MIT CSAIL
6.s953 Embodied Intelligence |
Spring 2024 |
|

Course Overview
We will study research papers, new and old, with the goal of understanding how to build physical robots that exhibit human-level intelligence. Detailed topics will include perception, decision-making, learning with limited data, learning to learn, partial observability, and interaction with other agents.
Course Information
Workload
Students will be expected to:
- Attend all classes and participate in discussion
- Read ~3 papers and submit a short homework assignment about those papers before each class meeting
- Present papers and lead discussion once or twice
- Do a final project based on course material or related topic (max team size: 2 students)
Logistics
- Class meetings: Tuesday, Thursday 1:00 - 2:30 PM in 36-112
- Office hours with instructors:
- Phillip: Mondays, 2-3pm, rm 45-741D
- William: Thursdays, 11am-12pm, rm 45-741P
- Enroll in Piazza to get announcements
- Contact instructors via private post on Piazza
- Upload a PDF to Canvas by 1PM before each class
- Listeners not allowed
Grading Policy
- 55%: final project, including checkpoint presentation and final paper
- 5% proposal
- 5% checkpoint presentation
- 10% final presentation
- 35% final report
- 25%: per-class homework and participation (including attendance)
- Possible grades: Good, Fair, Poor, None.
- Two lowest grades will be dropped
- Late assignments not accepted [except for serious medical/life issues, with support from S3 (for undergrads)
or GradSupport (for grad students)]
- 20%: paper presentations