Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Hamiltonian Monte Carlo frames sampling from a probability distribution as a physics problem. By endowing "particles" with momentum and simulating their energy and motion through Hamilton's equations you can efficiently explore a distribution.

Alec Helbling

10,792 subscribers

156,451 views • 7 months ago •via X (Twitter)

Education Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Langevin Monte Carlo allows you to draw samples from a probability distribution using its log gradient ∇ log p(x). By performing a sort of gradient ascent with noise you can navigate around the distribution. Langevin MC is heavily related to modern diffusion models.

Langevin Monte Carlo allows you to draw samples from a probability distribution using its log gradient ∇ log p(x). By performing a sort of gradient ascent with noise you can navigate around the distribution. Langevin MC is heavily related to modern diffusion models.

Alec Helbling

96,842 views • 1 year ago

Random Walks aim to explore a distribution through random perturbations. This is likely to move into low density regions, which is inefficient. Hamiltonian MC sails through a distribution much more rapidly by incorporating the distribution's structure into proposals.

Random Walks aim to explore a distribution through random perturbations. This is likely to move into low density regions, which is inefficient. Hamiltonian MC sails through a distribution much more rapidly by incorporating the distribution's structure into proposals.

Alec Helbling

32,767 views • 1 year ago

One of the simplest algorithms for sampling from a probability distribution is Random Walk Metropolis-Hastings. It proposes new samples by taking Gaussian-distributed steps, accepting or rejecting them to maintain the target distribution. I call this pdf the "fidget spinner".

One of the simplest algorithms for sampling from a probability distribution is Random Walk Metropolis-Hastings. It proposes new samples by taking Gaussian-distributed steps, accepting or rejecting them to maintain the target distribution. I call this pdf the "fidget spinner".

Alec Helbling

79,869 views • 1 year ago

Flow Matching aims to learn a "flow" that transforms a simple source distribution (e.g. Gaussian) to an arbitrarily complex target distribution. This video shows the evolution of the marginal probability path as a source distribution is transformed to a target distribution.

Flow Matching aims to learn a "flow" that transforms a simple source distribution (e.g. Gaussian) to an arbitrarily complex target distribution. This video shows the evolution of the marginal probability path as a source distribution is transformed to a target distribution.

Alec Helbling

150,399 views • 1 year ago

Startups achieve success slowly, building their company brick by brick. Founders can develop marketing leverage by building a viral or network effects strategy or… by partnering with larger companies who have great distribution themselves and trying to reach through to the end customer with premium offerings. Companies that build their distribution by coming in at a lower than market price should follow up with higher margin products or they will find themselves in a competition that we call “a race to the bottom.” It is mission critical for startups to work to build distribution and market share, but they need to then provide something that their customers can’t live without, and charge enough for it to support the continued distribution strategy. Startup founders: Look at every customer as a long-term partner, because you want your business to be a long-term survivor. Look at every supplier as a long-term partner and make sure they are healthy. Look at every employee as a long-term partner. You want them to grow with you and with the business.

Startups achieve success slowly, building their company brick by brick. Founders can develop marketing leverage by building a viral or network effects strategy or… by partnering with larger companies who have great distribution themselves and trying to reach through to the end customer with premium offerings. Companies that build their distribution by coming in at a lower than market price should follow up with higher margin products or they will find themselves in a competition that we call “a race to the bottom.” It is mission critical for startups to work to build distribution and market share, but they need to then provide something that their customers can’t live without, and charge enough for it to support the continued distribution strategy. Startup founders: Look at every customer as a long-term partner, because you want your business to be a long-term survivor. Look at every supplier as a long-term partner and make sure they are healthy. Look at every employee as a long-term partner. You want them to grow with you and with the business.

Tim Draper

10,607 views • 11 months ago

Web3 brands have a distribution problem. For Memeland ❤️ Memecoin and $MEME, distribution is our moat through our existing audience and 9GAG ❤️ Memeland. An extension of this is partnerships with Web2 brands/ products with the Potatoz IP being integrated. Stay tuned. 🥔

Web3 brands have a distribution problem. For Memeland ❤️ Memecoin and $MEME, distribution is our moat through our existing audience and 9GAG ❤️ Memeland. An extension of this is partnerships with Web2 brands/ products with the Potatoz IP being integrated. Stay tuned. 🥔

9GAGCEO

16,596 views • 8 months ago

You can use Mina's new passive to glide and traverse the map more efficiently. It's best used with some momentum, either from a wall bounce, vent, or dash jump. You can glide by aiming down sights.

You can use Mina's new passive to glide and traverse the map more efficiently. It's best used with some momentum, either from a wall bounce, vent, or dash jump. You can glide by aiming down sights.

Deadlock

55,393 views • 8 months ago

The physics of load distribution

The physics of load distribution

Science girl

2,959,962 views • 8 months ago

Visual annotation with Ray3 lets you explore multiple ad variations from a single frame. Sketch how scenes unfold, a can lifted, fruit bursting, or motion synced to energy. Now bringing ad-level direction to life through visual intent. Try Ray3 in Dream Machine.

Visual annotation with Ray3 lets you explore multiple ad variations from a single frame. Sketch how scenes unfold, a can lifted, fruit bursting, or motion synced to energy. Now bringing ad-level direction to life through visual intent. Try Ray3 in Dream Machine.

Luma AI

11,929 views • 8 months ago

A discrete Markov chain is basically a random walk on a graph, where each outgoing edge has a fixed probability. Cool fact: any initial distribution on a sufficiently nice* Markov chain converges to its stationary distribution after many steps—because it's a contraction mapping.

A discrete Markov chain is basically a random walk on a graph, where each outgoing edge has a fixed probability. Cool fact: any initial distribution on a sufficiently nice* Markov chain converges to its stationary distribution after many steps—because it's a contraction mapping.

Keenan Crane

181,771 views • 1 year ago

Perhaps the day will come when the truth about energy capture and distribution can no longer be hidden.

Perhaps the day will come when the truth about energy capture and distribution can no longer be hidden.

Bronze Giant

58,548 views • 1 year ago

Distribution connection diagram 🔥⚡ A simple visual guide to understanding electrical distribution connections and how power is safely routed.

Distribution connection diagram 🔥⚡ A simple visual guide to understanding electrical distribution connections and how power is safely routed.

ENGR FIX

17,826 views • 1 month ago

This animation shows how a uniform grid of points is deformed by a flow matching model. The visualization highlights how the learned flow warps the underlying space—stretching, compressing, and bending it—to transform one probability distribution into another.

This animation shows how a uniform grid of points is deformed by a flow matching model. The visualization highlights how the learned flow warps the underlying space—stretching, compressing, and bending it—to transform one probability distribution into another.

Alec Helbling

42,948 views • 1 year ago

Beautiful 3D grid map to visualize world population density. You can switch between years and see how the distribution changes over time. Explore here (by Raluca Nicola): ralucanicola.github .io/JSAPI_demos/world-population/

Beautiful 3D grid map to visualize world population density. You can switch between years and see how the distribution changes over time. Explore here (by Raluca Nicola): ralucanicola.github .io/JSAPI_demos/world-population/

Leon

27,517 views • 9 months ago

Messing around with classical mechanics. Here a simple 2D system with collision detection and conservation of momentum. #creativecoding #particles

Messing around with classical mechanics. Here a simple 2D system with collision detection and conservation of momentum. #creativecoding #particles

Darien Brito

88,272 views • 1 year ago

A Hamiltonian path is a route that passes through every node in a graph exactly once. Finding Hamiltonian paths can overload even the best-known algorithm for the job. Finding Eulerian paths that pass through every edge is computationally simpler.

A Hamiltonian path is a route that passes through every node in a graph exactly once. Finding Hamiltonian paths can overload even the best-known algorithm for the job. Finding Eulerian paths that pass through every edge is computationally simpler.

Quanta Magazine

69,419 views • 1 year ago

Google is giving away stablecoins, and it's only possible with Self's proof-of-human technology. Real humans can get real money through a first-of-its-kind distribution engine. Here's how it works. ↓

Google is giving away stablecoins, and it's only possible with Self's proof-of-human technology. Real humans can get real money through a first-of-its-kind distribution engine. Here's how it works. ↓

Self

11,348 views • 2 months ago

A source at one of the Gaza distribution sites tells me that Hamas set up a roadblock to prevent Gazans from getting aid. They broke through it and were shouting “thank you America” upon reaching the site.

A source at one of the Gaza distribution sites tells me that Hamas set up a roadblock to prevent Gazans from getting aid. They broke through it and were shouting “thank you America” upon reaching the site.

Kassy Akiva

4,107,403 views • 1 year ago

World population distribution by latitude and longitude - animated

World population distribution by latitude and longitude - animated

Vintage Maps

843,467 views • 2 months ago

This "impossible pit" is designed so you can escape using momentum and physics

This "impossible pit" is designed so you can escape using momentum and physics

Interesting AF

2,488,598 views • 2 months ago