Four NYU Researchers Win Sloan Fellowships for Brain Injury, Chemistry, Computing, and Sof

The Alfred P. Sloan Foundation's annual research fellowships have a simple premise: identify early-career scientists doing genuinely promising work and give them money and recognition to continue it. This year's 126 fellows, selected from 44 institutions in the United States and Canada, include four faculty members from New York University whose research spans neural science, organic chemistry, numerical computation, and software verification.

Each fellow receives $75,000 distributed over two years, without restrictions on how it is used. Since the fellowship program launched in 1955, 59 Sloan fellows have gone on to receive Nobel Prizes, and every winner of the John Bates Clark Medal in economics since 2009 has been a Sloan fellow. NYU has now had 102 faculty named to the program.

Danique Jeurissen: How Damaged Brains Reroute Information

Jeurissen, an assistant professor of neural science, studies a question that matters deeply for stroke and traumatic brain injury recovery: when a cortical area is damaged or disrupted, how does the brain find alternative pathways to keep processing information?

Her laboratory combines demanding behavioral tasks with causal manipulation techniques and electrophysiology. By temporarily disrupting specific neural circuits in controlled experiments, her team can map which pathways become more active in compensation and under what conditions that compensation succeeds or fails. The long-term goal is to understand the routing logic of neural information well enough to inform treatments that work with the brain's natural compensatory processes rather than around them.

Marvin Parasram: Building Molecules with Uncommon Atoms

Parasram, an assistant professor of chemistry, focuses on incorporating heteroatoms - atoms other than carbon and hydrogen - into organic molecules, including pharmaceutically relevant compounds. Heteroatoms appear in a wide range of life-saving drugs, but current methods for introducing them can be costly, low-yielding, and difficult to scale.

His approach uses light-activated 1,3-dipoles - reactive intermediates with a positive and negative charge spread across three atoms - to introduce heteroatoms while simultaneously activating the hydrocarbon framework. The same reagent acts as both the activator and the heteroatom source, streamlining what would otherwise require multiple separate chemical steps and providing a sustainable framework for synthesizing next-generation medicines.

Florian Schaefer: Statistical Inference Meets Numerical Computation

Schaefer, an assistant professor in NYU's Courant Institute School of Mathematics, Computing, and Data Science, works at the intersection of numerical computation and statistical inference. His group has developed algorithms with applications in computer graphics, materials modeling, and aircraft design optimization. His current focus is on information geometric mechanics - an interdisciplinary framework that revisits the statistical foundations of computational mechanics, with the potential to improve how physical simulations integrate with statistical modeling and optimization.

Joseph Tassarotti: Proving That Software Does What It Claims

Tassarotti, also at the Courant Institute, develops formal methods for verifying that software behaves correctly - not through testing, which can only check whether a program fails on cases the tester thought to try, but through mathematical proofs that the program cannot fail on any input.

His focus is on programs that are difficult to verify because their behavior is non-deterministic. The sources of non-determinism he studies include randomized algorithms, widely used in cryptography and privacy applications, and distributed systems, where multiple computers interact in ways that depend on message timing and network conditions. Tassarotti's group develops program logics - formal reasoning tools - that make such analysis tractable. As software increasingly controls safety-critical systems - medical devices, financial infrastructure, autonomous vehicles - the ability to formally verify correctness becomes practically important rather than merely theoretically interesting.