KRANIUM Participant Page
Welcome! This page will have the resources and information for the Kempner Research in Artificial & Natural Intelligence for Undergraduates with Mentorship (KRANIUM) students.

Kempner Research in Artificial & Natural Intelligence for Undergraduates with Mentorship (KRANIUM)
KRANIUM is a 10-week summer program sponsored by the Kempner Institute as part of the Harvard Summer Undergraduate Research Village (HSURV). KRANIUM aims to provide a formative and substantive research experience for — and build community among — a small cohort of Harvard College undergraduates interested in the study of natural and artificial intelligence.
Learn more about KRANIUMQuick Links
Logistics
- Denise Yoon, Associate Director for Educational Programs
- Lani O’Donnell, Educational Coordinator
The Kempner Institute is located on the 6th floor of the Harvard University Science and Engineering Complex (SEC).
Address:
150 Western Avenue,
Allston, MA 02134
All KRANIUM events will take place in the Kempner Conference Room (6.242) in the East wing unless otherwise noted.
Shuttles:
Harvard provides free shuttles to get around campus. You can see where on the map of Harvard Shuttle Stops.
Note: Kempner Institute’s stop is the bottom, middle stop called SEC. It is on the green Quad Stadium Express line, the purple Quad SEC Direct line, and the red Allston Loop line.
You will receive the first $1000 as soon as possible after being set up in Harvard’s procurement system, Buy2Pay.
You will receive the remaining $2000 in early July.
Register for Buy2Pay
Buy2Pay is the procurement system we use to pay your stipends.
To set up your account with Buy2Pay, contact Lani O’Donnell, and she will send you a supplier registration invitation via email that will come from ap_supplieronboarding@harvard.edu. See the detailed description of the registration process here. Harvard will review your registration details and notify you if anything is missing. You can expect this process to take about a week to complete from start to finish.
When completing your registration, please select the electronic payment option (Zelle). This is the fastest way to receive payment and safeguards us from having to track down and replace paper checks if they are lost in the mail.
After your profile is setup and confirmed, you can use this link to login. *Note: Foreign Individuals (Non-U.S. Citizens) may also be required to submit additional information to ensure compliance with U.S. regulations. New registrants may receive a separate email from Harvard’s Online Tax Compliance System, Sprintax Calculus (hutax@sprintax.com), after their Buy2Pay registration has been submitted. The link they send will only remain active for 24 hours so please respond quickly. You may set up a 1:1 appointment to get help with completing the Sprintax questionnaire and uploading required documents.
Stipend Financial Policy
To learn more about the tax and financial policies of your stipend, please see the following resources:
• There is a training and slide deck held on 4/17/24 that has some nice quick reference charts that may be helpful. See the main policy page under What’s New for the slide deck and recording.
• They also have a full written policy – Fellowship Payments and Reimbursements for Students and Non-employee Postdocs/Fellows | Financial Policy Office (harvard.edu)
• General Fellowship Information Brochure has some frequently asked questions.
• Fellowships Toolkit – Appendix B gives a more detailed breakdown on tax and reporting requirements that is a good reference.
Please refer to the HSURV handbook for information about housing, dining, and contacts for questions not addressed in the handbook.
You should receive an email invitation to the Kempner workspace. If you don’t, try logging into the Harvard Enterprise grid in order to see the Kempner workspace.
Please monitor the private kranium2025 channel and use it to chat/ask questions!
If you’re using the cluster, please make sure you join and monitor the cluster-users channel
Denise Yoon: Thursdays 12pm, Room 6.231 (message to schedule Zoom if you can’t come in person)
Lani O’Donnell: Wednesdays 2pm, Room 6.149 (message to schedule Zoom if you can’t come in person)
July 4th is a holiday. We may approve up to 3 other days off if your mentor approves, it isn’t a Wednesday, and it is not too many consecutive days off in a row. Please email Lani about vacation requests.
Events
Wednesday, June 11th | Orientation Mentor Lunch | |
Wednesday, June 18th | Compute Cluster Onboarding | |
Wednesday, June 25th | Presenting Your Research Pt. 1 | |
Wednesday, July 2nd | Industry Talk | |
Wednesday, July 9th | Faculty Talks | |
Wednesday, July 16th | Graduate School Panel | |
Wednesday, July 23rd | Presenting Your Research Pt. 2 | |
Wednesday, July 30th | TBD | |
Wednesday, August 6th | Poster Session |
HSURV will be communicating all announcements and events on Canvas. Please reach out and let us know if you do not have access.
Towards the end of the program, we will have a poster session where KRANIUM students will present posters of their research. Information on the poster session will be updated throughout the summer.
At the end of summer, you’ll fill out a short final report that consists of your abstract and a brief personal reflection.
Meet the 2025 KRANIUM Participants
Victoria Chen
Class Year: 2028
Primary Mentor: Ashley Thomas
Project Title: What’s in a Look? Infant Eye-Gaze and the Role of Collaborative Understanding in Human Intelligence

Hannah Kim
Class Year: 2026
Primary Mentor: Binxu Wang
Project Title: Diffusion Models and their Relational Compositions

Simon Ma
Class Year: 2026
Primary Mentor: Isabel Papadimitriou
Project Title: Investigating a Phonetic Subspace in Self-Supervised Speech Models

Ian Moore
Class Year: 2026
Primary Mentor: Susan Murphy
Project Title: Non-Interpolated Action Generalization In Meta Reinforcement Learning

Laasya Nagumalli
Class Year: 2028
Primary Mentor: Naomi Saphra
Project Title: Nonlinear Feature Interference

John Rho
Class Year: 2025
Primary Mentor: Kiante Brantley
Project Title: Guided Reinforcement Learning for Robotics Control Tasks

Carl Scandelius
Class Year: 2027
Primary Mentor: Haim Sompolinsky
Project Title: Separability Properties of General Perceptual Manifolds

Vincent Song
Class Year: 2028
Primary Mentor: Sam Gershman
Project Title: Inducing Meta-Generalisation in Machines

Alexandru-Raul Todoran
Class Year: 2028
Primary Mentor: Eran Malach
Project Title: Internalizing Branching Reasoning in Language Models: Toward Implicit and Parallel Exploration

Eric Xu
Class Year: 2028
Primary Mentor: Venkatesh Murthy
Project Title: Developing Models for Olfactory Perception

Gavin Ye
Class Year: 2028
Primary Mentor: Nada Amin
Project Title: Agentic AI for Automated Drug Design

Todd Zhou
Class Year: 2027
Primary Mentor: Mengyu Wang
Project Title: Evaluating and Enhancing Large Language Models (LLMs) for Fair and Accurate Medical Image Diagnosis

Past KRANIUM Participants
Ege Cakar
Primary Mentor: Cengiz Pehlevan
Project Title: Introducing Logical Thinking into Neural Networks
Emma Finn
Primary Mentor: Manos Theodosis
Project Title: Learning Artistic Signatures: Symmetry Discovery for Style Transfer
Ely Hahami
Primary Mentor: Kazuki Irie
Project Title: Large Language Models with Long-Term Memory
Kayla Huang
Primary Mentor: David Brandfonbrener
Project Title: Enhancing natural language capabilities of AI through hybrid MoE and GSSM models
Lavik Jain
Primary Mentor: Yasha Ektefaie
Project Title: Predicting Antibiotic Resistance in Mycobacterium tuberculosis with Interpretable Machine Learning
Justin Ji
Primary Mentor: Kenneth Li
Project Title: AttackVis: A visualization playground for large language models in response to adversarial attacks
Xiang Yi Sean Meng
Primary Mentor: Dr. Kevin Mastro
Project Title: Informing ML Explore/Exploit Balance through Insights from Age-Dependent PFC Circuit Modulation
Aneesh Muppidi
Primary Mentor: Wilka Carvalho
Project Title: Unsupervised Agent Discovery using Object-Centric Inverse Reinforcement Learning
Anne Mykland
Primary Mentor: Naomi Saphra
Project Title: Feature Compression and Direction Similarity in Neural Language Modeling
Alliyah Steele
Primary Mentor: Morgan Talbot
Project Title: Integrative Analysis and Predictive Modeling of Linguistic Brain Function Using LLMs
Johnathan Sun
Primary Mentor: Sammy Jelassi
Project Title: Search Algorithms and Value Functions in Programming Puzzles