This repository contains our final project for the Winter 2025 w233 Privacy Engineering course in the UC Berkeley MIDS program. All of our work is contained within the final_project.ipynb notebook. Our experimental setup and findings are recorded in the final_report.pdf document. This project used the "89k ChatGPT conversations" dataset from Kaggle:
Noah Persaud. (2023). 89k ChatGPT conversations [Data set]. Kaggle. https://doi.org/10.34740/KAGGLE/DSV/5601259