Michael Cardei

prof_pic.jpg

My name is Michael Cardei and I am Ph.D. student in Computer Science at the University of Virginia, where I am advised by Professor Ferdinando Fioretto. As a member of the RAISE lab, my research is focused on Responsible Generative AI and Generative AI for Scientific Applications. My research is supported by the National Science Foundation Graduate Research Fellowship (NSF GRFP).

I completed my BS in Computer Science at the University of Florida, where I performed research under the guidance of Dr. My Thai in the areas of LLM’s, Federated Learning, and XAI. I was part of the 2023 Robotics Institute Summer Scholars program RISS at the Robotics Institute at Carnegie Mellon University where I worked on Computer Vision applications for Autonomous Vehicles under the guidance of Dr. Srinivasa Narasimhan. In the summer of 2022, I was part of the Wake Forest University, Center for Artificial Intelligence summer REU program under the guidance of Dr. Umit Topaloglu researching ML privacy and bias. Most recently, I was a Staff Machine Learning Research Intern at Visa, focusing on generative AI for financial applications.

news

Nov 17, 2025 Happy to have recieved the NeurIPS Scholar Award! Very excited to attend NeurIPS in person this Decemeber
Oct 21, 2025 Paper titled “Contrained Tabular Diffusion for Finance” accepted to ACM International Conference on AI in Finance (ACM ICAIF) 2025
Sep 19, 2025 Paper titled Constrained Discrete Diffusion accepted to The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS)
Jun 25, 2025 Paper titled ROADWork Dataset: Learning to Recognize, Observe, Analyze and Drive Through Work Zones accepted to the International Conference on Computer Vision (ICCV).
Jun 10, 2025 Awarded the NSF GRFP Scholarship Award - 2025
May 16, 2025 Starting a new internship at Visa as a Staff Machine Learning Research Scientist.
May 10, 2025 Our paper Neuro-symbolic Generative Diffusion Models was awarded a DARPA Disruptive Idea Award (Top 5 Conference Submission).
Apr 20, 2025 Paper titled Neuro-symbolic Generative Diffusion Models for Physically Grounded, Robust, and Safe Generation accepted to NeuS 2025 and selected for Oral Presentation!
Mar 10, 2025 Excited to share a new preprint: Constrained Discrete Diffusion.
Jan 10, 2025 Paper titled Speculative Diffusion Decoding: Accelerating Language Generation through Diffusion accepted to NAACL 2025 and selected for Oral Presentation!
Aug 10, 2024 I have begun my Ph.D. in Computer Science at the University of Virginia, where I research Responsible Generative AI in the RAISE Lab under the guidance of Dr. Fioretto
Feb 28, 2024 Full version of Improving Equity in Deep Learning Medical Applications with the Gerchberg-Saxton Algorithm is published in the Journal of Healthcare Informatics Research; available (here).

selected publications

  1. cdd.jpg
    Constrained Discrete Diffusion
    Michael Cardei*, Jacob Christopher*, Thomas Hartvigsen, and 2 more authors
    The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS), 2025