Research

I am interested in the interaction of multiple learning pathways, both from an algorithmic perspective, as well as uncovering the multi-regional circuits that underlie learning. I study the neural dynamics that emerge over learning to support new task-relevant computations, with a focus on how feedback - sensory and internal feedback loops - can flexibly modulate effective cortical dynamics, and how these input-driven dynamics are learnt via multiple forms of errors.

Across projects, I integrate data-driven and/or normative modeling of behavioral strategies to analyze neural data in light of the actual computations at play and to better understand variability across individuals.

I am also motivated by a dynamical systems perspective that understanding neural computation requires not only identifying neural representations, but also uncovering the rules that govern how those representations evolve over time.

Projects

Data-driven discovery of shared latent dynamics

Identifying nonlinear dynamics from data, and aligning neural population activity across sessions.

Dynamical constraints on learning

How feedback, intrinsic dynamics, and controllability constrain learning timescales for fast adaptation.

Population dynamics in the cerebellar cortex

Network structure shapes inhibitory network dynamics and representational transformations across cerebellar circuits.

Role of perceptual uncertainty in reward-driven learning

Computational signatures of reinforcement learning when state inference is uncertain due to noisy perception.

Data-driven models for control of biomechanical bodies

Closed-loop neural controllers with biologically grounded sensory feedback for realistic locomotion in Drosophila.

Learning without plasticity

Fast adaptation via input-driven reorganization of dynamics, contextual inference, and flexible inter-area associations.