YoVDO

AlphaTensor by DeepMind Explained

Offered By: Yannic Kilcher via YouTube

Tags

Reinforcement Learning Courses Artificial Intelligence Courses Matrix Multiplication Courses Algorithm Optimization Courses Tensor Decomposition Courses

Course Description

Overview

Explore a comprehensive video lecture on DeepMind's AlphaTensor, a groundbreaking AI system that discovers efficient matrix multiplication algorithms. Delve into the fundamentals of matrix multiplication, its significance in science and engineering, and the surprising discovery that fewer than N^3 operations are needed for NxN matrices. Learn how AlphaTensor utilizes deep reinforcement learning to play TensorGame, uncovering novel matrix multiplication algorithms that optimize performance across various hardware. Gain insights into tensor decompositions, the formulation of algorithm discovery as a game, and the application of AlphaZero techniques. Discover the far-reaching implications of these advancements for computational efficiency, fundamental mathematics, and complexity theory.

Syllabus

- Intro
- Sponsor: Assembly AI link in description
- What even is Matrix Multiplication?
- A very astounding fact
- Trading multiplications for additions
- Matrix Multiplication as a Tensor
- Tensor Decompositions
- A formal way of finding multiplication algorithms
- How to formulate this as a game?
- A brief primer on AlphaZero / MCTS
- The Results
- Optimizing for different hardware
- Expanding fundamental math
- Summary & Final Comments


Taught by

Yannic Kilcher

Related Courses

Computational Neuroscience
University of Washington via Coursera
Reinforcement Learning
Brown University via Udacity
Reinforcement Learning
Indian Institute of Technology Madras via Swayam
FA17: Machine Learning
Georgia Institute of Technology via edX
Introduction to Reinforcement Learning
Higher School of Economics via Coursera