Content:
AI models like ChatGPT and LLAMA show impressive performance in a variety of tasks, sometimes even beating humans. The outputs of these models are, however, quite difficult to interpret. Given an input to the AI model and corresponding output, it is difficult for users to understand “why” the model generated this output. Interpretability is a key desideratum for successful adoption of these models in the real world. As a result, a productive field of research has developed in recent years with the goal of trying to understand and explain the outputs of these models.
In this seminar, we will read and discuss recent advances in interpretability of AI models. We will read papers from AI conferences like ICML, NeurIPS and ACL.
Learning Outcomes:
In this seminar, you will learn about:
- Surveying, understanding and critically analyzing scientific literature
- Delivering presentations to an audience who is familiar with but not expert in your topic
- State-of-the-art research in AI, specifically, about interpretability and explainability of AI models.
We will use the following workflow:
- Kickoff meeting: To take place on 18.10.2024 at 14:00. Each course participant will be assigned a research paper during this meeting. Your goal for the remainder of the course is to thoroughly understand the paper and provide critical feedback.
- Mid-point meeting: To take place roughly halfway through the semester. You will discuss your progress and any challenges you face with your instructor.
- Summary of the paper: You will write a short summary of the research paper and submit it. You will also be required to write a review of the summaries written by your fellow students.
- Final presentation: To take place at the end of the teaching period. We will block several small sessions for the presentations.
Exam:
Oral presentation, summary of the paper and review of the summary.
Requirements for the Awarding of Credits:
- Delivering an oral presentation to the group on your assigned topic.
- Actively participating in discussions when your fellow students are presenting.
- Writing a summary of the paper and reviewing the summaries written by others.