UMD Wraps Up 13th Annual Summer REU Program

The University of Maryland’s Department of Computer Science will wrap up its 13th annual Research Experiences for Undergraduates (REU) program this month, offering students from across the country a summer-long opportunity to engage in academic research alongside faculty and graduate mentors.

The program, running from June 1 through August 15, features students working in small groups on research spanning a range of computer science fields, including algorithms, machine learning, AI and quantum computing.

Participants were paired with faculty members and doctoral students to gain hands-on experience in the research process, from reading academic papers and coding to designing experiments and presenting findings. 

William Gasarch, a professor in the Department of Computer Science and one of the lead organizers of the REU initiative at UMD, said he has seen a marked shift in the preparation and drive of the students who apply.

“I've been running this program since 2013, and the students keep getting better and better in terms of how much they know coming in and how hard they work,” Gasarch said.

This summer’s projects tackled a wide array of research topics. One project focused on algorithms to infer the evolutionary histories of cancerous tumors, combining theoretical computer science and biological data analysis.

Luiz Mata Lopez, a senior majoring in computer science and mathematics, contributed to this work and reflected on the academic and social aspects of the summer program.

"REU has been a great experience, which has exposed me to several areas of computer science theory and applied research,” Lopez said. “This summer program was different from my previous experiences in that I got to participate in group activities that made me think hard and utilize my problem-solving skills. Thanks to these activities, I have been able to meet many new friends, both from UMD and from different universities, which bring diverse perspectives to the table.”

Quantum computing was also a major area of focus. Cassandra Hopkin, a recent graduate of the University of California, Davis, who majored in computer science and mathematics, worked on a project in quantum error correction, which she described as foundational for building large-scale quantum computers.

“This field of research is vital for the development of large-scale quantum computers,” Hopkin said. “My project would be useful for the classification of fracton models, as well as for finding logical gate restrictions on models in a given generalized family.”

Hopkin also emphasized the mentorship aspect of the program.

“My two faculty mentors have been very supportive and are always willing to answer my questions on Slack,” she said. “I meet with them once a week to check in with my progress and to set goals for the following week, which has been very helpful.”

Another participant, Soren Brown, a senior studying computer science and mathematics at UMD, contributed to a computational geometry project involving diagrams used to find the closest locations in a given space, known as Voronoi diagrams, in Hilbert geometry.

“My project expanded on a few papers in the past that solved nearest Voronoi and farthest Voronoi; we solved everything in between,” Brown said. “Additionally, we are expanding our results to 3D and any higher arbitrary dimension.”

Teamwork and knowledge-sharing were central to Brown’s experience.

“Working with others in this project has taught me about people and their different work styles,” he said. “A useful thing my group has done is delegate different parts of the project to people, and then discuss our findings with others as we progress. We work on our problems, but share our ideas with others along the way.”

Other projects this summer included work on space-efficient parallel algorithms, computational geometry, and phylogenetic reconstruction. In machine learning and AI, students explored autonomous driving systems and crop planning with multi-agent reinforcement learning. Quantum computing projects focused on error correction, code symmetries, and simulation methods. Participants presented their findings during a final research conference attended by faculty and peers.

Gasarch emphasized that beyond the research experience, REU programs serve an essential function in helping students assess whether to pursue graduate school or a research-focused career.

He also offered advice to prospective applicants.

“If you want to do research over the summer, then apply to lots of programs,” Gasarch said. “Even if you are not sure if research is for you, this is a great way to find out.”

2025 REU projects:

• Laxman Dhulipala, an assistant professor of computer science, led a project on parallel algorithm design. Students examined how to create succinct and space-efficient parallel algorithms, with an emphasis on work and depth complexity. The project explored topics such as compact data structures, parallel graph algorithms, and parallel programming techniques in C.

• David Mount, a professor of computer science with an appointment in UMIACS, and Auguste Gezalyan, a computer science doctoral student, guided a computational geometry project. Students investigated the Hilbert and Thompson geometries and their behavior in spaces like the probability simplex. The project included both theoretical work and contributions to a software package for visualizing geometric structures.

• Erin Molloy, an assistant professor of computer science, along with doctoral students Rachel Parsons and Junyan Dai, led a computational biology project. Students explored the reconstruction of evolutionary trees (phylogenies) under the influence of large-scale genomic changes. The work involved simulations, real-data analysis, and evaluation of discrete and statistical algorithms for phylogeny inference.

• Ming Lin, the Dr. Barry Mersky and Capital One E-Nnovate Endowed Professor and a Distinguished University Professor of computer science, directed a machine learning project focused on autonomous driving. Students worked with human driving data to develop machine learning models that anticipate behavior in mixed autonomy traffic environments. The project emphasized the application of theory to real-world autonomous vehicle safety challenges.

• Aviva Prins, a computer science doctoral student, developed a multi-agent reinforcement learning project. Students helped build a decision support tool for crop planning aimed at improving outcomes for small and marginal farmers in India. The project was conducted in collaboration with Kheyti, a nonprofit organization, and explored hierarchical episodes in reinforcement learning.

• Victor Albert, a fellow in the Joint Center for Quantum Information and Computer Science (QuICS), and Dominic Williamson, a theoretical physicist, led a project on quantum error correction using bivariate bicycle codes. Students adapted surface code techniques—such as modifying lattices and including defects—to the bicycle code setting, with possible directions including decoding, logical gate operations, and threshold improvements.

• Yuxin Wang and Yifan Hong, researchers in quantum information science, led a project on low-density parity-check (LDPC) codes. Students investigated how to reduce the spatial overhead of quantum LDPC codes while preserving error-correcting performance. The project included theoretical work and practical code construction for fault-tolerant quantum computing.

• Shubham Jain and Victor Albert supervised a project on qutrit quantum codes. Students explored extensions of binary quadratic residue codes to qutrit systems, to enable universal quantum computation. The project involved identifying ternary QR codes and adapting them into the CSS code framework.

• Michael Gullans, Thomas Steckman and Dominik Hangleiter led a project on robust Bell sampling. Students worked on error mitigation strategies for noisy quantum simulations, including fidelity estimation and shadow tomography. The project included theoretical analysis, numerical simulations and open-source code contributions.

• Ian Spielman and Gretchen Campbell, researchers at the Joint Quantum Institute (JQI), guided an experimental physics project. Students assisted in developing an apparatus for quantum simulations with dipolar erbium atoms. The project involved laser cooling, trapping techniques, and hands-on collaboration with graduate students and postdoctoral researchers to study quantum degenerate gases.

—Story by Samuel Malede Zewdu, CS Communications