HACKATHON 2023, Catholic university of america

Theme: the hackathon will combine Large Language Models and a Physics Problem. 

Instructions and a tutorial can be found at this link.

The hackathon will involve using an LLM (ChatGPT-3.5) to write a binary classifier for electromagnetic calorimeter events. This hackathon will involve no programming by the participants. Rather, you will gain experience in prompt engineering with an LLM to refine your physics result.

We used data from the GlueX Barrel Calorimeter (BCAL), and focus on neutron and photon showers classification.   

Powered by AWS

Organizers: Cris Fanelli (William & Mary),  James Giroux (William & Mary), Patrick Moran (William & Mary), Karthik Suresh (William & Mary)

Technical Assistance: Eric Walter (William & Mary, HPC)

List of teams (please form your team if you want AWS resources! Contact ai4eichackathon@gmail.com):

https://docs.google.com/spreadsheets/d/1UzTGyMYiT1I0SKEJrd1gHorz4kHITEIDk0oK53V6PeU/edit#gid=1691241232

Leaderboard: https://ai4eichackathon.pythonanywhere.com/leaderboard

Zenodo link (dataset and documentation): https://doi.org/10.5281/zenodo.7197023 

Powered by AWS

Organizers: Cris Fanelli (William & Mary/JLab), Diana McSpadden (JLab/Data Science), Kishan Rajput (JLab/Data Science)

Advisory and problem definition: Evaristo Cisbani (INFN), Wouter Deconinck (U. Manitoba)

Computing resources:  Eric Walter (William & Mary, IT)

Data generation,  Documentation,  Validation: James Giroux (U. Regina), Karthik Suresh (U. Regina)

Technical Assistance: Eric Walter (William & Mary, HPC),  James Giroux (U. Regina), Karthik Suresh (U. Regina)