Towards Understanding Deep Classifiers though Neural Collapse
1:13:23
Language Models as World Models
1:13:54
Towards Understanding the Implicit Regularization Effect of SGD
21:30
Similarity of neural network representations revisited
1:01:40
Panel Discussion: Open Problems in the Theory of Deep Learning
59:47
On the Role of Neural Collapse in Few-Shot Transfer Learning
29:43
The Surgery That Proved There Is No Free Will
1:09:42
The Mystery of Spinors
40:08