Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Google DeepMind, David Silver reveals: we built a system that used RL to discover its own RL algorithms. this AI-designed system outperformed all human-created RL algorithms developed over the years.

haider1's profile picture

Haider.

68,908 subscribers

397,236 Aufrufe • vor 1 Jahr •via X (Twitter)

Nachrichten & PolitikWissenschaft & TechnologieBildung
LIVE

Anya Rossi• Live Now

Private livecam show

10 Kommentare

Profilbild von Prinz Eugen, der edle Ritter
Prinz Eugen, der edle Rittervor 1 Jahr

"It's work that we did a few years ago, but is coming out now." Interesting.

Profilbild von Alger
Algervor 1 Jahr

Capitalize on the rapid expansion of AI technologies by investing in companies building and implementing AI solutions.

Profilbild von Noah Vandal
Noah Vandalvor 1 Jahr

this is just the start

Profilbild von Alex from OmniraAI
Alex from OmniraAIvor 1 Jahr

We're officially in the phase where AI is better at building AI than we are. Things are gonna move fast.

Profilbild von kay
kayvor 1 Jahr

Deepmind isn't respected enough. Absolute top tier talent in that company.

Profilbild von Tom Bennet
Tom Bennetvor 1 Jahr

So, AI is now designing better versions of itself. Wonder how long until it starts demanding a stake in the company? 🤔💸

Profilbild von Vedu
Veduvor 1 Jahr

Okay, now that feels scary… and weirdly exciting at the same time.

Profilbild von Deniz
Denizvor 1 Jahr

With genetic mutation, people back then trial and errored the shit out of plants for the „golden“ mutation. This analogy fits perfectly with the technological mutation. The perfect mutation for the golden new age.

Profilbild von Subba Reddy
Subba Reddyvor 1 Jahr

we are done 🤣 >RL to discover its own RL algorithms. >this AI-designed system outperformed all human-created RL algorithms developed over the years.

Profilbild von broadfield-dev
broadfield-devvor 1 Jahr

They made Thronglets.

Ähnliche Videos

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.
2:15

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 Aufrufe • vor 5 Monaten

Richard Sutton argues that AI must move beyond human-generated static data into the “Era of Experience,” where agents learn through continuous interaction with the world. This will require building upon RL with better algorithms capable of continual learning and meta-learning.
9:06

Richard Sutton argues that AI must move beyond human-generated static data into the “Era of Experience,” where agents learn through continuous interaction with the world. This will require building upon RL with better algorithms capable of continual learning and meta-learning.

The Humanoid Hub

31,493 Aufrufe • vor 1 Jahr

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
1:24

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

64,546 Aufrufe • vor 2 Jahren

Excited to announce AlphaEvolve A powerful AI coding agent developed by our team in Google DeepMind that is able to discover impactful new algorithms for important problems in Maths and Computing by combining the creativity of large language models with automated evaluators.
0:32

Excited to announce AlphaEvolve A powerful AI coding agent developed by our team in Google DeepMind that is able to discover impactful new algorithms for important problems in Maths and Computing by combining the creativity of large language models with automated evaluators.

Pushmeet Kohli

203,569 Aufrufe • vor 1 Jahr

Hiring RL Engineer! Started off as a curious project at Lossfunk to push the boundaries of LLMs in social reasoning - we are now building RL environments, data, and benchmarks to simulate more real-world scenarios. If you want to train SoTA RL models over multi-GPUs (H200s/B200s) to unlock next AI frontier, this is for you.
1:21

Hiring RL Engineer! Started off as a curious project at Lossfunk to push the boundaries of LLMs in social reasoning - we are now building RL environments, data, and benchmarks to simulate more real-world scenarios. If you want to train SoTA RL models over multi-GPUs (H200s/B200s) to unlock next AI frontier, this is for you.

Satpal Singh Rathore

46,010 Aufrufe • vor 1 Jahr

We developed a general recipe that allows VLAs to improve from experience. RL is back. (yes, this is 13 hours of coffee making)
9:42

We developed a general recipe that allows VLAs to improve from experience. RL is back. (yes, this is 13 hours of coffee making)

Karol Hausman

111,355 Aufrufe • vor 9 Monaten

I've been working on deformable object manipulation since my PhD. It was totally a nightmare years ago and my PhD advisor was telling me not to work on it for my own good. Today, at ByteDance Seed, we are dropping GR-RL, a new VLA+RL system that manages long-horizon precise dexterous manipulation of deformable objects. This is probably the first real-world RL system to make a robot: ✅ Lace up your shoes end to end ✅ Hit millimeter tolerance repeatedly ✅ Recover from mistakes (See video!) ✅ And complete continuous shoelace threading on a real bimanual platform 📈 Success rate: ↑ from 45.7% → 83.3% Yes, robots can now actually do this. Project page: ArXiv:
3:45

I've been working on deformable object manipulation since my PhD. It was totally a nightmare years ago and my PhD advisor was telling me not to work on it for my own good. Today, at ByteDance Seed, we are dropping GR-RL, a new VLA+RL system that manages long-horizon precise dexterous manipulation of deformable objects. This is probably the first real-world RL system to make a robot: ✅ Lace up your shoes end to end ✅ Hit millimeter tolerance repeatedly ✅ Recover from mistakes (See video!) ✅ And complete continuous shoelace threading on a real bimanual platform 📈 Success rate: ↑ from 45.7% → 83.3% Yes, robots can now actually do this. Project page: ArXiv:

Xiao Ma

110,461 Aufrufe • vor 9 Monaten

We used RL to train models that create curated context from long documents for downstream use by agents. The models sometimes learn to invent their own abbreviations and shorthand. Optimizing with RL for downstream use produces very different artifacts from ordinary summaries: shorter, denser, creatively concise. We call these neural cheat-sheets.
0:24

We used RL to train models that create curated context from long documents for downstream use by agents. The models sometimes learn to invent their own abbreviations and shorthand. Optimizing with RL for downstream use produces very different artifacts from ordinary summaries: shorter, denser, creatively concise. We call these neural cheat-sheets.

Applied Compute

37,007 Aufrufe • vor 2 Monaten

The network for machine intelligence Two years ago, we laid out our vision for a machine learning compute protocol. One that connects every device in the world into an open network for machine intelligence, with no gatekeepers or artificial boundaries. This week, we’ll be sharing some of our early progress, beginning with RL Swarm, a peer-to-peer system for collaborative reinforcement learning over the internet. Next month, we’ll open our Testnet, allowing anyone to contribute to the frontier of open machine intelligence. Introducing RL Swarm RL Swarm is a fully open source system for collaborative reinforcement learning over the internet. It is a live demo of our research findings, which show that models training with RL learn faster when they train as a collective swarm than they do on their own. Join our swarm now to see this in practice. You can participate with consumer hardware at home or a powerful GPU in the cloud. You can follow along with the swarm’s progress by following the links below.
0:28

The network for machine intelligence Two years ago, we laid out our vision for a machine learning compute protocol. One that connects every device in the world into an open network for machine intelligence, with no gatekeepers or artificial boundaries. This week, we’ll be sharing some of our early progress, beginning with RL Swarm, a peer-to-peer system for collaborative reinforcement learning over the internet. Next month, we’ll open our Testnet, allowing anyone to contribute to the frontier of open machine intelligence. Introducing RL Swarm RL Swarm is a fully open source system for collaborative reinforcement learning over the internet. It is a live demo of our research findings, which show that models training with RL learn faster when they train as a collective swarm than they do on their own. Join our swarm now to see this in practice. You can participate with consumer hardware at home or a powerful GPU in the cloud. You can follow along with the swarm’s progress by following the links below.

gensyn

228,915 Aufrufe • vor 1 Jahr

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more
0:41

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more

Prime Intellect

1,149,374 Aufrufe • vor 9 Monaten

Humanoid robots playing table tennis fully autonomously. The 'HITTER' system combines a model-based planner with a reinforcement learning (RL) whole-body controller. It is fully autonomous but relies on an external sensing system. A 9-camera OptiTrack motion capture setup tracks the ball, which is covered with reflective markers, to predict its precise trajectory and compute the optimal strike position, velocity, and timing. The RL controller then executes coordinated arm, leg, and waist movements, producing agile, human-like returns.
2:37

Humanoid robots playing table tennis fully autonomously. The 'HITTER' system combines a model-based planner with a reinforcement learning (RL) whole-body controller. It is fully autonomous but relies on an external sensing system. A 9-camera OptiTrack motion capture setup tracks the ball, which is covered with reflective markers, to predict its precise trajectory and compute the optimal strike position, velocity, and timing. The RL controller then executes coordinated arm, leg, and waist movements, producing agile, human-like returns.

The Humanoid Hub

91,896 Aufrufe • vor 1 Jahr

With AI, "we did 1,000,000,000 years of PHD time in one year." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on AlphaFold, the groundbreaking AI system developed to solve one of biology’s most complex problems: protein folding.
0:36

With AI, "we did 1,000,000,000 years of PHD time in one year." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on AlphaFold, the groundbreaking AI system developed to solve one of biology’s most complex problems: protein folding.

Reid Hoffman

113,494 Aufrufe • vor 1 Jahr

Lecture 10 of my course! Nominally on regularization in RL, so I discuss the evolving role of the KL penalty in RL, but also a set of nice RL papers that explain what RL helps models generalize better than SFT -- with theory supporting it. When going through these, it's so interesting how seasonal problems in ML are. Lots of problems from controlling reward models overopt will rhyme as we try to control rubrics for agents. 00:00 Intro & the role of regularization 02:50 The KL penalty in RL 10:53 RL as a reverse KL loss 21:09 Why RL generalizes better than SFT 25:15 Other regularization tools Just a few videos left as I get to the end of the course. Thanks all, and keep sending questions. Spread the word if you have a second.
31:33

Lecture 10 of my course! Nominally on regularization in RL, so I discuss the evolving role of the KL penalty in RL, but also a set of nice RL papers that explain what RL helps models generalize better than SFT -- with theory supporting it. When going through these, it's so interesting how seasonal problems in ML are. Lots of problems from controlling reward models overopt will rhyme as we try to control rubrics for agents. 00:00 Intro & the role of regularization 02:50 The KL penalty in RL 10:53 RL as a reverse KL loss 21:09 Why RL generalizes better than SFT 25:15 Other regularization tools Just a few videos left as I get to the end of the course. Thanks all, and keep sending questions. Spread the word if you have a second.

Nathan Lambert

89,404 Aufrufe • vor 1 Monat

In 2019, Edward Snowden warned that the infrastructure for a Chinese-style social credit system was being built in plain sight. "All of these things are increasingly being created and programmed and decided by algorithms. And those algorithms are fuelled by precisely the innocent data that our devices are creating all of the time constantly, invisibly, quietly, right now."
2:29

In 2019, Edward Snowden warned that the infrastructure for a Chinese-style social credit system was being built in plain sight. "All of these things are increasingly being created and programmed and decided by algorithms. And those algorithms are fuelled by precisely the innocent data that our devices are creating all of the time constantly, invisibly, quietly, right now."

redpillbot

45,225 Aufrufe • vor 3 Monaten

In 2019, Edward Snowden warned that the infrastructure for a Chinese-style social credit system was being built in plain sight. "All of these things are increasingly being created and programmed and decided by algorithms. And those algorithms are fuelled by precisely the innocent data that our devices are creating all of the time—constantly, invisibly, quietly—right now."
2:29

In 2019, Edward Snowden warned that the infrastructure for a Chinese-style social credit system was being built in plain sight. "All of these things are increasingly being created and programmed and decided by algorithms. And those algorithms are fuelled by precisely the innocent data that our devices are creating all of the time—constantly, invisibly, quietly—right now."

illuminatibot

47,949 Aufrufe • vor 6 Monaten

Over the years we’ve heard some memorable quotes in Rugby League. In 2013 Geoff Toovey produced this all time classic. Anymore RL quotes spring to mind ? #15-1Countdown.
0:53

Over the years we’ve heard some memorable quotes in Rugby League. In 2013 Geoff Toovey produced this all time classic. Anymore RL quotes spring to mind ? #15-1Countdown.

The Oracle

30,823 Aufrufe • vor 1 Jahr

Whatever happened to training physics based character controllers using RL? This was all the rage a few years ago. Every clip here looks cool af.
1:54

Whatever happened to training physics based character controllers using RL? This was all the rage a few years ago. Every clip here looks cool af.

AVB

28,378 Aufrufe • vor 23 Tagen