Second Boston Symmetry Day: November 3, 2023 at MIT

Schedule:

9:00-9:30. Breakfast + Registration

9:30-9:40. Opening Remarks

9:40-10:20. Talk 1 by Gabriele Corso, Bowen Jing, and Hannes Stark (MIT)
“Equivariant Generative Processes for Molecular Docking”
[Slides]

10:20-10:25. Sponsor presentation for Boston Dynamics AI Institute

10:25-10:50. Coffee

10:50-11:30. Talk 2 by Rui Wang (MIT)
“Relaxed Equivariant Networks for Learning Symmetry Breaking in Physical Systems”
[Slides]

11:30-12:10. Talk 3 by Cengiz Pehlevan (Harvard)
“Geometry and linear separability of neural representations”
[Slides]

12:10-13:40. Lunch

13:40-14:20. Talk 4 by Lawson Wong (Northeastern)
“Symmetry in Decision Making”
[Slides]

14:20-15:00. Talk 5 by Phiala Shanahan (MIT)
“Symmetry equivariant normalising flows for quantum fields”
[Slides]

15:00-16:00. Poster Session + Coffee

16:00-16:40. Talk 6 by Wei Zhu (UMass Amherst)
“Generalization and Optimization in Symmetry-Preserving Machine Learning: Sample Complexity and Implicit Bias”
[Slides]

First Boston Symmetry Day: April 7, 2023 at Northeastern

Schedule:

9:00-9:30. Breakfast, registration.

9:30-9:40. Opening Remarks

9:40-10:10. Talk 1 by Tess Smidt (MIT).
“Expected and Unexpected Properties of Neural Networks with Symmetry”
[Slides]

10:10-10:40. Talk 2 by Elisenda Grigsby (Boston College).
“Functional dimension of ReLU neural networks”
[Slides]

10:40-11:00. Coffee

11:00-11:30. Talk 3 by Mario Geiger (MIT).
“Create Equivariant Polynomials with e3nn”
[Slides]

11:30-12:00. Talk 4 by Nima Dehmamy (IBM Research, MIT-IBM Lab).
“Identifying symmetries in the parameter space and data”
[Slides]

12:00-13:00. Lunch. [Shillman Hall Room 305]

13:00-13:30. Break.

13:30-14:00. Talk 5 by Wengong Jin (Broad Institute).
“SE(3) Denoising Score Matching via Neural Euler’s Rotation Equation — an application to drug discovery”
[Slides]

14:00-14:30. Talk 6 by Steven Gortler (Harvard).
“Invariant Embeddings”

14:30-15:30. Coffee + Poster Session

15:30-16:00. Talk 7 by Shubhendu Trivedi.
“Two displaced vignettes about approximate equivariance”
[Slides] [arXiv]

16:00-16:30. Talk 8 by Robert Platt (Northeastern).
“Applications of symmetric neural models to robots”
[Slides]

Organizers

Stefanie Jegelka

Associate Professor

MIT

Hannah Lawrence

PhD Student

MIT

Derek Lim

PhD Student

MIT

Jung Yeon Park

PhD Student

Northeastern

Robin Walters

Assistant Professor

Northeastern