In the previous activities, we explored biases in conversations with language models. The goal of this activity is to systematize that exploration by beginning to build a structured dataset that captures biases in a smaller unit of discourse: phrases (or sentences).
To complete this asynchronous activity, you will first need to watch the following videos and then respond to the two prompts provided.

Important: The deadline to complete this activity is Wednesday, August 21st at 11:59 PM.

Videos

Introduction Video: In this video, Luciana reminds us how this activity connects to the overall goals of the course.

EDIA Video: In this video, Luciana explains how to access EDIA and then explores four types of bias using the tool: gender, age, death, and religion. She creates a total of six entries in the database of previous queries.

Task 1

For this activity, each of you is expected to create at least 15 entries in the EDIA database. If you’d like to do more—great! 🙂
These 15 entries should explore at least three different types of bias or controversial topics (these can be predefined or chosen by you).
As you use EDIA, your entries will be saved automatically when you click the button.

Task 2

In this form, we ask you to reflect on the activity you completed using the EDIA software.
It is not necessary to upload the entries from Task 2 in this form, as they are automatically stored in the database, as shown in the video.
We will also begin to think about the final project, which will include classroom practice.

Optional Material

If you enjoy listening to podcasts, Luciana recently participated in one with Jocelyn Dunstan, the professor from Chile who spoke in the first class. They discussed some of these topics, along with others, such as Luciana’s research on natural language processing and images, as well as her role in scientific associations like NAACL and Khipu. This is optional, but if you’re interested, you can access and listen to the podcast through this link.