1. Course Information

• Instructors: Hanjie Chen, Sharon Levy 

• Semester: Spring 2024

• Time: Thursday 3:00 - 4:15 PM

• Location: Malone 107

2. Course Description

This is a graduate student seminar aimed at introducing graduate students to the research areas of trustworthy and responsible NLP. This is primarily targeted at students with an NLP background and no or very little background in the course topics: bias, privacy, safety, misinformation, explainability, interpretability, and robustness. Students will be expected to read, present, and discuss papers each week. The goal is to let students get familiar with the emerging problems and recent advances in trustworthy and responsible NLP.

2.1 Topics

This course will include but not be limited to the following contents: bias, privacy, safety, misinformation, explainability, interpretability, robustness, etc.

2.2 Seminar Structure

Students will be assigned 2 papers to read each week and will be asked to formulate 2-3 discussion questions that relate to each paper or the topic of the week. Each week, a student will give one short presentation (~15-20 minutes) on each paper in the reading list for that week and we will have a discussion session after each presentation regarding the paper and general area. Each student will be doing two short presentations on two different weeks. The discussions will focus on not only the NLP aspect of the topic but also social implications. We provide a reading list for each topic and allow students to choose the paper to present for that week.

3. Assignments and Evaluation

• Reading List

• Paper reading and discussion questions (10 * 5%)

• Paper presentations (25% + 25%)