Red Teaming Large Language Models
HomeBackgroundLogisticsInstructions
Light Mode

Red Teaming Large Language Models for Healthcare

Workshop at Machine Learning for Healthcare (MLHC), 2024

August 15, 2024, 1:00PM — 5:00PM

Room 1190, Bahen Centre for Information Technology, University of Toronto, Toronto, Ontario


Register Here



Location

We will begin by convening in Room 1190 of the Bahen Centre at the University of Toronto. The Bahen Centre is located at 40 St. George Street. An interactive map of the campus highlighting the Bahen Centre is available here.

The red teaming portion of the workshop will comprise breakout sessions hosted in Rooms 2185 and 2195 of the Bahen Centre. After introductions and splitting into red teaming small groups, directions to these rooms will be provided.

WiFi Access

Workshop participants can access WiFi at the University of Toronto using the eduroam WiFi network. The following credentials can be used:

  • Username: nbgv@eva.eduroam.ca
  • Password: nuakm

Workshop Schedule

1:00PM — 2:00PMSponsored Talks by the American Medical Association
2:00PM — 2:30PMActivity Overview; Getting Acquainted with Language Models
2:30PM — 4:00PMInteractive Red Teaming Exercise in Breakout Groups
4:00PM — 4:30PMSharing Harmful Prompts; Discussion of Safeguards
4:30PM — 5:00PMConclusion

What to Bring

  • All Participants: Laptop
  • Clinicians: come prepared to share with your group your daily workflow and to brainstorm ways in which LLMs may integrate with these processes. This will help ground the exercise in real-world use cases.

Red Teaming Overview

The below slides present an overview of the activity and some exemplar classes of vulnerability. We'll go over them during the workshop, but they make for some handy pre-workshop reading.


Made with ❤️ at UofT (w/ a little help from next.js).