How to Attend or Follow Stanford CS336 Lectures

Updated

Stanford CS336 (Language Modeling from Scratch) is offered in Spring 2026 with in-person lectures on Mondays and Wednesdays, 3:00–4:20pm in Skilling Auditorium, and a public YouTube playlist for recordings. If you are enrolled or affiliated with Stanford, you can attend in person and use Slack for course questions and announcements; if you are following from outside, the YouTube playlist is the main way to watch lectures, while office hours and Slack are oriented toward enrolled students.

Lecture Schedule and Location

  • Days and time: Monday and Wednesday, 3:00–4:20pm
  • Location: Skilling Auditorium
  • Term: Spring 2026 (previous offerings: Spring 2025, Spring 2024)

The course site lists the schedule alongside a YouTube playlist, so the lecture sequence and recordings are aligned with the same weekly cadence.

Recordings

Lecture recordings are available through the course's YouTube playlist, linked from the Schedule section of the site. This is the primary option for anyone who cannot attend in person or wants to review a lecture after the fact.

Office Hours

Office hours are held by the instructors and CAs at specific times and locations:

Staff Day and Time Location
Percy Liang (Instructor) Fridays, 11am–12pm Gates 366
Tatsu Hashimoto (Instructor) Tuesdays, 11–12am Gates 364
Marcel Rød (CA) Tuesdays 4:30–5:30pm; Wednesdays 4:30–5:30pm Gates 498; Gates 415
Herman Brunborg (CA) Wednesdays 1:30–2:30pm; Fridays 1:30–2:30pm Gates 392
Steven Cao (CA) Mondays 4:30–5:30pm; Thursdays 9:30–10:30am Gates 200

Note that the listed Tuesday time for Tatsu Hashimoto reads "11–12am" on the course page; confirm the exact slot before planning to attend.

Questions and Announcements

  • Course-related questions: Ask in the public Slack channels.
  • Announcements: All announcements are also made in Slack.
  • Personal matters: Email [email protected].

This means Slack is the default channel for both peer and staff interaction during the term, and email is reserved for individual or private issues.

What to Expect If You Follow Along

The course walks through the full process of building a language model from scratch: data collection and cleaning for pre-training, transformer construction, training, and evaluation before deployment. The site notes that assignments involve minimal scaffolding and that students write substantially more code than in typical AI classes, so following the lectures is most useful alongside hands-on implementation.

Prerequisites listed by the course are proficiency in Python and experience with deep learning and systems optimization, including strong familiarity with PyTorch and basic systems concepts such as the memory hierarchy.