This competition is co-organized by the Center for Artificial Intelligence Innovation at the National Center for Supercomputing Applications. The main goal of the hackathon is to let talented UIUC students showcase their skills in a friendly competition while working on challenging problems involving computational science and machine learning using state-of-the-art computational systems at NCSA.
The competition will take place on October 25-26, 2025 at the National Center for Supercomputing Applications. We encourage you to sign up soon using this Registration Link. If you already have a team be sure to have the team lead complete the form and include the name and contact information for the full team. Deadline for registration has been extended to October 15, 2025.
Eligibility: Teams must have two or more students (undergraduate and/or graduate) with at least one currently enrolled in the Computer Science Department. Students are encouraged to form teams of up to five students.
Criteria: Teams will be evaluated on the following:
- Innovative approach with respect to utilizing LLMs to solve computationally challenging problems, generate and execute a workflow, as well as the use of machine learning, information visualization, and other computational techniques.
- Effective use of NCSA computing resources, including performance on its flagship AI platform.
- Quality of a written project summary and oral presentation.
- Relevance of the developed solution.
Prize: 1st place $3000, 2nd place $1500, 3rd place $750
HACKATHON PROJECT
Virtual CRISPR: Can LLMs Predict CRISPR Screen Results?
Science team contact: Volodymyr Kindratenko (kindrtnk@illinois.edu), Dou Kwark (dkwark2@illinois.edu)
Technical team contact: Rohan Marwaha (rohan13@illinois.edu), Abhijith Nagarajan ( an80@illinois.edu)
Problem: CRISPR-Cas systems enable systematic investigation of gene function, but experimental CRISPR screens are resource-intensive. Here, we would like to investigate the potential of Large Language Models (LLMs) to predict the outcomes of CRISPR screens in silico, thereby prioritizing experiments and accelerating biological discovery. Our previous work (https://aclanthology.org/2025.bionlp-1.30/) produced a benchmark dataset (https://github.com/czbiohub-chi/immune-llm-acl) derived from BioGRID-ORCS and manually curated sources. We also evaluated the performance of several LLMs across various prompting strategies, including chain-of-thought and few-shot learning. Furthermore, we proposed a novel, efficient prediction framework using LLM-derived embeddings, achieving significantly improved performance and scalability compared to direct prompting. The goal of this hackathon competition is to push this work further to improve the performance of the prior work.
Monday – September 8, 2025
Application form is open to register competition teams (on-line)
Wednesday – October 15, 2025
Deadline for registration
Monday – October 20, 2025
8:00 am — Teams are announced (on-line)
4:00 pm — Overview of the Hackathon rules, challenge problem, and a brief intro to the computing environment and LLM access (1104 NCSA)
Saturday – October 25, 2025
8:30 am — Teams work on the problem (light breakfast will be provided) (1104 NCSA)
Noon — Lunch (pizza will be provided) (1104 NCSA)
1:00 pm — Teams continue to work on the challenge problems (snacks will be provided) (1104 NCSA)
4:00 pm — Teams briefing (1104 NCSA)
5:00 pm – end
Sunday – October 26, 2025
8:30 am — Teams work on the problem (light breakfast will be provided) (1104 NCSA)
Noon — Lunch (pizza will be provided) (1104 NCSA)
1:00 pm — Teams continue to work on the challenge problems (snacks will be provided) (1104 NCSA)
4:00 pm — Teams briefing (1104 NCSA)
5:00 pm – end
Wednesday – October 29, 2025
2:00 pm – 6:00 pm — Teams present results (1104 NCSA)
Wednesday – November 5, 2025 – (1030 NCSA)
1:30 pm – 2:30 pm – Presentation of Certificates to winning teams by Bill Gropp, Director of the National Center for Supercomputing Applications
Winning teams announced online