CSCI-B 490 — Generative AI and Applications

Fall 2026 · Course Schedule

Monday & Wednesday, 75 minutes · August 24 – December 18, 2026

Contacting the teaching team
For anything about the course, email the whole teaching team at AI Group Email. It reaches the instructor and both Associate Instructors at once, so you get an answer from whoever is free rather than waiting on one person.
b490_associated_instr-aaaav3o2rmkyj6rzpbxyvqreeu@arkai-lab.slack.com
Your workspace now stops after 5 hours
Workspaces auto-stop 5 hours after they start, shortened from 8. The cluster itself kills any pod that runs past 6 hours, and when that happened the workspace looked like it was still running while your app was actually gone. Stopping first, on our own schedule, avoids that.
What this means for you. Anything in ~/persistent is safe, and so is everything you installed in your home directory. Just start the workspace again. If your Streamlit app stops responding but the page still says the workspace is running, stop it and start it again — that is this, and it is not something you did wrong.
Anonymous feedback — open all semester
One question, completely anonymous — no sign-in, no name, no email attached. You are welcome to send feedback any time during the semester, as often as you like. Tell me what would make this course better →
● INSTRUCTOR
Thai Le
◆ ASSOCIATE INSTRUCTOR
Binh Nguyen
◆ ASSOCIATE INSTRUCTOR
Charlie Furlow
Office hours: Tuesdays 11:30–12:30 on Zoom Join
Meeting ID 772 0882 0206 · Passcode 123
Other times can be arranged. Email AI Group Email to ask.
LECTURELABREADINGQUIZMILESTONE
Unit 1 — Foundations
Machine learning, embeddings, NLP and transformers, and the generative AI landscape
Mon Aug 24
LECTURECourse Overview, Logistics, and Policies
LABEnvironment Setup: Coder, LLM Access
Wed Aug 26
NOTESet up your environment before class — see the previous lab
LECTUREThe Generative AI Landscape
LECTUREResponsible AI: What Can Go Wrong?
QUIZIn-class Quiz #1
Mon Aug 31
MILESTONEHW1 RELEASE
NOTE
Already started HW1? Read this before re-copying the lab folder (show details)
Do not re-run cp -r ~/course/unit1 ~/persistent/hw1 if that folder already exists. The copy lands inside it as hw1/unit1/ . Your work is safe, but you end up with two copies and it is easy to edit the one that has no effect. To get just the finished lab file instead: cp ~/course/unit1/src/lab1_rag_chatbot.py ~/persistent/hw1/src/
LECTUREBasics of Machine Learning, Neural Networks, Representation Learning
Wed Sep 2
NOTE
Already started HW1? Read this before re-copying the lab folder (show details)
Do not re-run cp -r ~/course/unit1 ~/persistent/hw1 if that folder already exists. The copy lands inside it as hw1/unit1/ . Your work is safe, but you end up with two copies and it is easy to edit the one that has no effect. To get just the finished lab file instead: cp ~/course/unit1/src/lab1_rag_chatbot.py ~/persistent/hw1/src/
LECTUREBasics of Machine Learning, Neural Networks, Representation Learning (cont.)
LAB 1Your First AI App: a RAG Chatbot
~/course/unit1 from your workspace
Mon Sep 7
NOTELabor Day — no class
Wed Sep 9
LAB 1Your First AI App: a RAG Chatbot (cont.)
~/course/unit1 from your workspace
QUIZIn-class Quiz #2
Mon Sep 14
▶ WE ARE HERE
LECTUREBasics of Machine Learning, Neural Networks, Representation Learning (cont.)
Wed Sep 16
LECTURENLP History, Transformers, Transfer Learning (cont.)
QUIZIn-class Quiz #3
Unit 2 — Large Language Models
LLMs and chatbots, prompt engineering, RAG, fine-tuning and alignment
Mon Sep 21
MILESTONEMIDTERM PROJECT RELEASE
LECTURELLMs and Chatbots
Wed Sep 23
LECTUREPrompt Engineering, Context Engineering, and RAG
LECTUREPrompt Injection and Defenses
QUIZIn-class Quiz #4
Mon Sep 28
LECTURELLM Evaluations
LABEvaluating LLMs Across Prompting Strategies
Wed Sep 30
LECTUREReasoning Patterns and Loop Engineering
QUIZIn-class Quiz #5
Mon Oct 5
LECTUREReasoning Patterns and Loop Engineering (cont.)
LABLoop Engineering
Wed Oct 7
LECTURECustomizing LLMs: Fine-tuning, PEFT, and RLHF
QUIZIn-class Quiz #6
Fri Oct 9 – Sun Oct 11
NOTEFall Break — no effect on Mon/Wed; classes resume Oct 12
Mon Oct 12
LABFine-tuning with PEFT and MS-Swift
Wed Oct 14
MILESTONEMIDTERM SUBMISSION DUE
LABIntroduction to Agentic Coding
Mon Oct 19
MILESTONEMIDTERM PROJECT PRESENTATIONS
Unit 3 — Agentic AI
Agentic patterns, reflection and tool calling, multi-agent planning, evaluation, and trustworthy AI
Wed Oct 21
MILESTONEFINAL PROJECT RELEASE
LECTUREAgentic AI: Patterns, Memory, and Perception
LABBuild your First AI Agent
Mon Oct 26
LECTUREAgentic AI: Reflection and Tool Use
Wed Oct 28
LABImprove your Agent
Mon Nov 2
LABImprove your Agent (cont.)
Wed Nov 4
LECTUREMulti-Agent AI: Planning and Workflow
LABBuild a Multi-Agent System
Mon Nov 9
LECTUREMulti-Agent AI: Planning and Workflow (cont.)
LABBuild a Multi-Agent System (cont.)
Wed Nov 11
LECTUREMulti-Agent AI: Evaluation and Optimization
LABEvaluate your Multi-Agent System
Mon Nov 16
LABMini Multi-Agent Project
Wed Nov 18
LABMini Multi-Agent Project (cont.)
Mon Nov 23
NOTEThanksgiving Break — no class
Wed Nov 25
NOTEThanksgiving Break — no class
Mon Nov 30
MILESTONEFINAL PROJECT MILESTONE 1
LECTUREAI Interpretability, Observability, and Auditability
Wed Dec 2
LECTUREAI Interpretability, Observability, and Auditability (cont.)
Mon Dec 7
MILESTONEFINAL PROJECT INTERVIEW GROUP A
Wed Dec 9
MILESTONEFINAL PROJECT INTERVIEW GROUP B
Fri Dec 18
MILESTONEFINAL PROJECT SUBMISSION DEADLINE

Key dates

Date Milestone
Mon, Aug 31 HW1 released
Mon, Sep 21 Midterm project released
Wed, Oct 14 Midterm project submission due
Mon, Oct 19 Midterm project presentations
Wed, Oct 21 Final project released
Sun, Oct 25 Registrar: last day to drop with an automatic W
Mon, Nov 30 Final project milestone 1
Mon, Dec 7 Final project interview — Group A
Wed, Dec 9 Final project interview — Group B
Fri, Dec 18 Final project submission deadline

Links you will need

Where What for
CanvasAnnouncements, submissions, grades
Coder workspaceSign in with Indiana University
LLM API tokenGet your key here — treat it as a password
Open WebUIChat with the course models in a browser
Workspace Status CheckIs your workspace ready? Type your IU username to find your row

Grading

ComponentWeight
Participation & attendance20%
Labs & coding assignments30%
In-class quizzes10%
Midterm project15%
Final project25%

Quizzes — when they happen and what they cover

Quiz When What it covers
Quiz #1 Wed Aug 26 LECTURESCourse Overview, Logistics, and Policies; The Generative AI Landscape; Responsible AI: What Can Go Wrong?
Quiz #2 Wed Sep 9 READINGLECTURESRefresh: ML Basics, Discriminative vs Generative, Representation Learning  ·  Basics of Machine Learning, Neural Networks, Representation Learning
Quiz #3 Wed Sep 16 READINGLECTURESTransfer Learning for Boosting Neural Network Performance; A Brief History of NLP  ·  Basics of Machine Learning, Neural Networks, Representation Learning; NLP History, Transformers, Transfer Learning
Quiz #4 Wed Sep 23 READINGPrompt Engineering: a survey
Quiz #5 Wed Sep 30 READINGA Visual Guide to Reasoning LLMs
Quiz #6 Wed Oct 7 LECTURESReasoning Patterns and Loop Engineering; Customizing LLMs: Fine-tuning, PEFT, and RLHF

Quizzes are short, taken in class, and AI-free. If a session has a reading, the quiz comes from that reading. If it has no reading, it comes from the lectures of that session and the one before it. Where both tags appear, the quiz covers the reading and the lectures.

Attendance

Attendance is part of the 20% participation and attendance component, and it is checked during the session in various formats, a short in-class quiz, team discussion, or a lab submission. Arriving a few minutes late is not a penalty; failing to submit the attendance material is.
Two free misses. You may miss up to 2 attendances with no penalty, for any reason, and you do not need to explain them. Beyond 2, there are no accommodations without official IU documentation.

The schedule and grading policies are tentative and may be adjusted; changes will be announced in class and on Canvas.