Plasticity of neural representations

A major goal in systems neuroscience is to connect the activity of populations of neurons to specific behaviors. However, large scale recordings of neural activity during the execution of learned tasks and during the experience of familiar stimuli have revealed that neural activity patterns continually change over extended periods. This so-called Representational Drift is not accompanied by obvious alterations in behavior, learning or systemic physiology, which raises profound questions about its origin and its implications for learning and memory. For example, textbook theories of learning and memory assert that stable memories require stable relationships between neural activity and learned associations. Representational drift brings these theories into question, while raising practical problems for understanding neural data, designing experiments and developing technology such as brain-machine interfaces.  

This project uses a mix of data science, computational modelling and theory, and collaboration with experimentalists to understand the causes and implications of Representational Drift. We use a variety of statistical methods as well as modelling and analysis of artificial neural networks to generate and test hypotheses. We work closely with experimentalists in Harvard Medical School and UCL, and wish to find experimental partners in the NIH to further this research.  Key skills include proficiency in numerical methods, simulation, strong coding skills and a working knowledge of advanced statistical methods, including generalized linear models and Bayesian inference.  

Key recent publications include:  
Micou, C., & O'Leary, T. (2023). Representational drift as a window into neural and behavioural plasticity. Current opinion in neurobiology, 81, 102746. https://www.sciencedirect.com/science/article/pii/S0959438823000715  Rule, M. E., & O’Leary, T. (2022). 

Self-healing codes: How stable neural populations can track continually reconfiguring neural representations. Proceedings of the National Academy of Sciences, 119(7), e2106692119. https://www.pnas.org/doi/abs/10.1073/pnas.2106692119