CS234: Reinforcement Learning Winter 2025

Course Description & Logistics

To realize the dreams and impact of AI requires autonomous systems that learn to make good decisions. Reinforcement learning is one powerful paradigm for doing so, and it is relevant to an enormous range of tasks, including robotics, game playing, consumer modeling and healthcare. This class will provide a solid introduction to the field of reinforcement learning and students will learn about the core challenges and approaches, including generalization and exploration. Through a combination of lectures, and written and coding assignments, students will become well versed in key ideas and techniques for RL. Assignments will include the basics of reinforcement learning as well as deep reinforcement learning — an extremely promising new area that combines deep learning techniques with reinforcement learning.

Communication: We will use Ed discussion forums. We encourage all students to use Ed for the fastest response to your questions.

Lectures will be live every Tuesday and Thursday: Videos of the lecture content will also be made available to enrolled students through canvas.
Office hours: Will be announced in the first week of class

Platforms: All assignments and quizzes will be handled through Gradescope, where you will also find your grades. We will send out links and access codes to enrolled students through Canvas.

You can find Spring 2024 materials here.

Course Instructor

Emma Brunskill

Course Assistants

Hengyuan Hu

Shiyu Zhao

Keenon Werling

Prerequisites for This Class

Proficiency in Python
All class assignments will be in Python. There is a tutorial here for those who aren't as familiar with Python. If you have a lot of programming experience but in a different language (e.g. C/ C++/ Matlab/ Javascript) you will probably be fine.
College Calculus, Linear Algebra (e.g. MATH 51, CME 100)
You should be comfortable taking derivatives and understanding matrix vector operations and notation.
Basic Probability and Statistics (e.g. CS 109 or other stats course)
You should know basics of probabilities, Gaussian distributions, mean, standard deviation, etc.
Foundations of Machine Learning
We will be formulating cost functions, taking derivatives and performing optimization with gradient descent. Either CS 221 or CS 229 cover this background. Some optimization tricks will be more intuitive with some knowledge of convex optimization.

Learning Outcomes

By the end of the class students should be able to:

Define the key features of reinforcement learning that distinguishes it from AI and non-interactive machine learning (as assessed by the exam).
Given an application problem (e.g. from computer vision, robotics, etc), decide if it should be formulated as a RL problem; if yes be able to define it formally (in terms of the state space, action space, dynamics and reward model), state what algorithm (from class) is best suited for addressing it and justify your answer (as assessed by the exam).
Implement in code common RL algorithms (as assessed by the assignments).
Describe (list and define) multiple criteria for analyzing RL algorithms and evaluate algorithms on these metrics: e.g. regret, sample complexity, computational complexity, empirical performance, convergence, etc (as assessed by assignments and the exam).
Describe the exploration vs exploitation challenge and compare and contrast at least two approaches for addressing this challenge (in terms of performance, scalability, complexity of implementation, and theoretical guarantees) (as assessed by an assignment and the exam).

Course Lecture Materials (Videos and Slides)

See the Lecture Materials page.

Draft Course Schedule

	Monday	Tuesday	Wednesday	Thursday	Friday	Saturday	Sunday
Week 1	Jan 6	Jan 7	Jan 8	Jan 9	Jan 10	Jan 11	Jan 12
Lecture Materials		Introduction to Reinforcement Learning 1:30pm-2:50pm		Tabular MDP Planning 1:30pm-2:50pm [Assignment 1 Released]
Week 2	Jan 13	Jan 14	Jan 15	Jan 16	Jan 17	Jan 18	Jan 19
Lecture Materials		Policy Evaluation 1:30pm-2:50pm		Q-Learning and Function Approximation 1:30pm-2:50pm	Assignment 1 Due at 6pm [Assignment 2 Released]
Week 3	Jan 20	Jan 21	Jan 22	Jan 23	Jan 24	Jan 25	Jan 26
Lecture Materials		Policy Search 1 1:30pm-2:50pm		Policy Search 2 1:30pm-2:50pm
Week 4	Jan 27	Jan 28	Jan 29	Jan 30	Jan 31	Feb 1	Feb 2
Lecture Materials		Policy Search 3 1:30pm-2:50pm		Offline RL 1 / Imitation learning 1:30pm-2:50pm	Assignment 2 Due at 6pm
Week 5	Feb 3	Feb 4	Feb 5	Feb 6	Feb 7	Feb 8	Feb 9
Lecture Materials		Offline RL 2 / DPO 1:30pm-2:50pm		Midterm (in class)	[Assignment 3 released]
Week 6	Feb 10	Feb 11	Feb 12	Feb 13	Feb 14	Feb 15	Feb 16
Lecture Materials		Offline RL 3 1:30pm-2:50pm		Exploration 1 1:30pm-2:50pm
Week 7	Feb 17	Feb 18	Feb 19	Feb 20	Feb 21	Feb 22	Feb 23
Lecture Materials		Exploration 2 1:30pm-2:50pm		Exploration 3 1:30pm-2:50pm	Assignment 3 Due at 6pm
Week 8	Feb 24	Feb 25	Feb 26	Feb 27	Feb 28	Mar 1	Mar 2
Lecture Materials		Guest Lecture 1:30pm-2:50pm		Monte Carlo Tree Search / AlphaGo	Project Milestone Due at 6pm
Week 9	Mar 3	Mar 4	Mar 5	Mar 6	Mar 7	Mar 8	Mar 9
Lecture Materials		Monte Carlo Tree Search / AlphaGo 1:30pm-2:50pm		Quiz (in class) 1:30pm-2:50pm
Week 10	Mar 10	Mar 11	Mar 12	Mar 13	Mar 14	Mar 15	Mar 16
Lecture Materials		Wrap Up and Alignment		Final Project Poster Session 1:30pm-4:30pm
Week 11	Mar 17	Mar 18	Mar 19	Mar 20	Mar 21	Mar 22	Mar 23
Lecture Materials		Final Project Writeup Due at 6pm

Textbooks

There is no official textbook for the class but a number of the supporting readings will come from:

Reinforcement Learning: An Introduction, Sutton and Barto, 2nd Edition. This is available for free here and references will refer to the final pdf version available here.

Some other additional references that may be useful are listed below:

Reinforcement Learning: State-of-the-Art, Marco Wiering and Martijn van Otterlo, Eds. [link]
Artificial Intelligence: A Modern Approach, Stuart J. Russell and Peter Norvig.[link]
Deep Learning, Ian Goodfellow, Yoshua Bengio, and Aaron Courville. [link]
David Silver's course on Reinforcement Learning [link]

Grade Breakdown

Assignment 1: 10%
Assignment 2: 18%
Assignment 3: 18%
Midterm: 25%
Quiz: 5%
Course Project: 24%

Proposal: 1%
Milestone: 2%
Poster Presentation: 5%
Paper: 16%

0.5% bonus for participating [answering lecture polls for 80% of the days we have lecture with polls. You may participate in these remotely as well. These are due by Sunday at 6pm for the week of lecture. You should complete these by logging in with your Stanford sunid in order for your participation to count.]

Late Day Policy

You can use 5 late days total.
A late day extends the deadline by 24 hours.
You are allowed up to 2 late days for assignments 1, 2, 3, project proposal, and project milestone, not to exceed 5 late days total. You may not use any late days for the project poster presentation and final project paper. For group submissions such as the project proposal and milestone, all group members must have the corresponding number of late days used on the assignment, and if one or more members do not have a sufficient amount of late days, all group members will incur a grade penalty of 50% within 24 hours and 100% after 24 hours, as explained below.
If you use two late days and hand an assignment in after 48 hours, it will be worth at most 50%. If you do not have enough late days left, handing the assignment within 1 day after it was due (adjusting for the late days used) will be worth at most 50%. No credit will be given to assignments handed in after 24 hours they were due (adjusting for any late days. E.g. if you use 2 late days, then after this policy applies 24 hours after your 2 late days, e.g. after 72 hours). Please contact us if you think you have an extremely rare circumstance for which we should make an exception. This policy is to ensure that feedback can be given in a timely manner.

Exams

There will be one midterm and one quiz. See the schedule for the dates.
Exams will be held in class for on-campus students.

Conflicts: If you are not able to attend the in class midterm and quizzes with an official reason, please email us at cs234-spr2324-staff@lists.stanford.edu, as soon as you can so that an accommodation can be scheduled. (Historically this is either to ask you to take the exam remotely at the same time, or to schedule an alternate exam time).

Notes for the exams: You are welcome to bring a 1-sided 1 (letter sized) page of handwritten notes to the midterm. For the quiz you are welcome to bring a double sided (letter sized) page of handwritten notes. No calculators, laptops, cell phones, tablets or other resources will be allowed.

Assignments and Submission Process

Assignments: See Assignments page where all the assignments will be posted.

Computing Resources: We will have some cloud resources available for later assignments.

Submission Process: The submission instructions for the assignments can also be found on the Assignments page.

Communication

We believe students often learn an enormous amount from each other as well as from us, the course staff. Therefore to facilitate discussion and peer learning, we request that you please use Ed for all questions related to lectures and assignments.

For SCPD students, if you have generic SCPD specific questions, please email scpdsupport@stanford.edu or call 650-741-1542. In case you have specific questions related to being a SCPD student for this particular class, please contact us at cs234-spr2324-staff@lists.stanford.edu.

For exceptional circumstances that require us to make special arrangements, please email us at cs234-spr2324-staff@lists.stanford.edu. For example, such a situation may arise if a student requires extra days to submit a homework due to a medical emergency, or if a student needs to schedule an alternative midterm date due to events such as conference travel etc. They will be considered and approved on a case by case basis.

Regrading Requests

If you think that the course staff made a quantifiable error in grading your assignment or exam, then you are welcome to submit a regrade request. Regrade requests should be made on gradescope and will be accepted for three days after assignments or exams are returned.
Note that while doing a regrade we may review your entire assigment, not just the part you bring to our attention (i.e. we may find errors in your work that we missed before).

Academic Collaboration, AI Tools Usage and Misconduct

I care about academic collaboration and misconduct because it is important both that we are able to evaluate your own work (independent of your peer’s) and because not claiming others’ work as your own is an important part of integrity in your future career. I understand that different institutions and locations can have different definitions of what forms of collaborative behavior is considered acceptable. In this class, for written homework problems, you are welcome to discuss ideas with others, but you are expected to write up your own solutions independently (without referring to another’s solutions). For coding, you may only share the input-output behavior of your programs. This encourages you to work separately but share ideas on how to test your implementation. Please remember that if you share your solution with another student, even if you did not copy from another, you are still violating the honor code. Consistent with this, it is also considered an honor code violation if you make your assignment solutions publicly available, such as posting them online or in a public git repo.

We may run similarity-detection software over all submitted student programs, including programs from past quarters and any solutions found online on public websites. Anyone violating the Stanford University Honor Code will be referred to the Office of Judicial Affairs. If you think you made a mistake (it can happen, especially under stress or when time is short!), please reach out to Emma or the head CA; the consequences will be much less severe than if we approach you. We expect all students to submit their own solutions to CS234 homeworks, exams and quizzes, and for projects. You are permitted to use generative AI tools such as Gemini, GPT-4 and Co-Pilot in the same way that human collaboration is considered acceptable: you are not allowed to directly ask for solutions or copy code, and you should indicate if you have used generative AI tools. Similar to human collaboration help, you are ultimately responsible and accountable for your own work. We may check students' homework, exams and projects to enforce this policy.

Note that it is not acceptable to list a LLM as a collaborator on the project milestone or final report: as things stand, generative AI cannot accept fault or responsibility, and thus cannot be a collaborator in a final project.

Academic Accommodation

If you need an academic accommodation based on the impact of a disability, please share your Office of Accessible Education letter with us via an email to our course staff list as soon as it is convenient for you. This helps us ensure the course materials and staff support can comply with your needs. The OAS is located at 563 Salvatierra Walk (650-723-1066, http://studentaffairs.stanford.edu/oae).

Credit/No Credit Enrollment

If you're enrolled in the class on credit/no credit status, you will be graded on work as usual per standard Stanford rules. The only distinction with those taking the class for letter grade is that you must obtain a C- (C minus) grade or higher in the class, for you to be marked as CR. In past years, the threshold for obtaining a C- is typically 70%.