Stanford AI 课程地图 EN GitHub ↗
← 返回课程地图
Stanford · 2026—27 CS 329T LLM 与 Agent ★★★☆☆

核心课程 · LLM 与 Agent

可信机器学习:构建与评估 Agent 系统

Trustworthy Machine Learning: Building and evaluating agentic systems

可信 agent 系统:先搭 prototype,再用 eval 把它打磨到可靠。

3 学分 2026–27 未排课 Letter or Credit/No Credit ★★★☆☆
Agent可信Eval项目

2026-27 学年未在 ExploreCourses 排课。

课程定位

CS 329T 是一门项目制课程,主题是“prototype 很容易,可靠很难”。课程覆盖 foundation model、prompting、RAG、agentic 架构,并为每种架构配套评估方法;学生最终要用评估驱动的方式做一个可信的 agent 应用。

主要内容

  1. FM 与 prompting:能力边界、提示模式
  2. RAG:检索、引用、幻觉控制
  3. Agentic 架构:规划、工具、记忆、多 agent
  4. 评估:离线指标、LLM-as-judge、人工评估、红队
  5. 可靠性:失败模式、鲁棒性、监控
  6. 团队项目:构建并评估一个 agent 应用

先修

CS 229 或同等 Python ML 课;深度学习(CS 230/CS 231N);熟悉 scikit-learn/Keras 等框架。

考核与工作量

入门作业(RAG/agent)+ 小组项目 + 评估报告。具体以当季官网为准。

适合谁

想做 agent 产品、AI 评测、可信 AI 的人;已经会搭 demo、想补 eval 方法论的人。

学习建议

和 CS 329Z 搭配:329Z 更偏工程全栈,329T 更偏“可信 + 评估”。项目一定要先定义指标再写代码。

相关课程

官方课程描述(英文原文)

This project-based course will introduce students to building and evaluating agentic AI applications powered by foundation models. The overriding theme of the course is that building an initial prototype AI system can often be completed easily but refining a prototype into something that is useful and reliable requires iterative improvement based on clear evaluation metrics. We will cover background in foundation models, prompting, and retrieval-augmented generation (RAG) before introducing full agentic AI architectures. For each architecture, the course will study methods for evaluation. Students will complete introductory homework assignments to become familiar with retrieval-augmented generation (RAG) and agentic AI. Students will then work in pairs or small teams to develop applications using agentic or other approaches and evaluate them by adapting evaluation methods presented in the class.Prerequisites: CS229 or similar introductory Python-based ML class; knowledge of deep learning such as CS230, CS231N; familiarity with ML frameworks in Python (scikit-learn, Keras) assumed.