Lectures
The slides and resources for each lecture will be updated after the class. The schedule below is tentative and both the order and the content will almost definitely change as the semester goes along.
| Date | Lecture |
|---|---|
| 20-Aug | Class cancelled |
| 22-Aug | Class cancelled |
| 27-Aug | Introduction |
| 29-Aug | Neural networks refresher |
| 3-Sep | Symbolic logic |
| 5-Sep | Logic (continued) |
| 10-Sep | Neuro-symbolic methods: Overview |
| 12-Sep | Neuro-symbolic methods (continued) |
| 17-Sep | Neuro-symbolic methods (continued) |
| 19-Sep | Logic as loss: Overview |
| 24-Sep | Semantic loss |
| 26-Sep | Semantic loss (continued) |
| 1-Oct | Multi-valued logic and t-norm losses |
| 3-Oct | T-norm losses (continued) |
| 8-Oct | Fall break, no class |
| 10-Oct | Fall break, no class |
| 15-Oct | Using knowledge to design models |
| 17-Oct | (continued) |
| 22-Oct | Reasoning about model predictions, structured prediction |
| 24-Oct | (continued) |
| 29-Oct | Graph search for structures |
| 31-Oct | Integer linear programming inference |
| 5-Nov | Training with inference |
| 7-Nov | (continued) |
| 12-Nov | Training with symbols within networks |
| 14-Nov | The straight-through estimator |
| 19-Nov | Gumbel softmax |
| 21-Nov | REINFORCE |
| 26-Nov | Guest lecture: Kyle Richardson: Human preference tuning with LLMs |
| 28-Nov | Thanksgiving, no class |
| 3-Dec | REINFORCE (continued) |
| 5-Dec | Looking back at the semester |
Important dates
| Date | Deadline |
| 10-Sep | Project team information due |
| 19-Sep | Review 1 due |
| 1-Oct | Project proposals due |
| 29-Oct | Project intermediate status report due |
| 31-Oct | Review 2 due |
| 21-Nov | Review 3 due |
| 9-Dec | Project final report due |