Johns Hopkins researchers to present at NeurIPS 2025

Students and faculty from the Whiting School of Engineering’s Department of Applied Mathematics and Statistics will present their research at the 39th annual NeurIPS conference, held December 2 – 7 at the San Diego Convention Center.

NeurIPS is widely recognized as the flagship event in the machine learning and AI communities, bringing together cutting-edge work in fields such as artificial intelligence, statistics, optimization, and neuroscience.

This year, NeurIPS received 21,575 valid submissions in its main program track and accepted just 5,290 papers—an acceptance rate of 24.52%.

The department’s researchers will deliver both an in-person poster presentation and a recorded talk for the conference’s virtual program.

Their work spans a wide range of topics, including the following papers:

• Beyond Scores: Proximal Diffusion Models

Zhenghan Fang, Mateo Diaz, Sam Buchanan, Jeremias Sulam

• Extragradient Method for $(L_0, L_1)$-Lipschitz Root-finding Problems

Sayantan Choudhury, Nicolas Loizou

• Generalization Guarantees for Learning Branch-and-Cut Policies in Integer Programming

Hongyu Cheng, Amitabh Basu

• Multiplayer Federated Learning: Reaching Equilibrium with Less Communication

TaeHo Yoon, Sayantan Choudhury, Nicolas Loizou

• Nonlinear Laplacians: Tunable principal component analysis under directional prior information

Yuxin Ma, Dmitriy Kunisky

• On Transferring Transferability: Towards a Theory for Size Generalization

Eitan Levin, Yuxin Ma, Mateo Diaz, Soledad Villar