Summer '26 REU Opportunity w/Prof R. Zhang, Apply by 5/29

Summer 2026 REU Opportunity
Discrete Diffusion Models & Diffusion LLMs

PI: Prof. Ruqi Zhang, Department of Computer Science, Purdue University
Funding: NSF CISE REU — $9,000 stipend + $1,000 travel
Duration: 10 weeks, Summer 2026 (in-person at Purdue or remote)

Project
Diffusion models are transforming generative modeling, and recent work is pushing them into
discrete domains, such as text and code, as a competitor to autoregressive LLMs. This project
contributes to our group’s research on discrete diffusion and diffusion LLMs, including
training and inference efficiency, corruption schedules, and hybrid discrete-continuous diffusion.
The student will work with Prof. Zhang and a senior PhD mentor, aiming for a conference paper.

Who should apply
• Strong Python/PyTorch skills
• Solid foundations in probability, linear algebra, and machine learning
• Curiosity about generative models and LLMs; prior diffusion or LLM experience is a
bonus, not required
• Currently enrolled undergraduate; must be a U.S. citizen, U.S. national, or U.S.
permanent resident (NSF CISE REU eligibility)

How to apply
Email ruqiz@purdue.edu  with the following:
• CV / resume
• Unofficial transcript
• Short paragraph on your interests and fit
Application deadline: Friday, May 29, 2026