Kempner Institute Funds 21 Harvard Undergraduates in Its Largest KURE Cohort Yet

What KURE offers

KURE gives Harvard undergraduates funding during the fall and spring semesters to carry out research projects supervised by Kempner-affiliated faculty and mentors. According to the institute, the fellowships also provide mentorship for intensive projects related to intelligence.

The fall 2026 group is the sixth cohort of the term-time program and the biggest so far. The Kempner says the program has run since 2024.

The institute also runs a 10-week residential summer program, the Kempner Research in Artificial & Natural Intelligence for Undergraduates with Mentorship (KRANIUM), in partnership with the Harvard Summer Undergraduate Research Village (HSURV). For either KURE or KRANIUM, students must secure a research position with a Kempner-affiliated faculty member before they apply.

Research topics in the new cohort

The Kempner says this year's students are exploring the foundations of intelligence, with a focus on how systems learn and represent information. Areas range from speech generation and spatial reasoning to multilingual capabilities and collaboration between agents. The projects will also look at ways to improve model efficiency, interpretability and generalization.

Several projects apply AI to practical problems. The institute names clinical trial optimization, public housing resource allocation and scientific knowledge management as examples.

The published list shows the range of work. Examples include:

• Anthony Shen ’28, supervised by Greta Tuckute: causal steering dynamics of lexical commitment in a self-supervised speech model.
• Betina Kreiman ’28, supervised by Susan Murphy: partial pooling for sequential decisions in small clinical trials.
• Ikenna Ogbogu ’27, supervised by Kianté Brantley: optimizing housing voucher allocation through reinforcement learning.
• Dries Rooryck ’27, supervised by David Alvarez-Melis, with mentors Sara Kangaslahti and Nihal Nayak: code-switching scaling laws.
• Rubin Bekteshi ’29, supervised by Nada Amin: measuring specification strength in machine-generated formally verified programs.
• Yurui Zi ’28, supervised by Sam Gershman, with mentor Arthur Prat-Carrabin: computationally modelling belief updating using meta-Bayesian inference.

Students come from a mix of concentrations, including computer science, statistics, mathematics, physics, neuroscience, philosophy, comparative literature and a special concentration in rationality and decision theory.

The institute's approach and what it means for students

The Kempner describes its mission as understanding the basis of intelligence in natural and artificial systems by recruiting and training future generations of researchers. It studies intelligence from biological, cognitive, engineering and computational perspectives. Its stated premise is that natural and artificial intelligence are intimately interconnected, and that the next generation of AI will require the same principles the brain uses for fast, flexible natural reasoning.

For Harvard undergraduates, the growth of KURE points to more places in a research program that pairs students with faculty working across neuroscience and machine learning. The source does not give the number of applicants or the size of the awards. It does make clear that interested students need a commitment from a Kempner-affiliated faculty member before applying, and that more information on KURE and KRANIUM is available on the Kempner Institute website.