Machine Learning and Computational Statistics

The Center for Computational Mathematics is home to a large and vibrant group of researchers in the areas of machine learning and computational statistics. Particular areas of interest include:

• Deep learning and representation learning
• High dimensional data analysis
• Learning theory
• Machine learning for science
• Optimization
• Probabilistic modeling

Our group is part of a larger research effort in machine learning across all of Flatiron Institute, and many projects are pursued in collaboration with members of other centers.

Research scientists are responsible for the long-term vision of ML research. They also serve as mentors for postdocs and summer interns.

Visiting scholars have short-term or part-time appointments, while guest researchers visit frequently on a more informal basis. Both work closely with our postdocs and research scientists while providing a liaison to their home institutions.

Postdocs are appointed as Flatiron Research Fellows with three-year terms. They are free to pursue their own intellectual agenda, and they are encouraged to collaborate with researchers across all of Flatiron. As part of their support, they are offered a generous travel budget and access to subsidized housing in Greenwich Village.

September 2026: Welcome to newly arriving postdocs Alexandre Adam (joint with CCA), Simone Bombari(joint with NYU), Michael Crawshaw, and Siyuan Guo (joint with CCN).

July 2026: Researchers in CCM are co-authors on eight papers at ICML-2026 in Seoul, including work on non-Euclidean gradient descent, step-size stability, protein design, foundation models for continuum dynamics, neural operator splitting, energy-based generative models, and single-index learning.

July 2026: Welcome to newly arriving postdoc Stephen Zhang, who will work jointly between CCB and CCM, and visiting scholar Gemma Moran on sabbatical from Rutgers.

June 2026: Welcome to our newest guest researchers, Professors Yuejie Chi and Nicholas Loizou.

May 2026: A paper co-authored by David Persson and Robert Gower of CCM has received an Honorable Mention award at ICLR-2026. The authors were recognized for their work on The Polar Express: Optimal Matrix Sign Methods and their Application to the Muon Algorithm.

April 2026: Congratulations to guest researcher Andrew Gelman for his election to the National Academy of Sciences!

March 2026: Researchers in CCM have co-authored a slew of papers to appear at ICLR-2026 in Rio de Janeiro. The papers cover a variety of topics in representation learning, including any-subgroup equivariant networks, efficient finetuning of large models, fast hyperparameter transfer,generalized compressed sensing with diffusion models, generative modeling with stochastic interpolants, inference-time scaling of diffusion models, optimal matrix sign methods for the Muon algorithm, recall in transformers, Kolmogorov-Arnold neural operators, and the effects of personalization on prediction and explainability.

February 2026: Welcome to Leon Bottou, our newest ML visiting scholar, who will be working jointly between CCM and CCN.

January 2026: Welcome to Louis Grenioux, our newest Flatiron Research Fellow in CCM, and ProfessorHaewon Jeong, our newest ML visiting scholar, who will be working jointly between CCM and CCA.

November 2025: Welcome to Professor Bert Kappen, our newest ML visiting scholar, who will be working jointly between CCQ and CCM.

September 2025: Researchers in CCM have co-authored fifteen papers to appear at this year’s NeurIPS conference in San Diego, including an oral on Adam’s secret sauce and a spotlight on variational inference with Feynman parameterizations.

August 2025: Welcome to newly arriving postdocs Mark Goldstein , Zahra Kadkhodaie , Tetiana Parshakova , and visiting scholar Luhuan Wu.

May 2025: Researchers in CCM have co-authored four papers to appear at this year’s ICML in Vancouver, including one oral on in-context denoising with one-layer transformers.

May 2025: Charles Margossian and Lawrence Saul have received the best paper award at this year’s AI-STATS conference in Phuket for their work on variational inference in location-scale families.

March 2025: Jiequn Han has received a biannual SIAM Early Career Prize for his outstanding contributions to the fields of deep learning, stochastic control, stochastic differential equations, and molecular dynamics.

March 2025: Daniel Lee and Lawrence Saul have won the Flywire VNC Matching Challenge.

January 2025: Researchers in CCM have co-authored five papers to appear at this year’s ICLR in Singapore, including two spotlights on understanding factual recall in transformers and parallel end-to-end training in reversible architectures.