https://github.com/eeddaann/llm-privacy
gauge the proficiency of various LLMs in discerning between appropriate and inappropriate applications of CRM data.
https://github.com/eeddaann/llm-privacy
Last synced: over 1 year ago
JSON representation
gauge the proficiency of various LLMs in discerning between appropriate and inappropriate applications of CRM data.
- Host: GitHub
- URL: https://github.com/eeddaann/llm-privacy
- Owner: eeddaann
- Created: 2024-04-09T20:30:30.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-04-09T20:31:46.000Z (over 2 years ago)
- Last Synced: 2025-01-29T05:28:27.646Z (over 1 year ago)
- Language: Jupyter Notebook
- Size: 3.01 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Can we rely on AI to safeguard our privacy on CRM?
## Introduction
- This repository contains the code for the final project in Data Privacy Course.
- The goal of the project is to gauge the proficiency of various LLMs in discerning between appropriate and inappropriate applications of CRM data.
- Multiple LLMs were queried with different questions which related to the field of ethical CRM data utilization
- The results are visualized in the table below:
## Software Architecture
- The ```script``` refers to ```yes_no.py```
- During development we used [LM Studio](https://lmstudio.ai/) to manage the LLMs and deploy them locally. However since openai's API were used, any endpoint which support that API can be configured.
- The ```analysis.ipynb``` notebook contains the code for the analysis of the results.