Scalable Training of Diffusion Samplers via Fixed-point Score Matching

About this video

Hand-picked by a BrighterTube curator · found via round-3 back catalogue of a channel Lumen picks video by video · 82% confidence

Covers scalable training of diffusion samplers using fixed-point score matching. Technical research talk on optimization and sampling methods.

A 60-minute video from Microsoft Research: “Scalable Training of Diffusion Samplers via Fixed-point Score Matching”.

Science & Math Programming & CS

Video by Microsoft Research, played through YouTube's embedded player.

Quantum Computing for Computer Scientists Microsoft ResearchScience & Math

Advanced2.18M views

The Simple Essence of Automatic Differentiation - Conal Elliott Microsoft ResearchProgramming & CS

Advanced17K views

But what is a neural network? | Deep learning chapter 1 3Blue1BrownScience & Math

Intro25.1M views

Transformers, the tech behind LLMs | Deep Learning Chapter 5 3Blue1BrownProgramming & CS

Intro11.6M views

24:26

The Beauty of Bézier Curves Freya HolmérProgramming & CS

2.23M views

1:13:50

The Continuity of Splines Freya HolmérProgramming & CS

1.89M views

The Most Underrated Algorithm VeritasiumProgramming & CS

10.7M views

Coding Adventure: Simulating Fluids Sebastian LagueProgramming & CS

2.66M views

8:38

Diffie-Hellman Key Exchange Art of the ProblemProgramming & CS

938K views

A Better Default Colormap for Matplotlib | SciPy 2015 | Nathaniel Smith and Stéfan van der Walt EnthoughtProgramming & CS

Advanced177K views

16:31

Public Key Cryptography: RSA Encryption Art of the ProblemProgramming & CS

996K views

Category Theory 1.1: Motivation and Philosophy Bartosz MilewskiProgramming & CS

394K views

Quantum Computing for Computer Scientists

Microsoft Research

Advanced2.18M views

Andrew Helwer's Microsoft Research talk teaching quantum computing to programmers with only linear algebra: qubits, gates, and the Deutsch oracle problem.

The Simple Essence of Automatic Differentiation - Conal Elliott

Microsoft Research

Advanced17K views

Conal Elliott presents a simple, general formulation of automatic differentiation derived from category theory.

But what is a neural network? | Deep learning chapter 1

3Blue1Brown

Intro25.1M views

Builds up what a neural network actually is — neurons as numbers, layers, weights and biases — using handwritten-digit recognition as the running example.

Transformers, the tech behind LLMs | Deep Learning Chapter 5

3Blue1Brown

Intro11.6M views

How transformer models (the 'T' in GPT) turn text into vectors and predict the next token, step by step.

24:26

The Beauty of Bézier Curves

Freya Holmér

2.23M views

Freya Holmér builds Bézier curves from linear interpolation up, covering de Casteljau's algorithm, derivatives, arc length and their use in games and graphics.

1:13:50

The Continuity of Splines

Freya Holmér

1.89M views

A follow-up on splines: the different kinds of continuity (C and G), and how Hermite, Catmull-Rom, B-splines and NURBS trade them off.

The Most Underrated Algorithm

Veritasium

10.7M views

The history and inner workings of the Fast Fourier Transform, from nuclear test detection to everyday signal processing.

Coding Adventure: Simulating Fluids

Sebastian Lague

2.66M views

Building a fluid simulation step by step with smoothed-particle hydrodynamics.

8:38

Diffie-Hellman Key Exchange

Art of the Problem

938K views

Art of the Problem explains Diffie–Hellman key exchange with the paint-mixing analogy and then the modular-arithmetic version.

A Better Default Colormap for Matplotlib | SciPy 2015 | Nathaniel Smith and Stéfan van der Walt

Enthought

Advanced177K views

Nathaniel Smith and Stéfan van der Walt explain colour perception and perceptual uniformity, and how they designed viridis, Matplotlib's new default colormap.

16:31

Public Key Cryptography: RSA Encryption

Art of the Problem

996K views

How RSA public-key encryption works, from one-way functions through Euler's totient to the full algorithm.

Category Theory 1.1: Motivation and Philosophy

Bartosz Milewski

394K views

The first lecture of Bartosz Milewski's Category Theory for Programmers: why abstraction and composition matter, and what category theory offers programmers.