Главная
Study mode:
on
1
Intro
2
What is a generative model
3
Why generative models
4
Storyline
5
Punch Line
6
Gradient Descent
7
CSGM Framework
8
Plot
9
Theory
10
Proof
11
Generalization
12
NP Heart
13
Optimization
14
Approximate Metric Entropy
15
Intermediate Layer Optimization
16
Hyperparameters
17
Results
Description:
Explore the groundbreaking concept of generative models as the new frontier of sparsity in this 47-minute lecture by Alex Dimakis from UT Austin. Delve into the fundamentals of generative models, their significance, and their applications in various fields. Follow the storyline as Dimakis presents the CSGM Framework, theoretical proofs, and practical implementations. Examine the role of gradient descent, optimization techniques, and hyperparameter tuning in generative modeling. Gain insights into approximate metric entropy, intermediate layer optimization, and the challenges of NP-hard problems. Discover how generative models are revolutionizing data analysis and machine learning through this comprehensive talk presented at the Joint IFML/CCSI Symposium hosted by the Simons Institute.

Yes, Generative Models Are The New Sparsity

Simons Institute
Add to list
0:00 / 0:00