes: the relationship between anatomical neurons and functional network structure, and the temporal dynamics of trial-to-trial activity across tasks. Through a range of geometric, statistical, and structure-aware analyses, the investigation explores the extent to which neural activity patterns repeat, generalize, or reorganize in response to task context. This ongoing work seeks to uncover invariant substructures and context-specific modulations that support flexible    computation in the brain

Advisor : Prof.  Omri Barak