Probing cerebello-cortical motor learning via recurrent neural networks – CC-RNN

This project develops models inspired by artificial intelligence to study motor learning from high-dimensional neural data. The main model consists of two regions with different architectures, inspired respectively by the cerebellum and the motor cortex, and is trained to perform a motor task. It reproduces many results observed in classical motor adaptation studies. Like many other models of the motor system using neural networks, this model relies on supervised learning algorithms providing detailed error information to guide learning. Mathematically novel tensor-based data science tools were also developed to track changes in neural representations over learning, allowing the models to be tested and compared with experimental data.

The main result of this project is a theoretical framework for how the cerebellum and the neocortex are able to work together to control learning during behaviour. In this model, the cerebellum rapidly learns an internal error signal that can be used to both correct and slowly update neocortical motor memories. This results in a system in which learning is distributed across the two regions at different time scales, yet which work together as a unified system. A secondary result is series of tensor-based dimensionality reduction methods that can be used to track how neural representations change over learning. These methods can be used in the future to test our distributed learning model against data, and moreover have widespread applicability to trial-structured data in neuroscience.

The main focus of this project is to develop a theoretical model of how cerebellar error-based learning can modify neocortical motor memories during movement.

The scientific production of this project will be multiple scientific articles on cerebello-cortical interaction and cerebellar reinforcement learning for cognition.

Accumulating evidence suggests that naturalistic behaviours are driven by interactions between brain regions traditionally studied in isolation. For example, recent work has shown that the cerebellum and the motor cortex, two highly interconnected regions that are each crucial for motor behaviour, coordinate during motor learning and execution. However, progress towards understanding cerebello-cortical interaction has been hindered by the drastically different circuitries of these regions, coupled with the inaccessibility of the deep subcortical pathways connecting them. As a result, it remains unclear how to integrate leading theories of cerebellar and cortical function in motor control.

Recent theoretical and experimental evidence has argued that the motor cortex flexibly generates the spatiotemporal patterns necessary for complex movements by exploiting its rich dynamics. This concept is strongly influenced by the emergence of recurrent neural networks (RNNs), a class of AI-inspired models that can be taught to solve a wide range of motor and cognitive tasks via changes in recurrent connectivity. In contrast, the feedforward, evolutionarily-conserved circuitry of the cerebellum is thought to be optimized for learning sensorimotor relationships and error-based adaptation of movements. Yet this classic view of cerebellar motor control is not able to explain recent evidence of reward information in cerebellar Purkinje cells, nor its mysterious role in cognition.

I propose to use RNNs to probe how these two distinct neural circuits work together during motor learning. I will extend the traditional neocortical-like RNN architecture to incorporate a cerebellar-like module capturing its divergent-convergent structure and the sparser interregional pathways. I will train this cerebello-cortical RNN (CC-RNN) to perform motor and cognitive tasks using machine learning methods, and will use systems neuroscience and statistical learning tools to analyze the resulting learned structural and functional properties. With this approach, I will address four questions that have been challenging to answer with experiments or traditional circuit modeling: 1) how cerebellar and motor cortical representations co-emerge during learning, 2) how cerebellar adaptation reshapes motor representations in the neocortex, 3) the impact of cerebello-cortical scaling on complex behaviours, and 4) the role of cerebellar adaptation in cognition. More broadly, this project will lead to a fuller understanding of how cortical and subcortical regions interact to produce complex behaviours.