Aswathi Thrivikraman

PhD student

Aswathi Thrivikraman

Anatomically constrained recurrent neural networks, planning and internally generated cognition.

About

My work uses anatomically constrained recurrent neural networks to study how cognitive representations emerge and organise across neural systems. My current projects include developing cortically embedded recurrent neural networks (CERNNs) to investigate how anatomical constraints shape neural representations; modelling medial prefrontal representations of future states during planning; and developing models of internally generated cognition and mind-wandering.

More broadly, I am interested in how memory, planning, and internally generated representations interact across hippocampal, default-mode, and control systems to support flexible behaviour.

Current projects

Cortically Embedded RNNs

How anatomical constraints shape emergent neural representations and spatio-temporal dynamics.

Planning & future-state representations

Modelling medial prefrontal representations of future states during planning.

Internally generated cognition

Developing models of mind-wandering and interactions among hippocampal, default-mode and control systems.

Selected work

Papers, preprints & talks

Add up to five representative outputs with links.