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。

Reinforcement learning is mind-blowing! Here's a reinforcement learning agent that mastered Breakout and is playing at a superhuman level (score 350+). The crazy part: We didn't write the code or the pipeline. NEO, the first autonomous AI Engineer, did it. From a prompt! It trained over 24 million steps.... show more

Santiago

431,143 subscribers

21,701 次观看 • 7 个月前 •via X (Twitter)

游戏教育科学技术

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

The era of manual ML is over. Introducing NEO Neo AI the first Agentic ML Engineer. A system of 11 AI agents that designs, codes, and ships ML models automatically. Join the waitlist: Here’s how it works 👇 (VIDEO)

The era of manual ML is over. Introducing NEO Neo AI the first Agentic ML Engineer. A system of 11 AI agents that designs, codes, and ships ML models automatically. Join the waitlist: Here’s how it works 👇 (VIDEO)

SARAH

309,835 次观看 • 11 个月前

Jensen Huang says agentic AI is not a chatbot. It is a "digital robot" -- and the next bottleneck is not the model. It is first-person data. "Agentic AI is just a digital robot. It understands, it reasons, it plans, and it acts and uses tools." Then he makes the bigger jump: "The world's video data is from a third person, not first person. And so agentic systems, robotic systems, physical AI -- the data is the hardest problem." That is the hidden constraint in robotics. LLMs got the internet. Physical AI needs demonstrations, synthetic worlds, reinforcement learning, and first-person experience. This is why Jensen keeps calling it a stack: robots are not waiting for one magical model. They need the data engine, the simulator, the reward system, the runtime, and the computer. The next wave of AI is not just software that talks. It is software that sees, plans, acts, and eventually moves through the world.

Jensen Huang says agentic AI is not a chatbot. It is a "digital robot" -- and the next bottleneck is not the model. It is first-person data. "Agentic AI is just a digital robot. It understands, it reasons, it plans, and it acts and uses tools." Then he makes the bigger jump: "The world's video data is from a third person, not first person. And so agentic systems, robotic systems, physical AI -- the data is the hardest problem." That is the hidden constraint in robotics. LLMs got the internet. Physical AI needs demonstrations, synthetic worlds, reinforcement learning, and first-person experience. This is why Jensen keeps calling it a stack: robots are not waiting for one magical model. They need the data engine, the simulator, the reward system, the runtime, and the computer. The next wave of AI is not just software that talks. It is software that sees, plans, acts, and eventually moves through the world.

Karl Mehta

14,140 次观看 • 18 天前

Introducing NEO: The first Autonomous Machine Learning Engineer. NEO is a multi-agent system that automates the entire ML workflow, saving engineers thousands of hours of grunt work. Our recent advancements in multi-step reasoning and memory orchestration have enabled NEO to excel at solving complex Machine Learning problems on its own from data engineering to deployments of ML models. Benchmarks: NEO was tested against 50 Kaggle competitions and scored a medal in 26% of them, far superior than the previous state of the art performance of 16.9% by Open AI’s O1 with AIDE scaffolding on the MLE bench With these capabilities, NEO aims to help every ML engineer become superhuman and focus on real innovation instead of the existing grunt work Check out NEO in the thread below:

Introducing NEO: The first Autonomous Machine Learning Engineer. NEO is a multi-agent system that automates the entire ML workflow, saving engineers thousands of hours of grunt work. Our recent advancements in multi-step reasoning and memory orchestration have enabled NEO to excel at solving complex Machine Learning problems on its own from data engineering to deployments of ML models. Benchmarks: NEO was tested against 50 Kaggle competitions and scored a medal in 26% of them, far superior than the previous state of the art performance of 16.9% by Open AI’s O1 with AIDE scaffolding on the MLE bench With these capabilities, NEO aims to help every ML engineer become superhuman and focus on real innovation instead of the existing grunt work Check out NEO in the thread below:

Neo AI

321,145 次观看 • 1 年前

Its nice to see the recent excitement around Q-learning! Here's a clip from a video all about Deep Reinforcement Learning, including Deep Q Learning #machinelearning #ai

Its nice to see the recent excitement around Q-learning! Here's a clip from a video all about Deep Reinforcement Learning, including Deep Q Learning #machinelearning #ai

Steven Brunton

164,997 次观看 • 2 年前

Every ML team is about to 10x overnight. The first agentic Machine Learning Engineer is here, and it will blow your mind! Copilots help you write code, but that's about it. People still have to do most of the work. That's about to change!

Every ML team is about to 10x overnight. The first agentic Machine Learning Engineer is here, and it will blow your mind! Copilots help you write code, but that's about it. People still have to do most of the work. That's about to change!

Santiago

76,683 次观看 • 11 个月前

Introducing NEO: The first Agentic Machine Learning Engineer. It works like a full-stack ML engineer that never sleeps: handling data exploration, feature engineering, training, tuning, deployment, and monitoring, end to end. Powered by 11 specialized agents, NEO runs autonomously, saving ML engineers thousands of hours and making them 10x faster. Benchmarks: Tested on 75 Kaggle competitions, NEO scored a medal in 34% of them, significantly outperforming Microsoft’s RD Agent (22.4%) on the MLE Bench. This sets a new state of the art for autonomous ML systems. NEO runs on a novel multi-agent orchestrator, powered by a multi-step reasoning engine, context transfer protocol, and agent memory, built to solve complex workflows end to end. And with human-in-the-loop mode, you can guide, inspect, and override any step. You're always in command. NEO is built for real world workflows and ready for production. NEO is here to make every ML engineer truly superhuman. Watch NEO in action:

Introducing NEO: The first Agentic Machine Learning Engineer. It works like a full-stack ML engineer that never sleeps: handling data exploration, feature engineering, training, tuning, deployment, and monitoring, end to end. Powered by 11 specialized agents, NEO runs autonomously, saving ML engineers thousands of hours and making them 10x faster. Benchmarks: Tested on 75 Kaggle competitions, NEO scored a medal in 34% of them, significantly outperforming Microsoft’s RD Agent (22.4%) on the MLE Bench. This sets a new state of the art for autonomous ML systems. NEO runs on a novel multi-agent orchestrator, powered by a multi-step reasoning engine, context transfer protocol, and agent memory, built to solve complex workflows end to end. And with human-in-the-loop mode, you can guide, inspect, and override any step. You're always in command. NEO is built for real world workflows and ready for production. NEO is here to make every ML engineer truly superhuman. Watch NEO in action:

Neo AI

19,571,775 次观看 • 11 个月前

Can reinforcement learning from AI feedback unlock new capabilities in AI agents? Introducing Motif, an LLM-powered method for intrinsic motivation from AI feedback. Motif extracts reward functions from Llama 2's preferences and uses them to train agents with reinforcement learning. On the complex NetHack game, Motif solves previously unsolved tasks without needing any expert demonstrations. Surprisingly, Motif's reward leads to better game score than the one obtained by using the score itself as a reward. Given access to an event captioning mechanism, a few properties make Motif a general method: • it is entirely based on open models • the LLM doesn't need direct access to the environment dynamics (e.g., its source code) • the LLM doesn't need to understand observation and action spaces The best part? You can start using Motif right now, even on a small compute budget: the whole pipeline can take less than two GPU-days. Feel free to read our paper and try our code out. Paper: Code: Blog post: Work co-lead by Martin Klissarov and myself, with Shagun Sodhani Roberta Raileanu Pierre-Luc Bacon Pascal Vincent Amy Zhang Mikael Henaff Learn more in the thread 🧵

Can reinforcement learning from AI feedback unlock new capabilities in AI agents? Introducing Motif, an LLM-powered method for intrinsic motivation from AI feedback. Motif extracts reward functions from Llama 2's preferences and uses them to train agents with reinforcement learning. On the complex NetHack game, Motif solves previously unsolved tasks without needing any expert demonstrations. Surprisingly, Motif's reward leads to better game score than the one obtained by using the score itself as a reward. Given access to an event captioning mechanism, a few properties make Motif a general method: • it is entirely based on open models • the LLM doesn't need direct access to the environment dynamics (e.g., its source code) • the LLM doesn't need to understand observation and action spaces The best part? You can start using Motif right now, even on a small compute budget: the whole pipeline can take less than two GPU-days. Feel free to read our paper and try our code out. Paper: Code: Blog post: Work co-lead by Martin Klissarov and myself, with Shagun Sodhani Roberta Raileanu Pierre-Luc Bacon Pascal Vincent Amy Zhang Mikael Henaff Learn more in the thread 🧵

Pierluca D'Oro

311,883 次观看 • 2 年前

🚨 Copilots are not enough for ML engineering. Building ML pipelines with code snippets is like bringing a knife to a gunfight. Enter NEO by Neo AI: The first Agentic ML Engineer. 🔗

🚨 Copilots are not enough for ML engineering. Building ML pipelines with code snippets is like bringing a knife to a gunfight. Enter NEO by Neo AI: The first Agentic ML Engineer. 🔗

Riya Ai

89,187 次观看 • 11 个月前

"We have maybe the same size cake, and we just want to crush it with a giant reinforcement learning cherry." OpenAI's Dan Roberts discussed the controversial shift to prioritizing RL over pre-training at AI Ascent.

"We have maybe the same size cake, and we just want to crush it with a giant reinforcement learning cherry." OpenAI's Dan Roberts discussed the controversial shift to prioritizing RL over pre-training at AI Ascent.

Sequoia Capital

18,285 次观看 • 1 年前

Robot policies must be both reliable and highly capable to be useful; the best way to achieve this level of performance is with reinforcement learning. However, for reinforcement learning you are usually stuck between two difficult options: reinforcement in the real world is often risky and expensive, while reinforcement learning in a traditional simulator takes a lot of engineering work and has a persistent sim-to-real gap. What if instead you could train your robot purely in a world model? RISE by Jiazhi Yang et al. uses a compositional world model to predict the future and evaluate progress. This allows for a self-improving pipeline, which learns a world model from real data and then learns how the robot should perform different tasks. This pipeline results in a data-driven way to improve policy performance from real data but without real-world reinforcement learning. Watch Episode #86 of RoboPapers, with Chris Paxton and Jiafei Duan, to learn more!

Robot policies must be both reliable and highly capable to be useful; the best way to achieve this level of performance is with reinforcement learning. However, for reinforcement learning you are usually stuck between two difficult options: reinforcement in the real world is often risky and expensive, while reinforcement learning in a traditional simulator takes a lot of engineering work and has a persistent sim-to-real gap. What if instead you could train your robot purely in a world model? RISE by Jiazhi Yang et al. uses a compositional world model to predict the future and evaluate progress. This allows for a self-improving pipeline, which learns a world model from real data and then learns how the robot should perform different tasks. This pipeline results in a data-driven way to improve policy performance from real data but without real-world reinforcement learning. Watch Episode #86 of RoboPapers, with Chris Paxton and Jiafei Duan, to learn more!

RoboPapers

38,334 次观看 • 1 个月前

Here's a video explaining why The Elixir programming language's GenStage is designed the way it is –– and how that translates to better/faster pipeline performance:

Here's a video explaining why The Elixir programming language's GenStage is designed the way it is –– and how that translates to better/faster pipeline performance:

Anthony Accomazzo

12,348 次观看 • 1 年前

This is crazy... The first AI-native platform that builds, optimizes, and scales your e-commerce business is here. Runner AI takes you from one prompt to a fully functional, self-optimizing store in minutes. Here's how it works:

This is crazy... The first AI-native platform that builds, optimizes, and scales your e-commerce business is here. Runner AI takes you from one prompt to a fully functional, self-optimizing store in minutes. Here's how it works:

Angry Tom

12,268 次观看 • 4 个月前

Introducing NEO: The first Autonomous Machine Learning Engineer. It works like a full-stack ML engineer that never sleeps: handling data exploration, feature engineering, training, tuning, deployment, and monitoring, end to end. Powered by 11 specialized agents, NEO runs autonomously, saving ML engineers thousands of hours and making them 10x faster. Benchmarks: Tested on 75 Kaggle competitions, NEO scored a medal in 34.2% of them, significantly outperforming Microsoft’s RD Agent (22.4%) on OpenAI's MLE Bench. This sets a new state of the art for autonomous ML systems. NEO runs on a novel multi-agent orchestrator, powered by a multi-step reasoning engine, context transfer protocol, and agent memory, built to solve complex workflows end to end. And with human-in-the-loop mode, you can guide, inspect, and override any step. You're always in command. NEO is built for real world workflows and ready for production. NEO is here to make every ML engineer truly superhuman. Watch NEO in action:

Introducing NEO: The first Autonomous Machine Learning Engineer. It works like a full-stack ML engineer that never sleeps: handling data exploration, feature engineering, training, tuning, deployment, and monitoring, end to end. Powered by 11 specialized agents, NEO runs autonomously, saving ML engineers thousands of hours and making them 10x faster. Benchmarks: Tested on 75 Kaggle competitions, NEO scored a medal in 34.2% of them, significantly outperforming Microsoft’s RD Agent (22.4%) on OpenAI's MLE Bench. This sets a new state of the art for autonomous ML systems. NEO runs on a novel multi-agent orchestrator, powered by a multi-step reasoning engine, context transfer protocol, and agent memory, built to solve complex workflows end to end. And with human-in-the-loop mode, you can guide, inspect, and override any step. You're always in command. NEO is built for real world workflows and ready for production. NEO is here to make every ML engineer truly superhuman. Watch NEO in action:

Neo AI

859,951 次观看 • 10 个月前

NVIDIA's Jim Fan has verified from reliable sources that the video by EngineAI is a real robot demo and not CGI. The controller is a neural net trained in the NVIDIA Isaac simulator using reinforcement learning. There’s also another video from a different angle.

NVIDIA's Jim Fan has verified from reliable sources that the video by EngineAI is a real robot demo and not CGI. The controller is a neural net trained in the NVIDIA Isaac simulator using reinforcement learning. There’s also another video from a different angle.

The Humanoid Hub

148,535 次观看 • 1 年前

🚨 BREAKING: Language learning just entered its AI era. Meet Midoo AI — the world’s first AI language learning agent. Not another templated app. Not static flashcards. This is immersive, personalized, agentic learning. Here’s what makes it a game-changer 👇

🚨 BREAKING: Language learning just entered its AI era. Meet Midoo AI — the world’s first AI language learning agent. Not another templated app. Not static flashcards. This is immersive, personalized, agentic learning. Here’s what makes it a game-changer 👇

Pradeep Pandey

59,455 次观看 • 10 个月前

Wow. This is crazy. A developer trained an AI agent in simulation and deployed it onto a real robotic air hockey table using reinforcement learning. This robot can track the puck with millimeter-level accuracy and react in roughly 20 milliseconds, fast enough to challenge even skilled human players. We’re moving from robots that follow programmed rules to machines that learn strategies in simulation and execute them in the physical world.

Wow. This is crazy. A developer trained an AI agent in simulation and deployed it onto a real robotic air hockey table using reinforcement learning. This robot can track the puck with millimeter-level accuracy and react in roughly 20 milliseconds, fast enough to challenge even skilled human players. We’re moving from robots that follow programmed rules to machines that learn strategies in simulation and execute them in the physical world.

SciTech Era

1,583,835 次观看 • 1 个月前

David Silver says human data is AI's fossil fuel. "We mine it and burn it in our LLMs." It gave us a head start -- but it will run out. The sustainable path is reinforcement learning: systems that learn from experience, generate more, and keep learning.

David Silver says human data is AI's fossil fuel. "We mine it and burn it in our LLMs." It gave us a head start -- but it will run out. The sustainable path is reinforcement learning: systems that learn from experience, generate more, and keep learning.

vitrupo

20,190 次观看 • 1 年前

MITO AI is here. MITO AI is the tool we wish we had as filmmakers. It is not a prompt playground or a template generator. It is a control room for people who think in scenes, boards and timelines. You bring a story, references and intent. The tool helps you explore shots, workflows and versions faster than a traditional pipeline, without compromising taste. •⁠ ⁠Cinematic boards in minutes. •⁠ ⁠Keep your style consistent across images, scenes and videos. Designed by a studio that actually uses it every day. Join the early access list and help shape the tool:

MITO AI is here. MITO AI is the tool we wish we had as filmmakers. It is not a prompt playground or a template generator. It is a control room for people who think in scenes, boards and timelines. You bring a story, references and intent. The tool helps you explore shots, workflows and versions faster than a traditional pipeline, without compromising taste. •⁠ ⁠Cinematic boards in minutes. •⁠ ⁠Keep your style consistent across images, scenes and videos. Designed by a studio that actually uses it every day. Join the early access list and help shape the tool:

MITO AI

13,303 次观看 • 7 个月前