We will follow a role-playing format, inspired by Jacobson and Raffel. Most important: think deeply about the papers we read, and try to learn from them as much as possible (and then even more). If you do not understand something, we should discuss it and dissect it together. Whatever you think others understand, they understand less (instructor included), but together we will get it. As you read papers, please consider the following guidelines:

Identify the key questions the paper studies, and the answers it provides to these questions.
Consider the challenges of the problem or scenario studied, and how the paper’s approach addresses them.
Deconstruct the formal and technical parts to understand their fine details.
Note to yourself aspects that are not clear to you.

Each student assigned a role will discuss each paper within their role. The student will use slides added to a collaborative slidedeck before the class. The presentation may also use screen sharing (e.g., to show demos or browse data). You can create multiple consecutive slides. Except the presenter role, the maximum slides per role is 3-4 slides. Roles may be skipped if not appropriate for a paper. Aim for about 5min presentation/discussion for each role, except for the presenter (see guidelines below).

Paper Owner

Each paper has an owner. The owner selects the paper, and takes the presenter role. The owner also assigns all other roles using the assignment spreadsheet. When assigning roles, the owner has to take into account both what the student's existing assignment to maintain balance and the student's expertise. All assignments must be done at least a week before the paper presentation. The earlier you make your assignments, the more flexibility you will have.

Owner checklist, to be completed at least a week before the paper presentation:

📜 Identify the paper to read and confirm with Yoav. You can select a paper from the list of potential papers, or propose your own. If you are proposing your own, justify your choice and make sure the paper is of high quality.
👥 Assign all the roles using the role assignment spreadsheet, and notify all the assigned students (get an ack from them). The spreadsheet accounts for what roles students fulfilled. Please assign roles with the goal of balance and diversity.
🎥 Create a presentation deck by cloning and editing the template. The deck should be located in the shared folder.
📢 Announce the paper on Ed discussion. The announcement should include the title, author, abstract, and a link to the paper. If the paper is behind a paywall, please attach its PDF. Please also include a link to the collaborative deck.

The Roles

🧑‍🏫 Presenter

You are the first author of the paper, and you need to present the paper in a conference. Prepare a brief talk (10-15min) presenting the paper, including the motivation, problem, method, and results. You may use existing presentation material (e.g., slides from the original authors). Please may add the slides the collaborative deck, or present them from a separate deck. If presenting from a shared deck, add an accessible link to the collaborative deck.

👩🏽‍🔬 Scientific Peer Reviewer

The paper has not been published yet and is currently submitted to a top conference where you’ve been assigned as a peer reviewer. Complete a full NeurIPS-style review of the paper. This includes recommending whether to accept or reject the paper. Please follow the COLM reviewing guidelines.

🏺 Archaeologist

This paper was found buried under ground in the desert. You are an archeologist who must determine where this paper sits in the context of previous and subsequent work. Find and report on one older paper cited within the current paper that substantially influenced the current paper and one newer paper that cites this current paper. If the paper is too new to have signficant follow up, report on two previous works. Please try as much possible to thread the two papers into a single research narrative.

🍜 Academic Researcher

You are a researcher who is working on a new project in this area. Propose an imaginary follow-up project not just based on the current but only possible due to the existence of the current paper. When reading a paper outside of your area of research, it is advised to design a follow-up project in your area, or at least closer to your area of expertise. If your project requires data, please suggest how you will receive the data. Also, show examples of inputs and outputs. Clearly form the research questions you aim to study.

💰 Industry Practitioner

You work at a company or organization developing an application or product of your choice (that has not already been suggested in a prior session). Bring a convincing pitch for why you should be paid to implement the method in the paper, and discuss at least one positive and negative impact of this application. If the paper is does not present a new method (e.g., analysis or human study), instead of implementing consider incorporation of findings and/or proposed framework. Try to keep your discussion concrete, for example by advocating for a specific product or application. Propose avenues to address the requirements of the approach (e.g., as far as data, evaluation, compute, etc).

👾 Hacker

You are a hacker who needs a demo of this paper ASAP. Implement a small part or simplified version of the paper on a small dataset or toy problem. Prepare to share the core code of the algorithm to the class and demo your implementation. Do not simply download and run an existing implementation – though you are welcome to use (and give credit to) an existing implementation for “backbone” code.

🕵️ Private Investigator

You are a detective who needs to run a background check on one of the paper’s authors. Where have they worked? What did they study? What previous projects might have led to working on this one? What motivated them to work on this project? How does it contribute to their overall body of work (maybe it doesn't)? Feel free to contact the authors, but remember to be courteous, polite, and on-topic. Please do not depend on author response though. Please conclude with a slide synthesizing explicitly what perspective each author brought to this paper, which others couldn't bring (there is not always such contribution).

Identify how this paper self-assesses its (likely positive) impact on the world. Have any additional positive social impacts left out? What are possible negative social impacts that were overlooked or omitted?