This course explores active research in machine learning and AI, and topics may include probabilistic machine learning, generative AI models and their evaluation, AI agents, and applications. Students will gain broad exposure to these areas through recent papers, reflecting instructor interests. The course involves student-led weekly discussions of chosen papers, emphasizing motivation, context, and innovation. For every topic, an instructor will give a lecture on the background, facilitate group discussions and guide presentations. Students will also complete an in-depth course project on a chosen subject.
Liliya Lavitas, Zi Wang, Deqing Sun, Forrester Cole, Lily Zhang
Course correspondence will be through Canvas.
Friday 9:45am - 12:30pm
In the case of oversubscription, we will prioritize graduate students whose research interests will benefit from the material, but are open to other qualified and interested candidates. As an advanced course where we will discuss recent literature at a technical level, we do expect significant math, statistics and machine learning background.
Please submit your statement of interest using this form.
One of the following courses (or equivalent):
Class participation - 40%
Class presentations - 20%
Project proposal - 10%
Project presentation - 10%
Project report and code - 20%
We will consider the presence of AI-slop in presentation, proposal, code and report in final grading.
Each class meeting will be an in-depth discussion on a specific topic. Two students will present papers each week, and each student is expected to facilitate a discussion 1-2 times per semester. The presenters for each week are expected to coordinate with each other and with the course instructors in advance to divide up the assigned papers and any additional background material that will need to be discussed.
Discussions will center around:
Understanding the strengths and weaknesses of these methods.
Understanding the relationships between these methods, and with previous approaches.
Extensions or applications of these methods.
Experiments that might better illuminate their properties.
The ultimate goal is that these discussions will reveal gaps in current research, generate new ideas, and ideally generate novel research directions.
Students can work on projects individually or in pairs. The goal of the projects is to allow students to dive more deeply into one of the topics of the course. The project can be an extension of existing work, a novel application using existing methods, exploration of a new research idea or non-trivial implementation and experimentation using existing methods. The grade will depend on the ideas, how well you present them in the report, how clearly you position your work relative to existing literature, how illuminating your experiments are, and how well-supported your conclusions are.
Each group of students will write a short (2 pages max) research project proposal. It should include a problem description, “minimum viable product”, some nice-to-haves if time allows, and a short review of related work. Instructors will give feedback on the proposal and schedule office hours for discussions.
Towards the end of the course, everyone will present their project in a short presentation.
At the end of the class you'll hand in a project report (4 to 8 pages), prepared in the format of a machine learning conference paper such as NeurIPS or ICML.
Policy on dual submissions. It is not allowed to submit reports that are substantially similar to reports that the students finished before the course starts or submitted in parallel to other courses.
For course presentations, the presentations can be completed independently but we encourage students to collaborate to ensure there is less redundancy and overlap (e.g. by having similar introductions to a topic). However, we expect both students to each contribute substantially to the week's presentations. For the project, if students choose to work together, we will ask for a statement detailing the individual contributions of each student.
Attendance in person expected.
1 absence permitted per semester, unless otherwise discussed with the instructors.
Reading is expected to be done in advance, not during course time.
Note: this is a representative list of topics and papers and is subject to change. All topics are on generative AI.
Topic 1: AgentsTopic 2: Post-trainingTopic 3: EvaluationsTopic 4: MultimodalityExamples of papers suitable for presentations / discussion in class
[blog post] Harness Engineering for Self-Improvement, by Lilian Weng
[blog post] Scaling Laws, Carefully, by Lilian Weng
[educational video] How Attention Got So Efficient, by Jia-Bin Huang
Efficiently Reconstructing Dynamic Scenes One 🎯 D4RT at a Time
V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
3D Gaussian Splatting for Real-Time Radiance Field Rendering
Vision Banana: Image Generators are Generalist Vision Learners
[blog post] Diffusion Models for Video Generation, by Lilian Weng
[blog post] Learning the integral of a diffusion model, by Sander Dieleman
[blog post] Generative modelling in latent space, by Sander Dieleman
[blog post] Generative Modeling by Estimating Gradients of the Data Distribution, by Yang Song
Emerging Properties in Self-Supervised Transformers (DINO paper, possibly redundant with JEPA)
Multistep Distillation of Diffusion Models via Moment Matching
Simpler Diffusion: 1.5 FID on ImageNet512 with pixel-space diffusion
Gram: Assessing sabotage propensities via automated alignment auditing
Interpreting and Controlling Model Behavior via Constitutions for Atomic Concept Edits
[blog post] Demystifying evals for AI agents
AuditBench: Evaluating Alignment Auditing Techniques on Models with Hidden Behaviors
QuestBench: Can LLMs ask the right question to acquire information in reasoning tasks?
HiL-Bench (Human-in-Loop Benchmark): Do Agents Know When to Ask for Help?
Date TBD: Final project posters due (so that they can be printed before the poster session)
Date TBD: Final project poster session. Time TBD.
Date TBD: Final projects due
A project can contribute in any of these areas (or a combination of them):
Methods: systematic assessment of the strengths and weaknesses of a collection of novel or existing methods when applied to real or synthetic data.
Applications: use of machine learning to help solve a real-world problem.
Theory: formal statements concerning guarantees about machine learning problems or methods.
Exposition: presentation of a unified framework covering a set of existing theories or methods. The goal is to help provide accessible educational content to others as well as identify opportunities for development of novel methods and theory.
Software: development of machine learning / AI tools that are fast, general-purpose, and well-tested.
When evaluating your projects, we will be focusing on the following criteria:
Are your technical statements precise and correct?
Did you properly cite related work and explain the background concepts?
Given your specific machine learning / AI background, did the work stretch you outside of your comfort zone?
Is your write-up well-written and was your presentation engaging? If your project is software-based, was your code high-quality and reusable?
If working as part of a team, did you collaborate effectively?
Some example projects
Automatic failure or capability discovery for multimodal models or AI agents.
New evaluation approaches that improve evaluation efficiency.
Use existing AI tools to create a short educational movie on world models.
Write a tutorial covering the breadth of recent advancements in AI alignment.
Training Diffusion Language Models
Preventing “gaming” evals in LM Arena
An agent to perform a vision or graphics task such as rotoscoping