Sonia Cromp

Profile picture


CV

Contact

sonic@cs.wisc.edu
GitHub
Twitter
Instagram

About me

I’m fifth-year CS PhD candidate at the University of Wisconsin-Madison, in Fred Sala’s Sprocket Lab. My interests are machine learning for climate science and other scientific fields, with a particular focus on generative modeling for rare events and low-data tasks.

I also collaborate with and am mentored by Allegra LeGrande at NASA GISS, where I previously completed a year-long Research Assistant appointment through the Climate Change Research Initiative. Our ongoing work focuses on designing a diffusion model for simulating the evolution of cyclone storms. In Summer 2024 and Spring 2025, I also completed two NASA OSTEM internships at GISS, where I worked with Michael Hendrickson to enable analysis of climate datasets with LLMs.

December 2025 News: Tabby, my LLM-based approach to tabular and structured data synthesis, was just accepted to TMLR! 🎉

Publications

Tabby: A Language Model Architecture for Tabular and Structured Data Synthesis
Sonia Cromp, Satya Sai Srinath Namburi GNVV, Catherine Cao, Mohammed Alkhudhayri, Samuel Guo, Nicholas Roberts, Frederic Sala
To Appear in Transactions in Machine Learning Research (TMLR)
ArXiv, blog and codebase

Diffusion-Based Modeling of Extratropical Cyclone Wind Field and Storm Surge Hazards
Sonia Cromp, Kellie O’Grady, Grace Kisslinger, Allegra LeGrande
2025 Meeting of the American Meterological Society (AMS)

Pretrained Hybrids with MAD Skills
Nicholas Roberts, Samuel Guo, Zhiqi Gao, Satya Sai Srinath Namburi GNVV, Sonia Cromp, Chengjun Wu, Chengyu Duan, Frederic Sala
COLM 2025 (link)

Climate PAL: Enhancing Accessibility of Climate Model Data through Conversational AI
Sonia Cromp, Behrad Rabiei, Maxwell Elling, Alexander Herron
2024 Meeting of the American Geophysical Union (AGU24)
NeurIPS 2024 CCAI Workshop on Tackling Climate Change with Machine Learning
NeurIPS 2024 Workshop on Machine Learning and the Physical Sciences (link)

OTTER: Effortless Label Distribution Adaptation of Zero-shot Models
Changho Shin, Jitian Zhao, Sonia Cromp, Harit Vishwakarma, Frederic Sala
NeurIPS 2024 (link)

The Credential is Not Enough: Combining Honeypots and Fake Credentials for Cyber-Defense
Sonia Cromp, Mark Bilinski, Ryan Gabrys, Frederic Sala
GameSec 2023 (link)

Mitigating Source Bias for Fairer Weak Supervision
Changho Shin, Sonia Cromp, Dyah Adila, Frederic Sala
NeurIPS 2023 (link)

Geometry-Aware Adaptation for Pretrained Models
Nicholas Roberts, Xintong Li, Dyah Adila, Sonia Cromp, Tzu-Heng Huang, Jitian Zhao, Frederic Sala
NeurIPS 2023 (link)

Causal Omnivore: Fusing Noisy Estimates of Spurious Correlations
Dyah Adila, Sonia Cromp, Sicheng Mo, Frederic Sala
ICML 2022 Workshop on Spurious Correlations, Invariance, and Stability

Essay Revision and Corresponding Grade Change as Captured by Text Similarity and Revision Purposes
Sonia Cromp, Diane Litman
EDM 2021 Workshop for Undergraduates (link)
Best Paper Award

Relaxed Skiplist for Concurrent Priority Queues
Sonia Cromp, Michael Spear
PACT 2020
Finalist, ACM Student Research Competition

Education

Service

Fun facts