Sarthak Chandra

But he’ll do it using cold hard math. He is a theoretical physicist in the lab of Ila Fiete, director of the K. Lisa Yang Integrative Computational Neuroscience (ICoN) Center at MIT, and entered the field of neuroscience because he “was yearning to use mathematical tools in the real world.”
“A lot of my career so far has been going from abstract things to more-concrete things,” Chandra says. With two recent scientific papers published in Nature since the beginning of t he year, he’s doing just that. Chandra is studying grid cells, which he describes as “the CPS of the brain” that helps people and animals figure out where they are in space.

In the first paper, published in Nature in February 2025, he and his research team developed a new mathematical model that helps explain how grid cells form in the brain. These cells occur in distinct categories or modules, a pattern known as modularity. By integrating different biological features into their model. it churned out mathematical evidence for how modularity arises in grid cells.

Chandra asked another interesting question: Where else in nature can we see modularity? As it turns out. Chandra and his coauthors found that their model could also explain discrete ecological niches, the slot an organism fills in its ecosystem: apex predator, grass-eating herd, etc. “To apply a rule that matches a biological system so perfectly, it was an incredible experience,” Chandra says.

Chandra’s other paper, published in Nature in January 2025, focuses more on grid-cell function. “Our first paper focused on the ‘how· of grid cells. This paper focuses more on the ‘why’ of them,” he says. The team found that the grid-cell modules were strongly linked to the brain’s hippocampus, the center of episodic memory and spatial navigation.

Storing memories in a “virtual place” that one can visit in their mind and recall that information much more easily and reliably is a concept also known as a “mind palace,” Chandra explains. Most memory models hit a “memory cliff’- beyond a certain number of “memories,” the model collapses and forgets everything. But. when Chandra and his team developed a model linked to a scaffold of spatial representations, “we get much more realistic slow decay – this model that more closely mirrors reality,” he says.

“Towards the end of my PhD, I wanted to switch away from [the] purely theoretical. and neuroscience seemed like an interesting area where my theoretical skills could be useful.” Chandra adds. “Ila [Fiete] took a chance on me: she also has a background in physics. Since then, I’ve been looking at physics approaches in neuroscience