| Instructors | Ashwin Srinivasan, Tirtharaj Dash |
|---|
The course handout can be found here. The course lectures will be interleaved between "Good Old Fashioned AI" (GOFAI) and "New Fashioned AI" (NewFAI). GOFAI lectures will be taken by AS (mostly), and NewFAI lectures will be taken by TD (mostly). The official course portal will be Moodle: click here. This GitHub repository is a secondary source for almost everything about the course and can be used by students who are not officially registered.
| Week | Content |
|---|---|
| Week 1 | Introduction to AI (Past, Present, Future) |
| NewFAI: Introduction to Foundation Models I slides | |
| NewFAI: Introduction to Foundation Models II Ch1-2, video | |
| Week 2 | GOFAI: Intelligent Agents slides |
| GOFAI: Problem Solving Agents slides | |
| NewFAI: Prompt Engineering slides (Credit) | |
| Week 3 | Stuart Russell's Reith Lecture Part 1 view |
| GOFAI: More on search and agents | |
| NewFAI: LLMs as agents slides (credit: UC Berkeley) | |
| Week 4 | GOFAI: Probabilistic Intelligence (Basics} slides |
| GOFAI: Probabilistic Intelligence (Representation) slides | |
| GOFAI: Probabilistic Intelligence (Inference) slides | |
| Week 5 | NewFAI: LLMs with tools (1) paper |
| NewFAI: LLMs with tools (2) slides (Credit: Yu Meng@UVA) | |
| NewFAI: LLMs calling tools slides | |
| Stuart Russell's Reith Lecture Part 2 view | |
| Week 6 | GOFAI: Probabilistic Intelligence (Autoregressive Models) slides |
| GOFAI: Probabilistic Intelligence (Latent Variable Models) slides | |
| Week 7 | NewFAI: LLM with RAG slides |
| NewFAI: Fine-tuning slides | |
| Week 8 | Stuart Russell's Reith Lecture Part 3 view |
| Tutorial (pre-midsem discussions) | |
| Mid-semester exam, including Turing's paper as test (March 9, 2026) | |
| Week 9 | GOFAI: Neural Intelligence (also: NewFAI: Intro to Deep Learning) slides |
| NewFAI: Deep Learning (Transformers) notebook | |
| Week 10 | GOFAI: Reward-based Intelligence slides |
| NewFAI: Deep Reinforcement Learning slides | |
| NewFAI: Reinforcement Learning from Human Feedback slides | |
| Week 11 | GOFAI: Learning for Logical Intelligence slides |
| GOFAI: Learning for Probabilistic Intelligence slides | |
| NewFAI: Learning for Neural Intelligence (AR models) slides |
| Lab | Plan |
|---|---|
| Lab 0 | Using LLMs to generate code for solving search problems code |
| Lab 1 | QnA with LLM agents code (Credit: Riya); Sudoku code |
| Lab 2 | Planning with LLM tools code |
| Lab 3 | LLM with RAG code |
| Lab 4 | Homework: Take any open-source LLM and fine-tune for prediction tasks |
| Lab 5 | Implementation of an Enc-Dec transformer from scratch notebook |
| Lab 6 | Understanding word and sentence embeddings, retrieval notebook |
| Lab 7 | Q-learning: Treasure Hunt notebook |
| Lab 8 | Deep RL: TRex-Game notebook |
You can find the course project problem statement here. The project weighs
Textbooks: Some are below that we like. You are free to read other books.
- Russell and Norvig: Artificial Intelligence: A Modern Approach (any edition will work) book
- Huyen: AI Engineering book
- Raschka: Build LLM from Scratch book
Some relevant tutorials we (TD and AS) developed with our TAs back in the early 2020s. They still work, surprisingly, and are still very useful. You can use this link (BITS F464 Lab and Tutorial link) to access for lab and tutorial on Python and tutorials on Probability models and Bayesian Networks, etc.
If you can't get a book, it is okay to just go over the GitHub repositories or any other materials that you can find associated with these books. The AIMA book is readily available everywhere (1st-3rd editions).