Stanford CS336 | Language Modeling from Scratch
Official course website for Stanford CS336: Language Modeling from Scratch (Spring 2026), including logistics, schedule, assignments, and course mater...
Academic Policy
Summary
This document sets guidelines for AI assistants in Stanford CS336, defining them as teaching aids rather than solution generators. It prohibits writing code or completing assignments while encouraging dialog-based debugging and concept explanation to preserve student learning.
Highlights
- AI agents must function as teaching assistants, focusing on explanation and guidance rather than generating solutions.
- Prohibited actions include writing Python/pseudocode, completing TODOs, editing student repos, or implementing core components like transformers or optimizers.
- Allowed behaviors involve explaining error messages, pointing to course materials, and suggesting sanity checks or debugging strategies.
- The teaching approach emphasizes asking clarifying questions and referencing lecture materials to help students build understanding independently.
- The policy aims to preserve the implementation-heavy learning experience of the course by preventing AI from bypassing the coding effort.
auto-generated
Context
Audience
Students and AI developers
DomainComputer Science Education
Formatpolicy document
Accessopen source
Source Overview
This document defines strict guidelines for how AI coding assistants may interact with students in Stanford CS336, emphasizing a teaching and debugging support role rather than generating solutions.
It specifies allowed behaviors such as explaining concepts, referencing course materials, reviewing and critiquing student-written code, and guiding debugging through questions and suggested checks. It prohibits writing code or pseudocode, completing assignment TODOs, providing solutions or core implementations, pointing to third-party solutions, or directly fixing student code, and outlines a dialog-based teaching approach to maintain academic integrity.
Topics
Related
Stanford CS336AI EthicsAcademic IntegrityPythonPyTorchLarge Language Models
