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.

This is a vertically integrated end to end deep neural network performing forward pass inference real-time, controlling individual actuator's torque output for bidpedal gait generation in adverse, GPS denied, envs. ok its standard PPO rl trained in mjlab, strapped to a tractor.

Harrison Kinsley

101,538 subscribers

34,118 views • 6 months ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Optimus can now sort objects autonomously 🤖 Its neural network is trained fully end-to-end: video in, controls out. Come join to help develop Optimus (& improve its yoga routine 🧘) →

Optimus can now sort objects autonomously 🤖 Its neural network is trained fully end-to-end: video in, controls out. Come join to help develop Optimus (& improve its yoga routine 🧘) →

Tesla Optimus

14,210,271 views • 2 years ago

🧠Model-Based RL shows promises but has seen limited success in real-world robotics. 🌎Introducing Robotic World Model, a black-box end-to-end neural dynamics model that bridges this gap, where policies are trained purely in imagination. NeurIPS Conference 🎯

🧠Model-Based RL shows promises but has seen limited success in real-world robotics. 🌎Introducing Robotic World Model, a black-box end-to-end neural dynamics model that bridges this gap, where policies are trained purely in imagination. NeurIPS Conference 🎯

Chenhao Li

102,560 views • 7 months ago

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 views • 2 years ago

Here's your Saturday Indian tech white pill. - Aexo Aerospace has flown their UAV, Aer Genesis, 500+ times - lucidml built a neural network to simulate worlds in real time - perceptyne is showing off their vertically integrated semi-humanoid stack at India AI Impact Summit

Here's your Saturday Indian tech white pill. - Aexo Aerospace has flown their UAV, Aer Genesis, 500+ times - lucidml built a neural network to simulate worlds in real time - perceptyne is showing off their vertically integrated semi-humanoid stack at India AI Impact Summit

Runtime

48,322 views • 4 months ago

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 views • 2 years ago

Trained a small neural network in Verse to play Snake #UEFN #CreatedInFortnite #Verse

Trained a small neural network in Verse to play Snake #UEFN #CreatedInFortnite #Verse

Julianxd

12,007 views • 25 days ago

Brett Adcock says the latest autonomous demo of Figure 02 is fully end-to-end and uses a single neural network – camera frames in, actions out. “You cannot code your way out of this problem.”

Brett Adcock says the latest autonomous demo of Figure 02 is fully end-to-end and uses a single neural network – camera frames in, actions out. “You cannot code your way out of this problem.”

The Humanoid Hub

56,268 views • 1 year ago

Want to scale RL with your shiny new GPU? 🚀 In our ICML24 Oral we find that RL algorithms hit a barrier when data is scaled up. Our new algorithm, SAPG, proposes a simple fix. It scales to 25k envs and solves hard tasks where PPO makes no progress. 1/n

Want to scale RL with your shiny new GPU? 🚀 In our ICML24 Oral we find that RL algorithms hit a barrier when data is scaled up. Our new algorithm, SAPG, proposes a simple fix. It scales to 25k envs and solves hard tasks where PPO makes no progress. 1/n

Ananye Agarwal

64,403 views • 1 year ago

I trained a neural network to mimic a forest trail near my apartment (web demo in the post)

I trained a neural network to mimic a forest trail near my apartment (web demo in the post)

Ollin Boer Bohan

118,286 views • 1 year ago

In a world of PPO everything for reinforcement learning, I've been tinkering with SAC for training a quadruped gait. This gait is trained purely on CPU (training on one of the Dell GB10s) on a single environment. Training any particular run is obviously slower than PPO on an RTX Pro 6000 with 8092 envs, if you already know the exact hyperparams/rwd function for your PPO algo... but, if we're honest with ourselves, then we know we usually spend days tuning our PPO algo and fighting it to do what we want. In contrast, SAC has kind of been a breath of fresh air, very amenable to changing the reward function to tune behavior. So far, my first attempts to tune things have consistently just worked immediately rather than 15 different variations of reward hacking only to find previous tuned behaviors got lost in the process. There is also FastSAC, which I've not yet tried, but can speed things up potentially and introduce scale back into the equation. My main painpoint in getting SAC to work for gait was actually getting it to learn to step. It seems as though SAC is not as good as PPO at significant exploration on its own. I ended up starting with a sinusoidal gait (basically just a rule to make legs swing) as training wheels then blended it out through training as phase 1, then began working on smoothing things out after this. I think if we look at end to end dev time rather than any particular run that finally managed to work, SAC may actually be the "faster" algorithm to train. Quadruped gaits are inherently easier than bipedal and maybe there are areas where SAC falls short, but I'll definitely be spending more time with SAC.

In a world of PPO everything for reinforcement learning, I've been tinkering with SAC for training a quadruped gait. This gait is trained purely on CPU (training on one of the Dell GB10s) on a single environment. Training any particular run is obviously slower than PPO on an RTX Pro 6000 with 8092 envs, if you already know the exact hyperparams/rwd function for your PPO algo... but, if we're honest with ourselves, then we know we usually spend days tuning our PPO algo and fighting it to do what we want. In contrast, SAC has kind of been a breath of fresh air, very amenable to changing the reward function to tune behavior. So far, my first attempts to tune things have consistently just worked immediately rather than 15 different variations of reward hacking only to find previous tuned behaviors got lost in the process. There is also FastSAC, which I've not yet tried, but can speed things up potentially and introduce scale back into the equation. My main painpoint in getting SAC to work for gait was actually getting it to learn to step. It seems as though SAC is not as good as PPO at significant exploration on its own. I ended up starting with a sinusoidal gait (basically just a rule to make legs swing) as training wheels then blended it out through training as phase 1, then began working on smoothing things out after this. I think if we look at end to end dev time rather than any particular run that finally managed to work, SAC may actually be the "faster" algorithm to train. Quadruped gaits are inherently easier than bipedal and maybe there are areas where SAC falls short, but I'll definitely be spending more time with SAC.

Harrison Kinsley

26,758 views • 4 months ago

🌎World models can predict, but controlling real robots from imagination sees a long-standing failure due to hallucination. 🧠Introducing Uncertainty-Aware RWM: a black-box, end-to-end neural dynamics model with long-horizon uncertainty propagation. 🎯

🌎World models can predict, but controlling real robots from imagination sees a long-standing failure due to hallucination. 🧠Introducing Uncertainty-Aware RWM: a black-box, end-to-end neural dynamics model with long-horizon uncertainty propagation. 🎯

Chenhao Li

54,751 views • 5 months ago

Tesla is making a hugely ambitious leap with their upcoming Tesla FSD beta v12 which will use an end-to-end neural network approach for full self driving. Spoke with jimmah to discuss this game-changing move. (video released first on X and then on YouTube in a few hours)

Tesla is making a hugely ambitious leap with their upcoming Tesla FSD beta v12 which will use an end-to-end neural network approach for full self driving. Spoke with jimmah to discuss this game-changing move. (video released first on X and then on YouTube in a few hours)

Dave Lee

257,762 views • 2 years ago

AI developed at UMA Single neural network, End to End It operates robustly for hours, with precision, adapting to what it sees.

AI developed at UMA Single neural network, End to End It operates robustly for hours, with precision, adapting to what it sees.

Remi Cadene

35,551 views • 23 hours ago

Long video generation is a systems problem. Introducing LongLive-2.0 from NVIDIA Research: an end-to-end NVFP4 training and inference system for long video generation. Low-precision deployment often relies on post-training quantization, creating a gap between how models are trained and how they run. LongLive-2.0 aligns NVFP4-aware training, distillation, and W4A4 inference, maintaining strong benchmark quality while improving speed and memory efficiency.

Long video generation is a systems problem. Introducing LongLive-2.0 from NVIDIA Research: an end-to-end NVFP4 training and inference system for long video generation. Low-precision deployment often relies on post-training quantization, creating a gap between how models are trained and how they run. LongLive-2.0 aligns NVFP4-aware training, distillation, and W4A4 inference, maintaining strong benchmark quality while improving speed and memory efficiency.

NVIDIA AI

60,445 views • 1 month ago

Figure just released one of the most impressive humanoid demos yet. Figure 02 autonomously unloaded and reloaded a dishwasher, end to end. All powered by a single neural network controlling the entire robot. But here's ONE thing no one noticed behind the scenes:

Figure just released one of the most impressive humanoid demos yet. Figure 02 autonomously unloaded and reloaded a dishwasher, end to end. All powered by a single neural network controlling the entire robot. But here's ONE thing no one noticed behind the scenes:

Elad Inbar

233,889 views • 5 months ago

For the first time in over 150 years, Louisville’s majority-Black West End has its own hospital — thanks to the vision of a Black woman, Corenza Townsend! This is what real investment in Black communities looks like!

For the first time in over 150 years, Louisville’s majority-Black West End has its own hospital — thanks to the vision of a Black woman, Corenza Townsend! This is what real investment in Black communities looks like!

Ben Crump

45,431 views • 1 year ago

San Mateo, California–based Generalist showcases their real-time, end-to-end neural network that controls robots at 100Hz for dexterous manipulation tasks. The model handles fine-motor control, bi-manual coordination, tool use, and generalization across arms and environments.

San Mateo, California–based Generalist showcases their real-time, end-to-end neural network that controls robots at 100Hz for dexterous manipulation tasks. The model handles fine-motor control, bi-manual coordination, tool use, and generalization across arms and environments.

The Humanoid Hub

42,524 views • 1 year ago