Physics of Language Models: Part 2.2, How to Learn From Mistakes on Grade-School Math Problems
49:04
Physics of Language Models: Part 2.2, How to Learn From Mistakes on Grade-School Math Problems
Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process
1:00:53
Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process
ICML 2024 Tutorial: Physics of Language Models
1:53:43
ICML 2024 Tutorial: Physics of Language Models
Physics of Language Models: Part 1, Context-Free Grammar
1:39:15
Physics of Language Models: Part 1, Context-Free Grammar
Physics of Language Models: Part 3.1 + 3.2, Knowledge Storage, Extraction and Manipulation
1:18:49
Physics of Language Models: Part 3.1 + 3.2, Knowledge Storage, Extraction and Manipulation
Why Does Deep Learning Perform Deep Learning - MSR AI Seminar 08/11/2020
1:08:41
Why Does Deep Learning Perform Deep Learning - MSR AI Seminar 08/11/2020
Backward Feature Correction: How Deep Learning Performs Deep Learning (May 2020 by Yuanzhi Li)
1:11:59
Backward Feature Correction: How Deep Learning Performs Deep Learning (May 2020 by Yuanzhi Li)
Katyusha X: Practical Momentum Method for Stochastic Sum-of-Nonconvex Optimization
10:05
Katyusha X: Practical Momentum Method for Stochastic Sum-of-Nonconvex Optimization
Optimal Experimental Design via A New Regret Minimization Framework
47:08
Optimal Experimental Design via A New Regret Minimization Framework
How to Swing By Saddle Points: Faster Non-Convex Optimization Than SGD
52:31
How to Swing By Saddle Points: Faster Non-Convex Optimization Than SGD
ICML 2017 Tutorial: Recent Advances in Stochastic Convex and Non-Convex Optimization (audio fixed)
2:06:28
ICML 2017 Tutorial: Recent Advances in Stochastic Convex and Non-Convex Optimization (audio fixed)
First Efficient Convergence for Streaming k-PCA: a Global, Gap-Free, and Near-Optimal Rate
33:20
First Efficient Convergence for Streaming k-PCA: a Global, Gap-Free, and Near-Optimal Rate
ICML 2017 Tutorial: Recent Advances in Stochastic Convex and Non-Convex Optimization
2:06:28
ICML 2017 Tutorial: Recent Advances in Stochastic Convex and Non-Convex Optimization
Natasha: Faster Non-Convex Stochastic Optimization via Strongly Non-Convex Parameter
15:10
Natasha: Faster Non-Convex Stochastic Optimization via Strongly Non-Convex Parameter
Follow the Compressed Leader: Faster Online Learning of Eigenvectors and Faster MMWU
18:28
Follow the Compressed Leader: Faster Online Learning of Eigenvectors and Faster MMWU
Optimal Black-Box Reductions Between Optimization Objectives
5:10
Optimal Black-Box Reductions Between Optimization Objectives
LazySVD: Even Faster SVD Decomposition Yet Without Agonizing Pain
8:11
LazySVD: Even Faster SVD Decomposition Yet Without Agonizing Pain
Three ICML 2016 Talks on Optimization
43:47
Three ICML 2016 Talks on Optimization
Nearly-Linear Time Positive LP Solver with Faster Convergence Rate (STOC 2015)
19:06
Nearly-Linear Time Positive LP Solver with Faster Convergence Rate (STOC 2015)
Using Optimization to Solve Positive LPs Faster in Parallel
26:43
Using Optimization to Solve Positive LPs Faster in Parallel
Linear Coupling of Gradient and Mirror Descent
23:44
Linear Coupling of Gradient and Mirror Descent
Knightian Self Uncertainty in the VCG Mechanism for Unrestricted Combinatorial Auctions
17:26
Knightian Self Uncertainty in the VCG Mechanism for Unrestricted Combinatorial Auctions