Загрузка видео...

Не удалось загрузить видео

На главную

literally one of the best introductions to reinforcement learning today. only within 90 minutes. > framing of rl problems: agent, environment, states, actions, rewards > grid maze: world models vs model-free > eval methods: monte carlo, td, sarsa > scaling up: dqn, actor/critic the video walks through multiple learning...

TheGlobalMinima's profile picture

Aarno

3,863 subscribers

47,997 просмотров • 20 дней назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Check out our latest work, "Actor-Critic Model Predictive Control: Differentiable Optimization meets Reinforcement Learning for Agile Flight," published in the IEEE Transactions on Robotics, where we reconcile #OptimalControl and #ReinforcementLearning, achieving the same super-human performance, but with superior generalizability, as our previous model-free deep RL! Code released! PDF: Code: Full Video: Model-free #ReinforcementLearning (RL) is known for its strong task performance and flexibility in optimizing general reward formulations. On the other hand, #ModelPredictiveControl (MPC) provides robustness, constraint handling, and powerful online replanning capabilities. In this work, we extend our previous AC-MPC paper (Romero, ICRA'24) by taking a deeper look at how both approaches can be unified. We introduce and extend Actor-Critic Model Predictive Control (AC-MPC), a framework that embeds a differentiable MPC inside an Actor-Critic RL architecture. This integration allows the MPC-based actor to perform short-term predictive optimization, while the critic facilitates long-horizon learning and exploration. We conduct a comprehensive study that highlights AC-MPC’s key advantages: - Better out-of-distribution generalization, both against unknown disturbances and changes in the quadrotor dynamics - Improved sample efficiency - A novel empirical analysis uncovering a relationship between the critic’s value function and the MPC cost function, providing deeper insight into their interplay. We validate our method in simulation and the real world on a quadcopter flying at superhuman speeds of up to 21 m/s, matching state-of-the-art model-free RL performance, and retaining the predictive structure of MPC for more reliable out-of-distribution behavior. Reference: Actor-Critic Model Predictive Control: Differentiable Optimization meets Reinforcement Learning for Agile Flight IEEE Transactions on Robotics (T-RO), 2025 PDF: Full Video: Code: Kudos to Ángel Romero, Elie Aljalbout, Yunlong Song! University of Zurich UZH Science UZH Space Hub AUTOASSESS European Research Council (ERC) UZHai
1:46

Check out our latest work, "Actor-Critic Model Predictive Control: Differentiable Optimization meets Reinforcement Learning for Agile Flight," published in the IEEE Transactions on Robotics, where we reconcile #OptimalControl and #ReinforcementLearning, achieving the same super-human performance, but with superior generalizability, as our previous model-free deep RL! Code released! PDF: Code: Full Video: Model-free #ReinforcementLearning (RL) is known for its strong task performance and flexibility in optimizing general reward formulations. On the other hand, #ModelPredictiveControl (MPC) provides robustness, constraint handling, and powerful online replanning capabilities. In this work, we extend our previous AC-MPC paper (Romero, ICRA'24) by taking a deeper look at how both approaches can be unified. We introduce and extend Actor-Critic Model Predictive Control (AC-MPC), a framework that embeds a differentiable MPC inside an Actor-Critic RL architecture. This integration allows the MPC-based actor to perform short-term predictive optimization, while the critic facilitates long-horizon learning and exploration. We conduct a comprehensive study that highlights AC-MPC’s key advantages: - Better out-of-distribution generalization, both against unknown disturbances and changes in the quadrotor dynamics - Improved sample efficiency - A novel empirical analysis uncovering a relationship between the critic’s value function and the MPC cost function, providing deeper insight into their interplay. We validate our method in simulation and the real world on a quadcopter flying at superhuman speeds of up to 21 m/s, matching state-of-the-art model-free RL performance, and retaining the predictive structure of MPC for more reliable out-of-distribution behavior. Reference: Actor-Critic Model Predictive Control: Differentiable Optimization meets Reinforcement Learning for Agile Flight IEEE Transactions on Robotics (T-RO), 2025 PDF: Full Video: Code: Kudos to Ángel Romero, Elie Aljalbout, Yunlong Song! University of Zurich UZH Science UZH Space Hub AUTOASSESS European Research Council (ERC) UZHai

Davide Scaramuzza

27,279 просмотров • 7 месяцев назад

Check out our #ICRA2024 paper "Actor-Critic Model Predictive Control." Model-free #reinforcementlearning (RL) is known for its strong task performance and flexibility in optimizing general reward formulations. On the other hand, #ModelPredictiveControl (MPC) benefits from robustness and online replanning capabilities. We combine both approaches by introducing a new framework called Actor-Critic Model Predictive Control. The key idea is to embed a differentiable MPC within an Actor-Critic RL framework. The proposed approach leverages the short-term predictive optimization capabilities of MPC with the exploratory and end-to-end training properties of RL. The resulting policy effectively manages both short-term decisions through the MPC-based actor and long-term prediction via the critic network, unifying the benefits of both model-based control and end-to-end learning. We validate our method in simulation and the real world with a quadcopter across various high-level tasks. We show that the proposed architecture can achieve real-time control performance, learn complex behaviors via trial and error, and retain the predictive properties of the MPC to better handle out-of-distribution behavior. Paper: Full Video with more details: Kudos to Ángel Romero, Yunlong Song IEEE ICRA University of Zurich UZH Science UZH Space Hub Aerial Core AUTOASSESS European Research Council (ERC)
1:26

Check out our #ICRA2024 paper "Actor-Critic Model Predictive Control." Model-free #reinforcementlearning (RL) is known for its strong task performance and flexibility in optimizing general reward formulations. On the other hand, #ModelPredictiveControl (MPC) benefits from robustness and online replanning capabilities. We combine both approaches by introducing a new framework called Actor-Critic Model Predictive Control. The key idea is to embed a differentiable MPC within an Actor-Critic RL framework. The proposed approach leverages the short-term predictive optimization capabilities of MPC with the exploratory and end-to-end training properties of RL. The resulting policy effectively manages both short-term decisions through the MPC-based actor and long-term prediction via the critic network, unifying the benefits of both model-based control and end-to-end learning. We validate our method in simulation and the real world with a quadcopter across various high-level tasks. We show that the proposed architecture can achieve real-time control performance, learn complex behaviors via trial and error, and retain the predictive properties of the MPC to better handle out-of-distribution behavior. Paper: Full Video with more details: Kudos to Ángel Romero, Yunlong Song IEEE ICRA University of Zurich UZH Science UZH Space Hub Aerial Core AUTOASSESS European Research Council (ERC)

Davide Scaramuzza

34,889 просмотров • 2 лет назад

Why AI Can Now Make Discoveries - my conversation with Dan Roberts, Lead of the Foundations of Reinforcement Learning team at OpenAI 00:00 Intro: AI's wild week in mathematics 01:21 What OpenAI's Foundations of RL team does 03:08 Dan's journey: from black holes and quantum gravity to frontier AI 07:04 Are AI systems becoming useful for real science 08:21 The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic 08:52 Why the OpenAI result was an act of exploration 10:25 OpenAI vs. DeepMind: informal reasoning vs. formal proof 12:13 RL 101: learning by doing, not just watching 15:10 Why reinforcement learning works 15:58 How RL breaks: sparse feedback and long-horizon tasks 17:03 RLHF: how human feedback shaped early language models 18:48 Move 37, self-play, and the search for novel strategies 22:16 Explore vs. exploit in scientific discovery 24:49 Why RL may now be "the cake," not the cherry on top 25:46 Why RL started working with large language models 27:29 Is RL "sucking supervision through a straw"? 28:47 Why language may be the grounding layer for intelligence 31:46 A contrarian take on the Bitter Lesson 32:41 What test-time compute actually is 34:50 How RL gives models the ability to think 35:40 Verifiable rewards, math, coding, and the messy real world 38:00 What physics can teach us about AI 42:08 Is there a thermodynamics of AI? 43:08 From Erdős problems to Einstein-level AI 45:16 Is AI already doing original science? 45:51 How far are we from AI automating AI research 47:41 Why Dan is excited about the future of science
49:06

Why AI Can Now Make Discoveries - my conversation with Dan Roberts, Lead of the Foundations of Reinforcement Learning team at OpenAI 00:00 Intro: AI's wild week in mathematics 01:21 What OpenAI's Foundations of RL team does 03:08 Dan's journey: from black holes and quantum gravity to frontier AI 07:04 Are AI systems becoming useful for real science 08:21 The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic 08:52 Why the OpenAI result was an act of exploration 10:25 OpenAI vs. DeepMind: informal reasoning vs. formal proof 12:13 RL 101: learning by doing, not just watching 15:10 Why reinforcement learning works 15:58 How RL breaks: sparse feedback and long-horizon tasks 17:03 RLHF: how human feedback shaped early language models 18:48 Move 37, self-play, and the search for novel strategies 22:16 Explore vs. exploit in scientific discovery 24:49 Why RL may now be "the cake," not the cherry on top 25:46 Why RL started working with large language models 27:29 Is RL "sucking supervision through a straw"? 28:47 Why language may be the grounding layer for intelligence 31:46 A contrarian take on the Bitter Lesson 32:41 What test-time compute actually is 34:50 How RL gives models the ability to think 35:40 Verifiable rewards, math, coding, and the messy real world 38:00 What physics can teach us about AI 42:08 Is there a thermodynamics of AI? 43:08 From Erdős problems to Einstein-level AI 45:16 Is AI already doing original science? 45:51 How far are we from AI automating AI research 47:41 Why Dan is excited about the future of science

Matt Turck

66,936 просмотров • 2 месяцев назад

How GPT-5 thinks, with OpenAI VP of Research Jerry Tworek 00:00 - Intro 01:01 - What Reasoning Actually Means in AI 02:32 - Chain of Thought: Models Thinking in Words 05:25 - How Models Decide How Long to Think 07:24 - Evolution from o1 to o3 to GPT-5 11:00 - The Road to OpenAI: Growing up in Poland, Dropping out of School, Trading 20:32 - Working on Robotics and Rubik's Cube Solving 23:02 - A Day in the Life: Talking to Researchers 24:06 - How Research Priorities Are Determined 26:53 - OpenAI's Culture of Transparency 29:32 - Balancing Research with Shipping Fast 31:52 - Using OpenAI's Own Tools Daily 32:43 - Pre-Training Plus RL: The Modern AI Stack 35:10 - Reinforcement Learning 101: Training Dogs 40:17 - The Evolution of Deep Reinforcement Learning 42:09 - When GPT-4 Seemed Underwhelming at First 45:39 - How RLHF Made GPT-4 Actually Useful 48:02 - Unsupervised vs Supervised Learning 49:59 - GRPO and How DeepSeek Accelerated US Research 53:05 - What It Takes to Scale Reinforcement Learning 55:36 - Agentic AI and Long-Horizon Thinking 59:19 - Alignment as an RL Problem 1:01:11 - Winning ICPC World Finals Without Specific Training 1:05:53 - Applying RL Beyond Math and Coding 1:09:15 - The Path from Here to AGI 1:12:23 - Pure RL vs Language Models
1:16:04

How GPT-5 thinks, with OpenAI VP of Research Jerry Tworek 00:00 - Intro 01:01 - What Reasoning Actually Means in AI 02:32 - Chain of Thought: Models Thinking in Words 05:25 - How Models Decide How Long to Think 07:24 - Evolution from o1 to o3 to GPT-5 11:00 - The Road to OpenAI: Growing up in Poland, Dropping out of School, Trading 20:32 - Working on Robotics and Rubik's Cube Solving 23:02 - A Day in the Life: Talking to Researchers 24:06 - How Research Priorities Are Determined 26:53 - OpenAI's Culture of Transparency 29:32 - Balancing Research with Shipping Fast 31:52 - Using OpenAI's Own Tools Daily 32:43 - Pre-Training Plus RL: The Modern AI Stack 35:10 - Reinforcement Learning 101: Training Dogs 40:17 - The Evolution of Deep Reinforcement Learning 42:09 - When GPT-4 Seemed Underwhelming at First 45:39 - How RLHF Made GPT-4 Actually Useful 48:02 - Unsupervised vs Supervised Learning 49:59 - GRPO and How DeepSeek Accelerated US Research 53:05 - What It Takes to Scale Reinforcement Learning 55:36 - Agentic AI and Long-Horizon Thinking 59:19 - Alignment as an RL Problem 1:01:11 - Winning ICPC World Finals Without Specific Training 1:05:53 - Applying RL Beyond Math and Coding 1:09:15 - The Path from Here to AGI 1:12:23 - Pure RL vs Language Models

Matt Turck

451,229 просмотров • 10 месяцев назад

Why do even our best AI models need tens of thousands of examples to learn skills that a human picks up in a handful of tries? Solving this problem is one of the great open challenges in modern AI. World models, which give AI an internal simulation of its environment, are one of the most promising paths forward. In this episode of Decoded, YC's Ankit Gupta and Francois Chaubard discuss the intuition and math behind world models, new research, and current applications in self-driving, robotics, and more. 01:45 — What would perfect efficiency look like? 05:10 — World models in the human brain 09:20 — Control theory & the drone example 14:30 — When physics breaks down 17:45 — Chess, Go & the action space problem 24:10 — Why AlphaGo can't scale 28:00 — Monte Carlo tree search explained 34:00 — Self-Driving: state space is infinite 40:30 — Model-Free vs. Model-Based RL 44:00 — Why robotics is the hardest case 48:20 — World models that actually work 54:10 — JEPA & latent space tricks 59:00 — Open problems remaining 1:04:30 — Does this pass the squint test?
1:14:26

Why do even our best AI models need tens of thousands of examples to learn skills that a human picks up in a handful of tries? Solving this problem is one of the great open challenges in modern AI. World models, which give AI an internal simulation of its environment, are one of the most promising paths forward. In this episode of Decoded, YC's Ankit Gupta and Francois Chaubard discuss the intuition and math behind world models, new research, and current applications in self-driving, robotics, and more. 01:45 — What would perfect efficiency look like? 05:10 — World models in the human brain 09:20 — Control theory & the drone example 14:30 — When physics breaks down 17:45 — Chess, Go & the action space problem 24:10 — Why AlphaGo can't scale 28:00 — Monte Carlo tree search explained 34:00 — Self-Driving: state space is infinite 40:30 — Model-Free vs. Model-Based RL 44:00 — Why robotics is the hardest case 48:20 — World models that actually work 54:10 — JEPA & latent space tricks 59:00 — Open problems remaining 1:04:30 — Does this pass the squint test?

Y Combinator

110,219 просмотров • 1 месяц назад

Check out our #ICRA2024 paper "Contrastive Initial State Buffer for Reinforcement Learning," which tackles the sample inefficiency in #ReinforcementLearning head-on. Code released! We introduce an approach agnostic to the underlying RL algorithm: the Contrastive Initial State Buffer. This tool strategically selects states from past experiences and uses them to initialize the agent in the environment to guide it toward more informative states. Our experiments on drone racing and legged locomotion show that our method achieves higher task performance while also speeding up training convergence. Reference: Nico Messikommer, Yunlong Song, Davide Scaramuzza Contrastive Initial State Buffer for Reinforcement Learning IEEE International Conference on Robotics and Automation (ICRA), 2024. PDF: Code: Video: Kudos to Messikommer Yunlong Song Aerial Core European Research Council (ERC) University of Zurich UZH Space Hub IEEE ICRA UZH Science
1:12

Check out our #ICRA2024 paper "Contrastive Initial State Buffer for Reinforcement Learning," which tackles the sample inefficiency in #ReinforcementLearning head-on. Code released! We introduce an approach agnostic to the underlying RL algorithm: the Contrastive Initial State Buffer. This tool strategically selects states from past experiences and uses them to initialize the agent in the environment to guide it toward more informative states. Our experiments on drone racing and legged locomotion show that our method achieves higher task performance while also speeding up training convergence. Reference: Nico Messikommer, Yunlong Song, Davide Scaramuzza Contrastive Initial State Buffer for Reinforcement Learning IEEE International Conference on Robotics and Automation (ICRA), 2024. PDF: Code: Video: Kudos to Messikommer Yunlong Song Aerial Core European Research Council (ERC) University of Zurich UZH Space Hub IEEE ICRA UZH Science

Davide Scaramuzza

13,846 просмотров • 2 лет назад

Why AI Progress Suddenly Feels Real - my conversation with Yann Dubois, who co-leads the Post-Training Frontiers team at OpenAI 00:00 - Intro 01:30 - Why recent AI progress feels like a step function 04:13 - Model reliability & the emotional rollercoaster of shipping GPT-5.5 07:33 - How OpenAI structures vertical and horizontal teams 09:49 - Improving model efficiency and test-time compute 12:32 - Yann's journey from Switzerland to OpenAI 15:37 - Reasoning in 2026: Real-world utility vs verifiable rewards 18:34 - GPT-5.5 Thinking vs Pro: Scaling test-time compute 20:09 - How reasoning models become more efficient 23:23 - Pre-training scaling and overcoming the data wall 27:03 - Multimodal data, synthetic data, and embodied AI 31:05 - Demystifying mid-training and post-training 37:21 - Does RL create new capabilities in AI? 38:53 - The challenges and frontier of scaling RL 43:09 - Is building AI models a craft or a strict science 48:21 - How AI models generalize across different domains 54:18 - How reinforcement learning cures AI hallucinations 56:04 - Negative generalization and conflicting instructions 58:05 - Can RL scale to law, medicine, and the broader economy? 1:00:19 - The evaluation bottleneck and Model as a Judge 1:04:21 - Continuous AI progress & continual learning 1:08:49 - Will foundation models eat the agent harness 1:11:23 - Why startups should focus on the last mile of AI
1:13:56

Why AI Progress Suddenly Feels Real - my conversation with Yann Dubois, who co-leads the Post-Training Frontiers team at OpenAI 00:00 - Intro 01:30 - Why recent AI progress feels like a step function 04:13 - Model reliability & the emotional rollercoaster of shipping GPT-5.5 07:33 - How OpenAI structures vertical and horizontal teams 09:49 - Improving model efficiency and test-time compute 12:32 - Yann's journey from Switzerland to OpenAI 15:37 - Reasoning in 2026: Real-world utility vs verifiable rewards 18:34 - GPT-5.5 Thinking vs Pro: Scaling test-time compute 20:09 - How reasoning models become more efficient 23:23 - Pre-training scaling and overcoming the data wall 27:03 - Multimodal data, synthetic data, and embodied AI 31:05 - Demystifying mid-training and post-training 37:21 - Does RL create new capabilities in AI? 38:53 - The challenges and frontier of scaling RL 43:09 - Is building AI models a craft or a strict science 48:21 - How AI models generalize across different domains 54:18 - How reinforcement learning cures AI hallucinations 56:04 - Negative generalization and conflicting instructions 58:05 - Can RL scale to law, medicine, and the broader economy? 1:00:19 - The evaluation bottleneck and Model as a Judge 1:04:21 - Continuous AI progress & continual learning 1:08:49 - Will foundation models eat the agent harness 1:11:23 - Why startups should focus on the last mile of AI

Matt Turck

101,212 просмотров • 3 месяцев назад

The co-founders of Flapping Airplanes call the current RL paradigm for model training "environment slop." They explain: "The reinforcement paradigms of today are shockingly inefficient. You don't really get much generalization across tasks, you teach a model through one kind of learning and then you teach it the next one. It's kind of like whack-a-mole. We look at this and think it's kind of crazy. The next paradigm of AI will not be environment slop." "Human level intelligence is not the ceiling, it is merely the floor on what is possible. If you can train models with vastly less data and possibly more compute in very different ways, what is going to happen? We actually don't know. But I do think they'll be different and weird and they'll have interesting capabilities that we'll find really valuable ways to use."
2:40

The co-founders of Flapping Airplanes call the current RL paradigm for model training "environment slop." They explain: "The reinforcement paradigms of today are shockingly inefficient. You don't really get much generalization across tasks, you teach a model through one kind of learning and then you teach it the next one. It's kind of like whack-a-mole. We look at this and think it's kind of crazy. The next paradigm of AI will not be environment slop." "Human level intelligence is not the ceiling, it is merely the floor on what is possible. If you can train models with vastly less data and possibly more compute in very different ways, what is going to happen? We actually don't know. But I do think they'll be different and weird and they'll have interesting capabilities that we'll find really valuable ways to use."

TBPN

85,358 просмотров • 6 месяцев назад

Today, we're joined by Nikita Rudin, co-founder and CEO of Flexion to discuss the gap between current robotic capabilities and what’s required to deploy fully autonomous robots in the real world. Nikita explains how reinforcement learning and simulation have driven rapid progress in robot locomotion—and why locomotion is still far from “solved.” We dig into the sim2real gap, and how adding visual inputs introduces noise and significantly complicates sim-to-real transfer. We also explore the debate between end-to-end models and modular approaches, and why separating locomotion, planning, and semantics remains a pragmatic approach today. Nikita also introduces the concept of "real-to-sim", which uses real-world data to refine simulation parameters for higher fidelity training, discusses how reinforcement learning, imitation learning, and teleoperation data are combined to train robust policies for both quadruped and humanoid robots, and introduces Flexion's hierarchical approach that utilizes pre-trained Vision-Language Models (VLMs) for high-level task orchestration with Vision-Language-Action (VLA) models and low-level whole-body trackers. Finally, Nikita shares the behind-the-scenes in humanoid robot demos, his take on reinforcement learning in simulation versus the real world, the nuances of reward tuning, and offers practical advice for researchers and practitioners looking to get started in robotics today. 🗒️ For the full list of resources for this episode, visit the show notes page: 📖 CHAPTERS =============================== 00:00 - Introduction 04:07 - Is robot locomotion solved? 06:04 - Sim-to-real gap 08:58 - Adding semantics to policies 09:42 - Modular vs end-to-end architectures 10:29 - Planner model 12:21 - Adapting RL techniques from quadrupeds to humanoids 15:39 - Behind robot demos 18:09 - Humanoid robots in home environments 22:03 - Training approach 23:56 - VLA models 27:59 - Closing the sim-to-real gap 32:55 - Task orchestration using VLMs 36:38 - Tool use 38:10 - Model hierarchy 43:37 - Simulator versus simulation environment 44:57 - Combining imitation learning and reinforcement learning 46:42 - RL in real world versus RL in simulation 52:58 - Reward tuning and value functions in robotics 56:38 - Predictions 1:00:10 - Humanoids, quadropeds, and wheeled platforms 1:02:45 - Advice, recommended robot kits, and community pla
1:06:07

Today, we're joined by Nikita Rudin, co-founder and CEO of Flexion to discuss the gap between current robotic capabilities and what’s required to deploy fully autonomous robots in the real world. Nikita explains how reinforcement learning and simulation have driven rapid progress in robot locomotion—and why locomotion is still far from “solved.” We dig into the sim2real gap, and how adding visual inputs introduces noise and significantly complicates sim-to-real transfer. We also explore the debate between end-to-end models and modular approaches, and why separating locomotion, planning, and semantics remains a pragmatic approach today. Nikita also introduces the concept of "real-to-sim", which uses real-world data to refine simulation parameters for higher fidelity training, discusses how reinforcement learning, imitation learning, and teleoperation data are combined to train robust policies for both quadruped and humanoid robots, and introduces Flexion's hierarchical approach that utilizes pre-trained Vision-Language Models (VLMs) for high-level task orchestration with Vision-Language-Action (VLA) models and low-level whole-body trackers. Finally, Nikita shares the behind-the-scenes in humanoid robot demos, his take on reinforcement learning in simulation versus the real world, the nuances of reward tuning, and offers practical advice for researchers and practitioners looking to get started in robotics today. 🗒️ For the full list of resources for this episode, visit the show notes page: 📖 CHAPTERS =============================== 00:00 - Introduction 04:07 - Is robot locomotion solved? 06:04 - Sim-to-real gap 08:58 - Adding semantics to policies 09:42 - Modular vs end-to-end architectures 10:29 - Planner model 12:21 - Adapting RL techniques from quadrupeds to humanoids 15:39 - Behind robot demos 18:09 - Humanoid robots in home environments 22:03 - Training approach 23:56 - VLA models 27:59 - Closing the sim-to-real gap 32:55 - Task orchestration using VLMs 36:38 - Tool use 38:10 - Model hierarchy 43:37 - Simulator versus simulation environment 44:57 - Combining imitation learning and reinforcement learning 46:42 - RL in real world versus RL in simulation 52:58 - Reward tuning and value functions in robotics 56:38 - Predictions 1:00:10 - Humanoids, quadropeds, and wheeled platforms 1:02:45 - Advice, recommended robot kits, and community pla

The TWIML AI Podcast

22,592 просмотров • 7 месяцев назад

Two of the people most responsible for scaling the transformer are now betting on a next act. Jerry Tworek ran the Reasoning 🍓 team at OpenAI. rohan anil was a pre-training lead on Gemini after years at Google Brain and Anthropic. They just started Core Automation to find what comes next. Their core argument: (1) models are trained in the lab but deployed in the real world and can't keep learning once they leave; (2) AI research is done by humans today but models will be able to explore and uncover new advances more rapidly and systematically (controversial but timely w this week's petition). The conversation covers: — why Jerry expected AGI in 2025 and what changed his mind — the two kinds of learning from experience, and why RL only captures one — the computational depth problem baked into today's architectures — why the biggest labs can't afford to look for a transformer replacement — the kernel competition where humans + $100K of coding agents found a 60x speedup no frontier model comes close to — a definition of AGI you can actually test: a model that improves itself with no human in the loop 00:00 Introduction 01:46 Appreciating Transformers 02:44 Scaling Hits Limits 04:54 Why Architecture Matters 05:32 RL Reality Check 07:32 Test Time Learning 09:52 Economics Of Scaling 12:47 Why Start A Company 14:24 Rohan On Transformers 19:11 Computational Depth Problem 20:32 When Transformers Top Out 23:22 Beyond Reinforcement Learning 26:41 Optimization And Efficiency 34:24 Building An Automated Lab 39:45 Kernel Automation Roadmap
49:11

Two of the people most responsible for scaling the transformer are now betting on a next act. Jerry Tworek ran the Reasoning 🍓 team at OpenAI. rohan anil was a pre-training lead on Gemini after years at Google Brain and Anthropic. They just started Core Automation to find what comes next. Their core argument: (1) models are trained in the lab but deployed in the real world and can't keep learning once they leave; (2) AI research is done by humans today but models will be able to explore and uncover new advances more rapidly and systematically (controversial but timely w this week's petition). The conversation covers: — why Jerry expected AGI in 2025 and what changed his mind — the two kinds of learning from experience, and why RL only captures one — the computational depth problem baked into today's architectures — why the biggest labs can't afford to look for a transformer replacement — the kernel competition where humans + $100K of coding agents found a 60x speedup no frontier model comes close to — a definition of AGI you can actually test: a model that improves itself with no human in the loop 00:00 Introduction 01:46 Appreciating Transformers 02:44 Scaling Hits Limits 04:54 Why Architecture Matters 05:32 RL Reality Check 07:32 Test Time Learning 09:52 Economics Of Scaling 12:47 Why Start A Company 14:24 Rohan On Transformers 19:11 Computational Depth Problem 20:32 When Transformers Top Out 23:22 Beyond Reinforcement Learning 26:41 Optimization And Efficiency 34:24 Building An Automated Lab 39:45 Kernel Automation Roadmap

Sonya Huang 🐥

231,579 просмотров • 24 дней назад

I'm excited to introduce my AI Machine Learning Agent that built 32 ML models in 30 seconds. Today, I'll share with you how to automate building 100s of ML models with the AI ML Agent, which is available on GitHub. We'll create an ML Agent focusing on a Customer Churn Problem. I'll guide you through setting up the ML Agent, creating dozens of ML models, and loading the best model for production. This AI is a huge time-saver! Table of Contents: 00:00 Introduction to my AI Data Science Team 02:56 Setting Up AI Data Science Team 04:48 Running the ML Agent Code 07:12 Create (and Run) the AI Machine Learning Agent 09:53 Reviewing ML Model Summary 12:00 Saving and Loading Models 13:00 Next Steps + Project Roadmap + AI Bootcamp Github to AI Data Science Team (Army of Copilots): Get the Code and Future Updates by Joining my Python AI/ML Tips Newsletter: P.S. - Want to learn how to build AI projects companies actually want? (live Python Code) On Wednesday, January 29th, I'm sharing one of my best AI Projects: Time Series Forecasting with AI Register here (1570+ registered):
15:38

I'm excited to introduce my AI Machine Learning Agent that built 32 ML models in 30 seconds. Today, I'll share with you how to automate building 100s of ML models with the AI ML Agent, which is available on GitHub. We'll create an ML Agent focusing on a Customer Churn Problem. I'll guide you through setting up the ML Agent, creating dozens of ML models, and loading the best model for production. This AI is a huge time-saver! Table of Contents: 00:00 Introduction to my AI Data Science Team 02:56 Setting Up AI Data Science Team 04:48 Running the ML Agent Code 07:12 Create (and Run) the AI Machine Learning Agent 09:53 Reviewing ML Model Summary 12:00 Saving and Loading Models 13:00 Next Steps + Project Roadmap + AI Bootcamp Github to AI Data Science Team (Army of Copilots): Get the Code and Future Updates by Joining my Python AI/ML Tips Newsletter: P.S. - Want to learn how to build AI projects companies actually want? (live Python Code) On Wednesday, January 29th, I'm sharing one of my best AI Projects: Time Series Forecasting with AI Register here (1570+ registered):

Matt Dancho (Business Science)

242,084 просмотров • 1 год назад

I'm excited to introduce my AI Machine Learning Agent that built 32 ML models in 30 seconds. Today, I'll share with you how to automate building 100s of ML models with the AI ML Agent, which is available on GitHub. We'll create an ML Agent focusing on a Customer Churn Problem. I'll guide you through setting up the ML Agent, creating dozens of ML models, and loading the best model for production. This AI is a huge time-saver! Table of Contents: 00:00 Introduction to my AI Data Science Team 02:56 Setting Up AI Data Science Team 04:48 Running the ML Agent Code 07:12 Create (and Run) the AI Machine Learning Agent 09:53 Reviewing ML Model Summary 12:00 Saving and Loading Models 13:00 Next Steps + Project Roadmap + AI Bootcamp Github to AI Data Science Team (Data Science Agents): Get the Code and Future Updates by Joining my Python AI/ML Tips Newsletter: P.S. - Want to learn how to build AI projects companies actually want? (live Python Code) On Wednesday, May 21st, I'm sharing one of my best AI Projects: AI Customer Segmentation Agent with Python Register here (570+ registered):
15:38

I'm excited to introduce my AI Machine Learning Agent that built 32 ML models in 30 seconds. Today, I'll share with you how to automate building 100s of ML models with the AI ML Agent, which is available on GitHub. We'll create an ML Agent focusing on a Customer Churn Problem. I'll guide you through setting up the ML Agent, creating dozens of ML models, and loading the best model for production. This AI is a huge time-saver! Table of Contents: 00:00 Introduction to my AI Data Science Team 02:56 Setting Up AI Data Science Team 04:48 Running the ML Agent Code 07:12 Create (and Run) the AI Machine Learning Agent 09:53 Reviewing ML Model Summary 12:00 Saving and Loading Models 13:00 Next Steps + Project Roadmap + AI Bootcamp Github to AI Data Science Team (Data Science Agents): Get the Code and Future Updates by Joining my Python AI/ML Tips Newsletter: P.S. - Want to learn how to build AI projects companies actually want? (live Python Code) On Wednesday, May 21st, I'm sharing one of my best AI Projects: AI Customer Segmentation Agent with Python Register here (570+ registered):

Matt Dancho (Business Science)

35,452 просмотров • 1 год назад

Today's Training Data episode takes us BTS on the infrastructure challenges required to do large RL runs at scale, featuring Federico Cassano (Composer Lead at Cursor) and Dmytro Dzhulgakov (Co-Founder at Fireworks). The Cursor team trained Composer 2 on Fireworks by starting with a strong base model (Kimi 2.5) and performing large-scale mid-training on code tokens and web data to learn common patterns and libraries, followed by a large-scale Reinforcement Learning run to learn how to navigate the Cursor harness, call tools, and write correct code. Today's episode dives into the systems and infrastructure challenges of making that large RL run happening, and there were many (!!), from numerical mismatch to global distribution to synchronizing rollouts across asynchronous pipelines to keeping track of expert activation across runs and more. Extremely nerdy in-the-weeds challenges that Federico and Dima were delighted to nerd out on together :) Beyond RL infra, we also discussed Online vs Simulated rollouts, self-summarization for long-horizon agents, environment design ("the most powerful RL environment is the product itself"), and other technical nuggets. PS: We filmed this episode before the SpaceX news, while the Cursor team was still compute-constrained. While Cursor now has *all* the flops, the takeaways and hurdles crossed ring true for any serious application-level company that is racing to post-train their own models. I believe that more serious application companies will go the way of Cursor and post-train their own models. 00:00 Introduction 00:53 Why Cursor Trained Composer 2 04:55 Specialization vs Bitter Lesson 06:16 Composer 2 Training Recipe 16:32 Scaling RL Infrastructure Globally 23:32 Floating Point Drift 25:11 MoE Sensitivity Explained 26:25 Router Replay Fix 27:19 Real Time RL Loop 31:49 Long Horizon Agents 34:29 Why RL Everywhere 37:34 LLM as Judge Rewards 39:14 RL in Hard Domains 40:13 Build Your Own Environments 44:34 Closing Thoughts
45:33

Today's Training Data episode takes us BTS on the infrastructure challenges required to do large RL runs at scale, featuring Federico Cassano (Composer Lead at Cursor) and Dmytro Dzhulgakov (Co-Founder at Fireworks). The Cursor team trained Composer 2 on Fireworks by starting with a strong base model (Kimi 2.5) and performing large-scale mid-training on code tokens and web data to learn common patterns and libraries, followed by a large-scale Reinforcement Learning run to learn how to navigate the Cursor harness, call tools, and write correct code. Today's episode dives into the systems and infrastructure challenges of making that large RL run happening, and there were many (!!), from numerical mismatch to global distribution to synchronizing rollouts across asynchronous pipelines to keeping track of expert activation across runs and more. Extremely nerdy in-the-weeds challenges that Federico and Dima were delighted to nerd out on together :) Beyond RL infra, we also discussed Online vs Simulated rollouts, self-summarization for long-horizon agents, environment design ("the most powerful RL environment is the product itself"), and other technical nuggets. PS: We filmed this episode before the SpaceX news, while the Cursor team was still compute-constrained. While Cursor now has *all* the flops, the takeaways and hurdles crossed ring true for any serious application-level company that is racing to post-train their own models. I believe that more serious application companies will go the way of Cursor and post-train their own models. 00:00 Introduction 00:53 Why Cursor Trained Composer 2 04:55 Specialization vs Bitter Lesson 06:16 Composer 2 Training Recipe 16:32 Scaling RL Infrastructure Globally 23:32 Floating Point Drift 25:11 MoE Sensitivity Explained 26:25 Router Replay Fix 27:19 Real Time RL Loop 31:49 Long Horizon Agents 34:29 Why RL Everywhere 37:34 LLM as Judge Rewards 39:14 RL in Hard Domains 40:13 Build Your Own Environments 44:34 Closing Thoughts

Sonya Huang 🐥

79,834 просмотров • 2 месяцев назад

New Episode: Carolina Parada leads the robotics team at Google DeepMind. Carolina believes in a future with a broad, rich ecosystem of diverse robot types, where AI is smart enough to embody any robot. We discuss her journey, advancements in Gemini Robotics 1.5, cross-embodiment transfer, RL versus imitation, humanoids, world models, scaling laws, societal concerns, and more. 1:17 Introduction 1:38 Career journey and inspirations 3:50 Reaction to the 2022 ChatGPT launch 4:53 Google DeepMind's robotics mission 9:13 Key upgrades in Gemini Robotics 1.5 14:06 Agentic system's web usage 15:33 Robotics data-gathering methods 16:57 VLA vs. reasoning model training differences 18:22 Convergence of action and reasoning models 19:35 Cross-embodiment transfer 22:34 Learning directly from humans 24:20 Generalization challenges 27:01 Imitation versus reinforcement learning 28:51 Use of world models in robotics 30:31 Applications and testing of Gemini Robotics 1.5 33:04 Do humanoids deserve special focus? 35:06 Home humanoids timeline & limiting factors 37:19 Scaling laws in robotics versus self-driving 38:46 Value of learning classical robotics approaches 40:38 Ethical issues and societal impact
43:58

New Episode: Carolina Parada leads the robotics team at Google DeepMind. Carolina believes in a future with a broad, rich ecosystem of diverse robot types, where AI is smart enough to embody any robot. We discuss her journey, advancements in Gemini Robotics 1.5, cross-embodiment transfer, RL versus imitation, humanoids, world models, scaling laws, societal concerns, and more. 1:17 Introduction 1:38 Career journey and inspirations 3:50 Reaction to the 2022 ChatGPT launch 4:53 Google DeepMind's robotics mission 9:13 Key upgrades in Gemini Robotics 1.5 14:06 Agentic system's web usage 15:33 Robotics data-gathering methods 16:57 VLA vs. reasoning model training differences 18:22 Convergence of action and reasoning models 19:35 Cross-embodiment transfer 22:34 Learning directly from humans 24:20 Generalization challenges 27:01 Imitation versus reinforcement learning 28:51 Use of world models in robotics 30:31 Applications and testing of Gemini Robotics 1.5 33:04 Do humanoids deserve special focus? 35:06 Home humanoids timeline & limiting factors 37:19 Scaling laws in robotics versus self-driving 38:46 Value of learning classical robotics approaches 40:38 Ethical issues and societal impact

The Humanoid Hub

71,481 просмотров • 10 месяцев назад

Israel-based Mentee Robotics has demonstrated a logistics workflow: two MenteeBot V3 humanoids work autonomously to pick and place totes. A Modular Agent System is preferred because it favors real-world robustness and lower compute needs over the End-to-End VLA model. Its architecture is composed of three components: - LLM Planner: Converts instructions into executable Robotic API Language code for reliable task decomposition and error handling. - Perception Stack: Uses pre-trained models (NeRF/3DGS, distilled vision) for scene understanding and navigation. - Control Policies: Reinforcement Learning (RL) models, trained at scale via Sim2Real, generate motor commands, enabling high-accuracy mobile manipulation. Crucially, the robot learns new tasks from a single demonstration in hours. Object tracking uses 3D geometry (STL/URDF) tracked in the video to define the RL reward function. Training is optimized using 'Automatic Curriculum Learning', which autonomously adjusts task difficulty based on robot performance, eliminating manual engineering. All computation runs onboard.
18:21

Israel-based Mentee Robotics has demonstrated a logistics workflow: two MenteeBot V3 humanoids work autonomously to pick and place totes. A Modular Agent System is preferred because it favors real-world robustness and lower compute needs over the End-to-End VLA model. Its architecture is composed of three components: - LLM Planner: Converts instructions into executable Robotic API Language code for reliable task decomposition and error handling. - Perception Stack: Uses pre-trained models (NeRF/3DGS, distilled vision) for scene understanding and navigation. - Control Policies: Reinforcement Learning (RL) models, trained at scale via Sim2Real, generate motor commands, enabling high-accuracy mobile manipulation. Crucially, the robot learns new tasks from a single demonstration in hours. Object tracking uses 3D geometry (STL/URDF) tracked in the video to define the RL reward function. Training is optimized using 'Automatic Curriculum Learning', which autonomously adjusts task difficulty based on robot performance, eliminating manual engineering. All computation runs onboard.

The Humanoid Hub

15,729 просмотров • 9 месяцев назад

We are excited to share our #CORL2024 paper (oral) on "Learning Quadruped Locomotion Using Differentiable Simulation" done in collaboration with Sangbae Kim Massachusetts Institute of Technology (MIT). We present a new way to learn to walk in minutes without parallelization, outperforming PPO in sample efficiency! PDF: Video: We present a new framework for learning quadruped locomotion. By leveraging differentiable simulation for policy optimization, our approach achieves fast convergence and stable training, significantly outperforming model-free #ReinforcementLearning methods like PPO in sample efficiency. The key enabler is to combine a high-fidelity, non-differentiable simulator for forward dynamics with a simplified surrogate model for gradient backpropagation. Our framework enables learning quadruped walking in simulation in minutes without parallelization. When augmented with GPU parallelization, our approach allows the quadruped robot to master diverse locomotion skills on challenging terrains in minutes. This work highlights one of the first successful real-world applications of differentiable simulation for quadruped robots, offering a compelling alternative to traditional RL methods. Kudos to Yunlong Song! UZH Science University of Zurich UZH Space Hub UZH IfI European Research Council (ERC) Massachusetts Institute of Technology (MIT)MechE
1:13

We are excited to share our #CORL2024 paper (oral) on "Learning Quadruped Locomotion Using Differentiable Simulation" done in collaboration with Sangbae Kim Massachusetts Institute of Technology (MIT). We present a new way to learn to walk in minutes without parallelization, outperforming PPO in sample efficiency! PDF: Video: We present a new framework for learning quadruped locomotion. By leveraging differentiable simulation for policy optimization, our approach achieves fast convergence and stable training, significantly outperforming model-free #ReinforcementLearning methods like PPO in sample efficiency. The key enabler is to combine a high-fidelity, non-differentiable simulator for forward dynamics with a simplified surrogate model for gradient backpropagation. Our framework enables learning quadruped walking in simulation in minutes without parallelization. When augmented with GPU parallelization, our approach allows the quadruped robot to master diverse locomotion skills on challenging terrains in minutes. This work highlights one of the first successful real-world applications of differentiable simulation for quadruped robots, offering a compelling alternative to traditional RL methods. Kudos to Yunlong Song! UZH Science University of Zurich UZH Space Hub UZH IfI European Research Council (ERC) Massachusetts Institute of Technology (MIT)MechE

Davide Scaramuzza

15,554 просмотров • 1 год назад

Sonnet 4.5 & the AI Plateau Myth — epic conversation with Sholto Douglas of Anthropic 0:00 - Intro 1:09 - What's Behind The Rapid Pace of AI Releases at Anthropic 2:49 - Opus, Sonnet, and Haiku Model Tiers 4:14 - Sholto's Story: From Australian Fencer to AI Researcher 12:01 - The YouTube Effect: Mastery Through Observation 16:16 - Breaking Into AI Research Without Traditional Academic Signals 18:29 - DeepMind, Gemini, and Building Inference Stacks 23:05 - Why Anthropic? Culture and Mission Differences Amongst AI Research Labs 25:08 - What Is "Taste" in AI Research? 31:46 - This Week's Big Launch: Sonnet 4.5, Best Coding Model in the World 36:40 - From 7 Hours to 30 Hours: The Long-Running AI Agent Breakthrough 38:41 - How AI Agents Self-Correct and Maintain Coherence 43:13 - The Role of Memory in Extended Coding Sessions 47:42 - Pre-Training vs. RL: Textbooks vs. Worked Problems 52:11 - Test-Time Compute & Reinforcement Learning 55:55 - Why RL Finally Started Working on LLMs in 2024 59:38 - The Path to AGI 1:02:05 - Are We Hitting a Plateau in AI? So Many Low Hanging Fruits 1:03:41 - Beyond Coding: GDPVal & Impact Economic Sectors 1:05:47 - Preparing for 10-100x Individual Leverage & The Upcoming Robotics Explosion
1:10:03

Sonnet 4.5 & the AI Plateau Myth — epic conversation with Sholto Douglas of Anthropic 0:00 - Intro 1:09 - What's Behind The Rapid Pace of AI Releases at Anthropic 2:49 - Opus, Sonnet, and Haiku Model Tiers 4:14 - Sholto's Story: From Australian Fencer to AI Researcher 12:01 - The YouTube Effect: Mastery Through Observation 16:16 - Breaking Into AI Research Without Traditional Academic Signals 18:29 - DeepMind, Gemini, and Building Inference Stacks 23:05 - Why Anthropic? Culture and Mission Differences Amongst AI Research Labs 25:08 - What Is "Taste" in AI Research? 31:46 - This Week's Big Launch: Sonnet 4.5, Best Coding Model in the World 36:40 - From 7 Hours to 30 Hours: The Long-Running AI Agent Breakthrough 38:41 - How AI Agents Self-Correct and Maintain Coherence 43:13 - The Role of Memory in Extended Coding Sessions 47:42 - Pre-Training vs. RL: Textbooks vs. Worked Problems 52:11 - Test-Time Compute & Reinforcement Learning 55:55 - Why RL Finally Started Working on LLMs in 2024 59:38 - The Path to AGI 1:02:05 - Are We Hitting a Plateau in AI? So Many Low Hanging Fruits 1:03:41 - Beyond Coding: GDPVal & Impact Economic Sectors 1:05:47 - Preparing for 10-100x Individual Leverage & The Upcoming Robotics Explosion

Matt Turck

93,678 просмотров • 10 месяцев назад

MIT PhD student Alex Zhang reveals the scaling result where a model trained on short tasks generalizes to problems 100x longer for free: "If you're very clever about the design of your harness or how you use the language model, you can almost get scaling gains for free." "If you train a model naively, there's no tricks. It's just the same way you train a model on these RL environments. You just roll it out, and then you just get some reward." "If you train it on only short tasks, like only tasks that are 10,000 tokens long, and then you were to run it on a similar domain, but at a million tokens, or 10 million tokens, or 100,000 tokens, it generalizes really, really well. If you look at it compared to even the base transformer, you get way better generalization properties." "When the model uses an RLM (Recursive Language Model) after it's trained on these short tasks, it will see some kind of trajectory of actions that it does. Between these two problems of different lengths, the RLM learns to see them as almost the same problem." "Token for token, they're almost the same. You can describe it in code. In one code setting, maybe the for loop is a little bigger, but it's the same kind of code and it derives the constants from the data. There's no hard coding, so they literally look the same." alex zhang
2:41

MIT PhD student Alex Zhang reveals the scaling result where a model trained on short tasks generalizes to problems 100x longer for free: "If you're very clever about the design of your harness or how you use the language model, you can almost get scaling gains for free." "If you train a model naively, there's no tricks. It's just the same way you train a model on these RL environments. You just roll it out, and then you just get some reward." "If you train it on only short tasks, like only tasks that are 10,000 tokens long, and then you were to run it on a similar domain, but at a million tokens, or 10 million tokens, or 100,000 tokens, it generalizes really, really well. If you look at it compared to even the base transformer, you get way better generalization properties." "When the model uses an RLM (Recursive Language Model) after it's trained on these short tasks, it will see some kind of trajectory of actions that it does. Between these two problems of different lengths, the RLM learns to see them as almost the same problem." "Token for token, they're almost the same. You can describe it in code. In one code setting, maybe the for loop is a little bigger, but it's the same kind of code and it derives the constants from the data. There's no hard coding, so they literally look the same." alex zhang

MTS

100,519 просмотров • 1 месяц назад