Fall 2026
  • Piazza
  • Slack
  • Syllabus
  • Labs
  • Project

CSCI 382 Responsible AI/ML

The theory, science and practice of designing and deploying socio-technically responsible machine learning systems.


Instructor: Lucas Rosenblatt. Please call me Lucas.

Class Times: Mondays and Thursdays, 1:10-2:25pm in [ROOM].

Office Hours: In TCL 308. Open drop-in on Wednesdays, 1:30-2:30pm. Appointment blocks Mondays 4:00-5:00pm and Thursdays 2:30-3:30pm; see the syllabus for how to book.

Communication: Piazza for content questions; Slack for casual discussion and team coordination. Please try to use these channels before emailing me; see the syllabus for which to use when.

Labs: Mondays, 2:30-4:00pm in [LAB_ROOM]. Alternating problem weeks and presentation weeks. See the Labs page.

Quizzes: Short quizzes on Mondays covering the prior week’s lectures or a recent homework. 10 total, 1 point each and additive, and you start with 1 point banked, so you can miss one and still earn full credit.

Exams: Midterm in class on Thu Oct 15. We will vote to decide whether the final is in the December exam period or in the last lecture block on Thu Dec 10.

Project: Semester-long project in groups of two. See the Project page.

Some Potentially Useful Resources: Fairness and Machine Learning | CS 860: Algorithms for Private Data Analysis | Review of Probability Theory

Week Monday Thursday Lab Project Assignments
Module 1: AI as Normal Technology
Week 1 (Sep 10) No class The “State of Play”
Week 2 (Sep 14 and Sep 17) Normal Technology, Part 1 Normal Technology, Part 2 Lab 1 (A)
Module 2: Fairness and Bias in ML
Week 3 (Sep 21 and Sep 24) Fairness and Bias in ML Fairness Criteria and Impossibility Lab 2 (B) Quiz 1 | HW1 out
Week 4 (Sep 28 and Oct 1) Thresholds and Post-Processing Bias Across the ML Pipeline Lab 3 (A) Partner matching due (Thu, 11:59pm) Quiz 2
Module 3: Data Privacy
Week 5 (Oct 5 and Oct 8) Data Privacy Defining Differential Privacy Lab 4 (B) Quiz 3 | HW2 out | HW1 due Oct 9, 5pm
Week 6 (Oct 12 and Oct 15) Fall Reading Period (No Class) Midterm Exam No lab
Week 7 (Oct 19 and Oct 22) The Laplace Mechanism Composition and What DP Promises No lab Quiz 4
Week 8 (Oct 26 and Oct 29) DP Synthetic Data Approximate DP and Gaussian Noise Lab 5 (A) Ideation meeting (book a slot) Quiz 5
Module 4: Interpretability and Explainability
Week 9 (Nov 2 and Nov 5) Local DP in Practice Interpretability and Explainability Lab 6 (B) Proposal due (Mon, 1:10pm) Quiz 6 | HW3 out | HW2 due Nov 6, 5pm
Week 10 (Nov 9 and Nov 12) Shapley Values SHAP at Scale Lab 7 (A) Quiz 7
Week 11 (Nov 16 and Nov 19) Auditing Ad Delivery Transparency and the Law Lab 8 (B) Initial results due (Mon, 1:10pm) | Results meeting (book a slot) Quiz 8
Week 12 (Nov 23 and Nov 26) Probing, Patching, and Circuits Thanksgiving Recess (No Class) Lab 9 (A) Quiz 9
Module 5: LLM Risks and the State of AI Research
Week 13 (Nov 30 and Dec 3) R/AI and LLMs (Part 1) R/AI and LLMs (Part 2) Lab 10 (B) Quiz 10 | HW3 due Dec 4, 5pm
Week 14 (Dec 7 and Dec 10) Project Presentations Final Exam (if class votes for this slot) No lab Presentations (Mon, in class) | Final paper due (Fri Dec 11, 11:59pm)

In the Lab column, (A) marks a problem week and (B) a presentation week. Three Mondays have no lab: Oct 12 (reading period), Oct 19 (midterm solutions), and Dec 7 (project presentations run long). See the Labs page.