正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

can a neural network learn to walk as a physical object in a physics simulation? here I train walking neural nets with an evolutionary algorithm. The input nodes/feet are activated by sine waves at learned phases & connections between two neurons extend based on their difference

Matt Henderson

80,126 subscribers

339,220 次观看 • 3 年前 •via X (Twitter)

科学技术教育

Anya Rossi• Live Now

Private livecam show

10 条评论

Matt Henderson 的头像

Matt Henderson3 年前

aka “embodied neural networks”

Matt Henderson 的头像

Matt Henderson3 年前

Remember soda constructor?

Patrick Ꝺoyle 的头像

Patrick Ꝺoyle3 年前

Whatever the result, they seem to be enjoying themselves.

Eric Jang 的头像

Eric Jang3 年前

this is amazing. would you consider open-sourcing the code? I'd love to extend this to optimizing the architecture of the net itself, so it ends up having to trade off physical bulk with function approximation power

Tom 的头像

Tom3 年前

Looks kinda similar

Austen Lamacraft 的头像

Austen Lamacraft3 年前

Surely a CNN will work better because of the strides?

c7ddfc 的头像

c7ddfc3 年前

reminds me of this, also controlled by a neural net

Mike Vella 的头像

Mike Vella3 年前

How do you think of this stuff??

Matt Henderson 的头像

Matt Henderson3 年前

I saw a cool demo of a neural net controlling an inverted pendulum, and was thinking it would be cool to make a nnet control something. I was also thinking about that old game soda constructor. Then I thought of this idea like a bad joke about learning to walk

Connor McCormick 的头像

Connor McCormick3 年前

You should make it so that the ground is a treadmill with numbers on it, and the fitness function rewards it both for walking and for adding the numbers it's touching correctly.

相关视频

A Neural Network Can Grow New Neurons Where It Is Confused? In 1994, Bernd Fritzke published A Growing Neural Gas Network Learns Topologies. He introduced a network that starts small, follows incoming data, and inserts new neurons where its error is highest. In the animation, the fog is the drifting data. The glowing nodes are neurons. The fibers are learned connections. The network grows into a living skeleton of the manifold.

A Neural Network Can Grow New Neurons Where It Is Confused? In 1994, Bernd Fritzke published A Growing Neural Gas Network Learns Topologies. He introduced a network that starts small, follows incoming data, and inserts new neurons where its error is highest. In the animation, the fog is the drifting data. The glowing nodes are neurons. The fibers are learned connections. The network grows into a living skeleton of the manifold.

Mathelirium

38,191 次观看 • 1 个月前

Implemented a visualization for the neural network today. This is a neural network being trained to learn the snake game in a terminal. It's built in Rust using Ratatui More input details in thread

Implemented a visualization for the neural network today. This is a neural network being trained to learn the snake game in a terminal. It's built in Rust using Ratatui More input details in thread

Bones

165,655 次观看 • 2 年前

Neuron cell communication Neurons connect to other neurons and transmit information via neurites. These cultured neurons also extend several neurites to connect with neighboring neurons, forming a neural network.

Neuron cell communication Neurons connect to other neurons and transmit information via neurites. These cultured neurons also extend several neurites to connect with neighboring neurons, forming a neural network.

Adrian Dittmann

3,661,004 次观看 • 1 年前

In the era of 10T-parameter foundation models 😈, can we still build next-generation world representations using 🦥 a single RTX 3090 GPU? We introduce WorldString (a name inspired by string theory), a tiny neural network representation for reconstructing dynamic world instances, ranging from articulated motion and skeletal skinning to soft-body deformation. 🔥 We view object representation as the core primitive of world modeling: the basic executable unit from which perception, interaction, simulation, and generation can be built. With a compact neural basis, WorldString aims to connect dynamic object modeling with neural simulation and foundation models, enabling a scalable path toward richer representations of the physical world.

In the era of 10T-parameter foundation models 😈, can we still build next-generation world representations using 🦥 a single RTX 3090 GPU? We introduce WorldString (a name inspired by string theory), a tiny neural network representation for reconstructing dynamic world instances, ranging from articulated motion and skeletal skinning to soft-body deformation. 🔥 We view object representation as the core primitive of world modeling: the basic executable unit from which perception, interaction, simulation, and generation can be built. With a compact neural basis, WorldString aims to connect dynamic object modeling with neural simulation and foundation models, enabling a scalable path toward richer representations of the physical world.

Xueyan Zou

37,725 次观看 • 13 天前

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Harper Carroll

28,368 次观看 • 7 个月前

Physics-Informed Neural Operators: Learning The Solver, Not Just One Solution Our PINN scene learned one solution field for one PDE setup. A Physics-Informed Neural Operator learns the map from input fields, like material coefficients or source terms, to the full solution across a whole family of PDE problems. So, the goal is no longer just one approximate answer, but a reusable solver-shaped object guided by the physics itself.

Physics-Informed Neural Operators: Learning The Solver, Not Just One Solution Our PINN scene learned one solution field for one PDE setup. A Physics-Informed Neural Operator learns the map from input fields, like material coefficients or source terms, to the full solution across a whole family of PDE problems. So, the goal is no longer just one approximate answer, but a reusable solver-shaped object guided by the physics itself.

Mathelirium

25,935 次观看 • 3 个月前

The Inventor of the Neural Network, Dr. Warren McCulloch explains how Artificial Neurons are able to mimic Brain Neurons algorithmically. I grew up around the peek of Bell Labs in New Jersey and I knew quite a few folks like Dr McCulloch walking around the Campus.

The Inventor of the Neural Network, Dr. Warren McCulloch explains how Artificial Neurons are able to mimic Brain Neurons algorithmically. I grew up around the peek of Bell Labs in New Jersey and I knew quite a few folks like Dr McCulloch walking around the Campus.

Brian Roemmele

39,035 次观看 • 1 年前

The Inventor of the Neural Network, Dr. Warren McCulloch explains how Artificial Neurons are able to mimic Brain Neurons algorithmically:

The Inventor of the Neural Network, Dr. Warren McCulloch explains how Artificial Neurons are able to mimic Brain Neurons algorithmically:

ₕₐₘₚₜₒₙ

152,056 次观看 • 1 年前

🌟Your static 3D world models are now alive and interactable! 🚀Introducing NeuROK, a neural simulation framework that turns any static 3D object into an interactive 4D asset — no per-category physics, no physical annotations for training. 📄 🧵 1/n

🌟Your static 3D world models are now alive and interactable! 🚀Introducing NeuROK, a neural simulation framework that turns any static 3D object into an interactive 4D asset — no per-category physics, no physical annotations for training. 📄 🧵 1/n

Chen Geng

31,637 次观看 • 23 天前

Hopfield recall Spiking Neural network now this one is cool isn't it? -open source, link in comment- This is a Spiking Neural Network (SNN) that can learn and recall patterns through Hebbian learning, similar to a Hopfield network but using biologically-inspired spiking dynamics rather than energy-based settling.

Hopfield recall Spiking Neural network now this one is cool isn't it? -open source, link in comment- This is a Spiking Neural Network (SNN) that can learn and recall patterns through Hebbian learning, similar to a Hopfield network but using biologically-inspired spiking dynamics rather than energy-based settling.

echo.hive

31,452 次观看 • 5 个月前

Scientists have mapped the entire neural network of the fruit fly. 50 million synapses between 139,000 neurons.

Scientists have mapped the entire neural network of the fruit fly. 50 million synapses between 139,000 neurons.

Massimo

1,081,311 次观看 • 1 年前

Individual Neuron: Neural Network A neuron in a neural network performs a weighted sum of inputs, adds a bias, and applies an activation function like sigmoid, ReLU, or tanh, introducing non-linearity. This output helps the neuron learn and represent patterns in the data.

Individual Neuron: Neural Network A neuron in a neural network performs a weighted sum of inputs, adds a bias, and applies an activation function like sigmoid, ReLU, or tanh, introducing non-linearity. This output helps the neuron learn and represent patterns in the data.

ₕₐₘₚₜₒₙ — e/acc

47,467 次观看 • 2 年前

Check out 🌟Vid2Sim: Generalizable, Video-based Reconstruction of Appearance, Geometry & Physics for Mesh-Free Simulation #CVPR2025, from Lingjie Liu’s lab at UPenn. Congrats to Chuhao Chen! Vid2Sim aims to achieve system identification by reconstructing geometry, appearance, and physical properties directly from video. It combines learned data priors with closed-loop optimization: a feed-forward predictor trained on physical prior, followed by fast refinement via Neural Jacobian and mesh-free simulation. The system delivers simulation-ready outputs in minutes, with strong generalization across objects and materials. 🏠Project page: #PhysicalAI #AIGC #CV #CG #simulation #graphics

Check out 🌟Vid2Sim: Generalizable, Video-based Reconstruction of Appearance, Geometry & Physics for Mesh-Free Simulation #CVPR2025, from Lingjie Liu’s lab at UPenn. Congrats to Chuhao Chen! Vid2Sim aims to achieve system identification by reconstructing geometry, appearance, and physical properties directly from video. It combines learned data priors with closed-loop optimization: a feed-forward predictor trained on physical prior, followed by fast refinement via Neural Jacobian and mesh-free simulation. The system delivers simulation-ready outputs in minutes, with strong generalization across objects and materials. 🏠Project page: #PhysicalAI #AIGC #CV #CG #simulation #graphics

Zhiyang (Frank) Dou

12,393 次观看 • 1 年前

you can and should ask claude to train a neural network and then compile into excel formulas

you can and should ask claude to train a neural network and then compile into excel formulas

nate parrott

185,354 次观看 • 9 个月前

I was bored, so I created a self-learning Flappy Bird from scratch using a neural network and a genetic algorithm.

I was bored, so I created a self-learning Flappy Bird from scratch using a neural network and a genetic algorithm.

Tanay🔳

4,495,802 次观看 • 10 个月前

Scientists have developed a neural network-based planning framework that enables robots to navigate complex environments and mazes autonomously. Learn more in Science Robotics:

Scientists have developed a neural network-based planning framework that enables robots to navigate complex environments and mazes autonomously. Learn more in Science Robotics:

Science Magazine

25,783 次观看 • 1 年前

Can you calculate a one-node neural network by hand ✍️

Can you calculate a one-node neural network by hand ✍️

Tom Yeh

10,362 次观看 • 7 个月前

Here's Elon Musk in 2022 predicting that all human code could be replaced by neural nets in Tesla's self-driving software. This approach materialized two years later with FSD v12. “The thing that's happened over time with respect to Tesla Autopilot, is that the neural nets have gradually absorbed more and more software. In the limit, of course, you could simply take the videos as seen by the car and compare those to the steering inputs from the steering wheel and pedals, which are very simple inputs. In principle, you could train with nothing in between, because that's what humans are doing with a biological neural net, you could train based on video, and what trains the video is the moving of the steering wheel and the pedals, with no other software in between. We're not there yet, but it's gradually going in that direction.” Tesla AI Day, September 30, 2022

Here's Elon Musk in 2022 predicting that all human code could be replaced by neural nets in Tesla's self-driving software. This approach materialized two years later with FSD v12. “The thing that's happened over time with respect to Tesla Autopilot, is that the neural nets have gradually absorbed more and more software. In the limit, of course, you could simply take the videos as seen by the car and compare those to the steering inputs from the steering wheel and pedals, which are very simple inputs. In principle, you could train with nothing in between, because that's what humans are doing with a biological neural net, you could train based on video, and what trains the video is the moving of the steering wheel and the pedals, with no other software in between. We're not there yet, but it's gradually going in that direction.” Tesla AI Day, September 30, 2022

ELON CLIPS

16,644 次观看 • 1 年前