@shao__meng: Stanford CS224V "Agentic AI" Course: How Monica Lam's Team Transforms Hallucination-Prone LLMs into Trustworthy, Accoun…

Stanford CS224V “Agentic AI” Course: How Monica Lam’s Team Transforms Hallucination-Prone LLMs into Trustworthy, Accountable Agents

Stanford CS224V “Agentic AI” https://web.stanford.edu/class/cs224v/index.html…

The entire course revolves around one central question: how to transform inherently unreliable, hallucination-prone LLMs into trustworthy, accountable agents. The course believes that such systems hold “extraordinary promise for accelerating scientific discovery and democratizing medical/legal/educational services,” but only if reliability is addressed first.

The course’s technical main thread: An “anti-hallucination” knowledge stack If you look at the 15 lectures and reading list together, the course actually builds a complete technical stack from the ground up, with each layer addressing “where does reliability come from”:

1. Base layer: The LLM itself
2. Knowledge Curation (STORM family 9/28 lecture STORM/Co-STORM):
3. Reliable Q&A on free text (RAG) Retrieval (ColBERT, RankGPT) + Generation + Verification
4. Agents on structured and heterogeneous data—this is the most distinctive part of the course
5. Formal methods and accountable AI (the course’s most unique label)
6. Task-oriented conversational Agent Genie Worksheet
7. Meta-Agents and meta-optimization (a forward-looking lecture)

The first two lectures’ slides are currently publicly available: https://web.stanford.edu/class/cs224v/lectures_2026/l-introduction.pdf…

Source: https://web.stanford.edu/class/cs224v/index.html

About CS 224V

AI agents powered by Large Language Models are already transforming how we work, learn, and solve problems — and we are only at the beginning. As these systems grow more capable, they hold extraordinary promise for accelerating scientific discovery and democratizing access to high-quality medical, legal, and educational services worldwide. This is a project course coupled with rigorous lectures on the principles, methodologies, and cutting-edge research underlying agentic AI. Students undertake a substantial quarter-long project in either foundational methodology research or building novel agents in a domain of their choice.

Topics include:

• Minimizing hallucination in question-answering and task-oriented agents using Retrieval-Augmented Generation (RAG) and formal task descriptions.
• Hybrid knowledge reasoning over databases, knowledge bases, and unstructured text.
• AI-driven knowledge curation and discovery for scientific research.
• Improving the accuracy and interpretability of decision-making agents through formal methods.
• Automated techniques for improving the accuracy and efficiency of long-horizon agents.

Office Hours

Vidhyakshaya KannanMon 1:00-3:00pm Gates 392 (B Wing)Harshit JoshiTue 4:00-5:00pm Gates 392Yu HeWed 4:30-6:30pm CoDa E206Mahathi MangipudiThu 8:00-10:00am ECON 106Jeongyeon KimFri 8:00-10:00am Huang 130 Additional mentors and outside advisors are listed with the individual projects in theproject proposals.

Logistics

Lectures:Monday/Wednesday 3:00-4:20pmin personinCoDa B80. Attendance is mandatory.

**Recordings:**video recordings of the lecture can be found onCanvas.

**Slides:**can be found on theScheduleand in the lecture slides folder onCanvas. Posted lecture slides are missing important details to facilitate student participation. Please make sure you watch the lectures.

**Homework:**can be found on theScheduleand submissions will be onGradescope.

**Contact:**Students should ask all course-related questions onEd, where you will also find announcements. For external inquiries, personal matters, or in emergencies, you can send an email to our staff email[email protected]

**Academic accommodations:**If you need an academic accommodation based on a disability, you should initiate the request with the Office of Accessible Education (OAE). The OAE will evaluate the request, recommend accommodations, and prepare a letter for the teaching staff. Once you receive the letter, send it to the course staff email at[email protected]. Students should contact the OAE as soon as possible since timely notice is needed to coordinate accommodations.

Audit Requests: To audit the course, please send an email to course staff email at[email protected], with the subject line “audit cs224v request.”

**Course Participation:**We offer the course on SCPD to serve remote students; it is not to allow students take conflicting courses. In-class attendance and participation are an important part of the course. We allocate 15% of the course grade to class participation, which is important to make the most out of the course.

1. If you are a local student, 5% of the course grade is allocated to in-class attendance and participation.
2. We allocate 10% and 15% of the grade to local and remote students, respectively, to (1) your weekend updates, and (2) interaction with your project mentor every week.
3. Contributions in helping others on Ed will be awarded with bonus points.

Enrollment

CS 224V has limited enrollment so as to provide adequate project and research supervision to students. The class is currentlywaitlist only(class #2029on Navigator).

**Prerequisites:**one of LINGUIST 180/280, CS 124, CS 224N, CS 224S, 224U.