Course Project

Overview

The goal of the course project is to give students hands-on experience in a research area of their choice, related to the topics covered in class. Teams of 2 or 3 students are recommended. Solo projects are allowed with a justification (e.g., if the project is part of your ongoing research).

The following types of projects are welcome and encouraged:

  • Paper reproduction: Reproduce any paper discussed in class (presented, upcoming, or listed as suggested reading), or related work from ML/NLP conferences. Your project should:
    • Include a subset of the original experiments, along with meaningful extensions (e.g., new datasets, base models, or ablations).
    • Provide novel insights or results beyond the original paper.
    • Meta-reproducibility studies across related papers are also encouraged.
  • Discussion-inspired projects: Develop ideas inspired by in-class discussion.
    • If you’d like to pursue an idea proposed by a classmate, offer them to join your team, and they can choose to join or opt out.
  • Ongoing research: You may build on your current research (including work with collaborators outside the class), as long as it clearly relates to the course topic and is noted in all deliverables (see the project FAQ). If unsure about relevance, consult the instructor.

Please keep in mind the compute resources available to you. The instructor can’t provide compute resources for the project.

There are three main deliverables: project proposal, the project presentation, and the final paper. See below for details (all deadlines are at 6pm PST).

Milestone 1: Project matching survey [deadline: 09/17 Wed]

  • Fill out the project matching survey (to be posted on Slack) with your preferred teammates and project topic (optional). The topic does not have to be finalized; it will only primarily be used for team matching.
  • By 09/19 (Fri), project team assignment will be announced on Slack.

Milestone 2: Project proposal [deadline: 10/09 Thu]

  • Format: Use the unmodified COLM template (find the latex template here).
  • Up to 3 pages, not including references. You are welcome to include an Appendix with no page limit, but the evaluation will primarily be based on the main content.
  • Your proposal must have the following sections: Abstract, Introduction, Related work, Proposed approach, Experiments and evaluation plan (including baselines, evaluation datasets, and metrics), (Optional) Preliminary results, Timeline, Contribution statement (thus far, and plans going forward), and References.

Required, 30-min meeting to discuss project proposal (10/13–17)

  • The team should have a 30-min meeting with the instructor to discuss the project proposal. All team members should be present in the meeting.
  • There will be the calendar link where you choose the slot (to be posted on Slack).

Milestone 3: Project presentation (11/18 and 11/20)

  • Time: 15 min total per team (this can change depending on the number of teams we have)
    • ~10 min for the talk, split between all team members
    • ~5 min for QnA
  • The goal of the presentation is to convey the important high-level ideas and takeaways of your project, rather than all the details.
  • All group members should participate in the presentation. You can split it any way that you see fit, as long as each person presents a significant chunk of it.
  • Demos are strongly encouraged where possible!

Optional meetings to discuss the final paper (11/24–12/10)

  • The team can have optional, up to three 30-min meetings with the instructor to discuss the final paper.
  • Send your draft paper and meeting agenda (e.g., requested feedback topics) to the instructor 24 hours in advance.
  • There will be the calendar link where you choose the slot (to be posted on Slack).

Milestone 4: Final paper [deadline: 12/10 Wed]

  • Format: Use the unmodified COLM template. Find the latex template here.
  • Up to 8 pages, not including references. You are welcome to include an Appendix with no page limit, but the evaluation will primarily be based on the main paper.
  • The final paper should be comparable in quality to a conference or workshop submission. If you’re interested in submitting your work for real, feel free to reach out to the instructor for help.

Project FAQ

Using the same project for CS294-288 and another class

While the projects can be related and use a shared codebase, you may not submit an identical project as another class project. If any part of the project is done for another course, please clearly indicate in the Contributions section of your report which part of the project was done for CS294-288 and which part was not.

Using your ongoing research as your CS294-288 project

This is allowed. In the Contributions section of your writeup, you should indicate this and describe which parts were done prior to the start of the course vs. which parts were done for the course. You will be evaluated on the parts that were done after the start of the course, i.e., you may not reuse a previously completed project.

Collaborating with people outside this course

This is allowed (e.g., your advisor and labmates might be involved in your ongoing research). In the Contributions section of your writeup, you should indicate this and describe which parts you were responsible for. If you are repurposing text or slides that were written by your collaborators, you should also declare this. You will be evaluated on the parts that you worked on.

Generative AI policy

See the course policies page.