Ulyana Piterbarg

Ulyana Piterbarg

up2021 [at] cims.nyu.edu

[Bluesky] [Google Scholar] [Github] [CV]

About

I’m a Ph.D. candidate in the CILVR lab at NYU Courant, co-advised by Rob Fergus and Lerrel Pinto. My research is supported by a DeepMind Ph.D. Scholarship and an NSF Graduate Research Fellowship.

I’m interested in generative models that can solve hard tasks in settings like code synthesis, decision-making, and open-ended/agentic interaction. Recently, I’ve been thinking about:

Currently, I’m a research scientist intern at Meta working on Llama 4 post-training for collaborative multi-agent tasks, supervised by Gregoire Mialon and Thomas Scialom. I will be on the job market in Fall 2025/Spring 2026!

Previously, I’ve spent time working on improving small language model reasoners with the GenAI/AI Frontiers Teams at Microsoft Research and studying ML-powered weather/climate simulators with the Applied Science Team at Google Research.

I did my undergrad in mathematics and computer science at MIT, during which I was exceptionally lucky to be mentored by Kelsey R. Allen and Josh Tenenbaum.


News

July 2025
Our work on scalable generation of synthetic mid-training data for human-like software development was accepted to COLM 2025 in Montreal ⚜️
June 2025
I gave an invited talk on agents at the 7th Deep Learning Ulaanbaatar Summer School in Mongolia. Thanks to Sainaa for coordinating!
April 2025
LintSeq won an outstanding paper award at the New England NLP Workshop! 🎉
April 2025
I gave an invited talk on LM priming for long-horizon agent tasks and participated in a panel on the risks & opportunities of open-ended self-improvement at the Workshop on Self-Improving Foundation Models at ICLR 2025 in Singapore