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。

We developed RECAP Physical Intelligence to apply RL and interventions to π0.6, achieving high success rates and throughput on several challenging tasks! Watching these policies operate successfully for hours gives an appreciation for what the method can do

Michael Equi

1,331 subscribers

14,241 次观看 • 9 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.

Physical Intelligence

440,982 次观看 • 5 个月前

Traditional tokenization methods for robotic actions struggle with high-frequency, dexterous tasks due to redundancy and inefficiency. Inspired by JPEG compression, Physical Intelligence has developed a compressed action representation that accelerates VLA model training 5x.

Traditional tokenization methods for robotic actions struggle with high-frequency, dexterous tasks due to redundancy and inefficiency. Inspired by JPEG compression, Physical Intelligence has developed a compressed action representation that accelerates VLA model training 5x.

The Humanoid Hub

26,198 次观看 • 1 年前

Learning dexterous policies from human videos is challenging due to differences between human and robot hands. We present HuDOR, a method that learns dexterous policies within the robot's physical constraints using just one human video and an hour of online interactions! [1/n]

Learning dexterous policies from human videos is challenging due to differences between human and robot hands. We present HuDOR, a method that learns dexterous policies within the robot's physical constraints using just one human video and an hour of online interactions! [1/n]

Irmak Guzey

65,141 次观看 • 1 年前

[1/16] The real world is large, and we want our AI models to operate on the scale of reality. Over the past two years, my colleagues and I have developed fVDB, a deep learning framework for large-scale, high performance spatial intelligence. We finally announced this work at SIGGRAPH and have released the framework in early access. Here’s a small tour of what fVDB is, and what we’ve used it to do.

[1/16] The real world is large, and we want our AI models to operate on the scale of reality. Over the past two years, my colleagues and I have developed fVDB, a deep learning framework for large-scale, high performance spatial intelligence. We finally announced this work at SIGGRAPH and have released the framework in early access. Here’s a small tour of what fVDB is, and what we’ve used it to do.

Francis Williams

34,026 次观看 • 2 年前

Claude Code /simplify is now /code-review Now accepts an effort parameter: - low: short, scoped, latency-sensitive tasks that aren't intelligence-sensitive - medium: cost-sensitive work that can trade off some intelligence for fewer tokens - high: balances tokens and intelligence. Minimum for intelligence-sensitive work, or to cut spend vs xhigh - xhigh: best results for most coding and agentic tasks. Recommended default on Opus 4.7 - max: can improve performance on demanding tasks but shows diminishing returns and tends to overthink. Test before adopting broadly

Claude Code /simplify is now /code-review Now accepts an effort parameter: - low: short, scoped, latency-sensitive tasks that aren't intelligence-sensitive - medium: cost-sensitive work that can trade off some intelligence for fewer tokens - high: balances tokens and intelligence. Minimum for intelligence-sensitive work, or to cut spend vs xhigh - xhigh: best results for most coding and agentic tasks. Recommended default on Opus 4.7 - max: can improve performance on demanding tasks but shows diminishing returns and tends to overthink. Test before adopting broadly

Daniel San

85,822 次观看 • 3 个月前

Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.

Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.

Generalist

3,340,320 次观看 • 9 天前

I’m really hoping there is an appreciation for what we are watching Ohio State and Ryan Day do in real time, and the effects it’s having on opposing coaching staffs…

I’m really hoping there is an appreciation for what we are watching Ohio State and Ryan Day do in real time, and the effects it’s having on opposing coaching staffs…

Kris Drew

40,806 次观看 • 9 个月前

Super impressive, this AI agent can use the web browser like a human. Just describe what you want it to do, and it will automatically operate Chrome for you to achieve your task. Apply for early access to HyperWrite:

Super impressive, this AI agent can use the web browser like a human. Just describe what you want it to do, and it will automatically operate Chrome for you to achieve your task. Apply for early access to HyperWrite:

Lior Alexander

217,712 次观看 • 3 年前

We have news! We created a new robotics model called Loop Model 1. On the zip-tie insertion task, it achieves 20x more throughput per unit of data than "Pi06 + RLT" from Physical Intelligence, a top model for such tasks. It’s the missing piece that makes MicroFactory work, because now deployment becomes so simple and fast that our users can do it themselves.

We have news! We created a new robotics model called Loop Model 1. On the zip-tie insertion task, it achieves 20x more throughput per unit of data than "Pi06 + RLT" from Physical Intelligence, a top model for such tasks. It’s the missing piece that makes MicroFactory work, because now deployment becomes so simple and fast that our users can do it themselves.

Igor Kulakov (MicroFactory)

77,013 次观看 • 3 个月前

National Truck Driver Appreciation Week is such an important time for us at U.S. Department of Transportation to talk to truckers about what we can do to support them. These men and women keep our store shelves stocked and America moving day and night, rain or shine. We thank them for their work! 🚚🇺🇸

National Truck Driver Appreciation Week is such an important time for us at U.S. Department of Transportation to talk to truckers about what we can do to support them. These men and women keep our store shelves stocked and America moving day and night, rain or shine. We thank them for their work! 🚚🇺🇸

Secretary Sean Duffy

131,619 次观看 • 11 个月前

Knights we’ve 10 hours before voting lines closes Have you voted for kola? Do you think the vote is enough or better to apply more pressure? It’s the last hours of voting and we need to apply more pressure and vote for kola!! VOTE FOR KOLA #KolaOmotoso #BBNaija

Knights we’ve 10 hours before voting lines closes Have you voted for kola? Do you think the vote is enough or better to apply more pressure? It’s the last hours of voting and we need to apply more pressure and vote for kola!! VOTE FOR KOLA #KolaOmotoso #BBNaija

Kola Omotoso

42,705 次观看 • 11 个月前

18-18 Hours of Work.....What a Dedication: From showing Green Flags to Buses to This.....Such Important tasks for the Nation!!! Only a True Nationalist and Patriot can do such Jobs...

18-18 Hours of Work.....What a Dedication: From showing Green Flags to Buses to This.....Such Important tasks for the Nation!!! Only a True Nationalist and Patriot can do such Jobs...

𝐒𝐢𝐝𝐝

78,174 次观看 • 10 个月前

RL is a powerful mechanism for training company-specific models on their unique work and data. This is what we do at Applied Compute. A key challenge is how to make RL efficient, because we need runs to be fast (delivered in days), cheap (scalable unit economics), and predictable (not just fast, but reliably fast). Here are some takeaways: • Synchronous RL is wasteful with time and compute. • Asynchronous RL is more efficient but introduces staleness, which causes learning instabilities. • Modeling and simulations can help analytically solve for what configuration leads to optimal efficiency. This allows us to rapidly prototype training configurations, without burning expensive compute cycles on trial runs. Two of our co-founders, Rhythm Garg and Linden Li, discussed some of this research at AI Engineer recently, with a focus on the following subproblem: what is the highest throughput way to do RL given a maximum staleness and compute budget?

RL is a powerful mechanism for training company-specific models on their unique work and data. This is what we do at Applied Compute. A key challenge is how to make RL efficient, because we need runs to be fast (delivered in days), cheap (scalable unit economics), and predictable (not just fast, but reliably fast). Here are some takeaways: • Synchronous RL is wasteful with time and compute. • Asynchronous RL is more efficient but introduces staleness, which causes learning instabilities. • Modeling and simulations can help analytically solve for what configuration leads to optimal efficiency. This allows us to rapidly prototype training configurations, without burning expensive compute cycles on trial runs. Two of our co-founders, Rhythm Garg and Linden Li, discussed some of this research at AI Engineer recently, with a focus on the following subproblem: what is the highest throughput way to do RL given a maximum staleness and compute budget?

Applied Compute

45,652 次观看 • 8 个月前

Today I’m excited to introduce micro1 Intelligence, the world’s most advanced platform for training frontier AI models. Achieving AGI is bottlenecked by one main thing: high-quality data. Data based on real-world environments that capture human expert workflows, complex decision-making, and reward signals models need to learn. With micro1 Intelligence, frontier labs can train on RL environments across every subject matter all in one place.

Today I’m excited to introduce micro1 Intelligence, the world’s most advanced platform for training frontier AI models. Achieving AGI is bottlenecked by one main thing: high-quality data. Data based on real-world environments that capture human expert workflows, complex decision-making, and reward signals models need to learn. With micro1 Intelligence, frontier labs can train on RL environments across every subject matter all in one place.

Ali Ansari

332,850 次观看 • 9 个月前

Historically, RL policies for robots have been trained in synthetic, untextured environments, limiting perceptive policies to depth images where the sim-to-real gap is manageable. RGB has always had more potential, but leveraging it to train policies in simulation remained an open problem. Partnering with Niantic Spatial 🌎 and NVIDIA Robotics, we built a pipeline that addresses exactly that. We can now scan a real deployment site with off-the-shelf hardware, reconstruct it into a photorealistic Gaussian splat, and run massively parallel RL training. The policies trained in our Gym environment then transfer zero-shot to the real robot and environments they were trained for. This enables faster deployment of more capable and robust policies for the end user. The new resulting capabilities are a big step towards solving sim-to-real and also apply well beyond navigation. Read the full technical breakdown on our blog; link in the comments. #HumanoidRobots #Flexion #NianticSpatial #NVIDIA

Historically, RL policies for robots have been trained in synthetic, untextured environments, limiting perceptive policies to depth images where the sim-to-real gap is manageable. RGB has always had more potential, but leveraging it to train policies in simulation remained an open problem. Partnering with Niantic Spatial 🌎 and NVIDIA Robotics, we built a pipeline that addresses exactly that. We can now scan a real deployment site with off-the-shelf hardware, reconstruct it into a photorealistic Gaussian splat, and run massively parallel RL training. The policies trained in our Gym environment then transfer zero-shot to the real robot and environments they were trained for. This enables faster deployment of more capable and robust policies for the end user. The new resulting capabilities are a big step towards solving sim-to-real and also apply well beyond navigation. Read the full technical breakdown on our blog; link in the comments. #HumanoidRobots #Flexion #NianticSpatial #NVIDIA

Flexion

71,971 次观看 • 1 个月前