Marcus Ghosh.

Marcus Ghosh

Research fellow
Imperial College London

Marcus Ghosh is a research fellow at Imperial College London. In his work, he uses artificial neural networks to model neural circuits. His research focuses on how the brain combines information across senses and how neural circuit structure shapes function. 

Ghosh completed his B.Sc., Ph.D. and medical degree (M.B.B.S.) at University College London before pursuing three postdoctoral fellowships—first at Sorbonne Université and then at Imperial College London. In his Ph.D., he designed a machine-learning tool kit to describe animal behavior at multiple timescales. Across his fellowships, he has modeled neural circuits using a range of neural network architectures and learning algorithms. Ghosh has received funding from Marie Skłodowska-Curie Actions, Schmidt Sciences and Imperial College London.

From this contributor

Explore more from The Transmitter

Mice playing, engaged in learning and memory: some of their brains are yellow, and some are red, matching a red and yellow legend that runs through the image to underscore both memory and learning.

This paper changed my life: A technical breakthrough for memory studies

In a 2007 Science paper, Mark Mayford and his colleagues found that some neurons activated during learning are also recruited during memory retrieval. Noelia Weisstaub shares how this study and others drove an era of great progress in tool development.

By Noelia Weisstaub
14 August 2026 | 5 min read

Early experience reshapes zebrafish retinal cells, alters behavior

The cells undergo a classic form of plasticity previously thought to occur only in downstream visual circuits.

By Siddhant Pusdekar
13 August 2026 | 0 min watch
Research image of expression of MEIS2 in PV interneurons.

Loss of interneuron plasticity may be shared hallmark of neurodevelopmental conditions

Restoring an experience-dependent interneuron plasticity gene in adulthood improves memory and cuts seizures in CNTNAP2 knockout mice, a new study shows.

By Claudia López Lloreda
12 August 2026 | 5 min read