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。

How can we ensure robots using #foundationmodels, like LLMs, won’t “hallucinate” when executing tasks in complex, previously unseen environments? Our new SAFRON Advanced Research Concept seeks ideas to make sure #robots behave only as directed & intended.

DARPA

269,736 subscribers

11,052 次观看 • 1 年前 •via X (Twitter)

#foundationmodels #robots

Anya Rossi• Live Now

Private livecam show

9 条评论

Jay Geis 的头像

Jay Geis1 年前

Instead of making Super Soldiers, this is what the wizards are doing. It's still cool, don't get me wrong, it just shatters my dreams of being the leader of Hero Team.

Christine 的头像

Christine1 年前

Unplug them. Now.

LPK 的头像

LPK1 年前

GIORDANO 😈🤬😡👿👺😡👹😈🤬

⁴² Ministère de la sécurité cognitive du Québec 的头像

⁴² Ministère de la sécurité cognitive du Québec1 年前

I'm pretty sure hallucination can be weaponize, You know, when it need to be parralel but not accurate due to legal restriction. Or pseudo hallucination made to mess with the human pareidolia "muscle" for stealthy harassement

CurbsideRX 的头像

CurbsideRX1 年前

Are (we) the robots?

StarDrive Engineering Company.LLC 的头像

StarDrive Engineering Company.LLC1 年前

Sharefacts1985yrD.O.D we build artfactial intelligence warp dimensional drive saucer spacecrafts with 300,000 hours teaching vocabulary with verbal skills and social interaction to see with video cameras and mathematics skills And science skills and capabilities run spacecraft

LPK 的头像

LPK1 年前

I'm being tortured by this AH and brother Jay Giordano these men r EVIL AS HELL MIND BRAIN CONTROL TORTURE LYNN ROTHCHILD TO YHE THE SEX TRAFFICKING POS

Bull Meechum 的头像

Bull Meechum1 年前

@trevorlanting @Quantum_Murray @dwavequantum REAL quantum annealing can help w/that Same tech that did this QBTS

No Such Project 的头像

No Such Project1 年前

SchizoBots

相关视频

Robots Using Their Feet to Manipulate Objects ETH Zurich researchers have advanced "pedipulation," where legged robots use their feet for manipulation instead of robotic arms. Their new approach trains robots to navigate and manipulate objects while avoiding static and dynamic obstacles. Tested on the ANYmal quadruped, the method enables precise foot control in complex, unseen environments after minimal training in simulation. Will this research open up new possibilities for efficient, obstacle-aware robot manipulation?

Robots Using Their Feet to Manipulate Objects ETH Zurich researchers have advanced "pedipulation," where legged robots use their feet for manipulation instead of robotic arms. Their new approach trains robots to navigate and manipulate objects while avoiding static and dynamic obstacles. Tested on the ANYmal quadruped, the method enables precise foot control in complex, unseen environments after minimal training in simulation. Will this research open up new possibilities for efficient, obstacle-aware robot manipulation?

Circuit

11,802 次观看 • 1 年前

Robot AI brains, aka Vision-Language-Action models, cannot adapt to new tasks as easily as LLMs like Gemini, ChatGPT, or Grok. LLMs can adapt quickly with their in-context learning (ICL) capabilities. But can we inject ICL abilities into a pre-trained VLA like pi0? Yes! Introducing RICL (Retraining for In-Context Learning), our Conference on Robot Learning (CoRL) 2025 paper. Our RICL-pi0 model can adapt to unseen objects, novel motions, and new scenes with just ICL and RAG (retrieval-augmented generation). RICL-pi0 also boosts performance on the long-tail of tasks. A quick 1 minute video summary:

Robot AI brains, aka Vision-Language-Action models, cannot adapt to new tasks as easily as LLMs like Gemini, ChatGPT, or Grok. LLMs can adapt quickly with their in-context learning (ICL) capabilities. But can we inject ICL abilities into a pre-trained VLA like pi0? Yes! Introducing RICL (Retraining for In-Context Learning), our Conference on Robot Learning (CoRL) 2025 paper. Our RICL-pi0 model can adapt to unseen objects, novel motions, and new scenes with just ICL and RAG (retrieval-augmented generation). RICL-pi0 also boosts performance on the long-tail of tasks. A quick 1 minute video summary:

Kaustubh Sridhar

52,158 次观看 • 11 个月前

Our vision is for AI that uses world models to adapt in new and dynamic environments and efficiently learn new skills. We’re sharing V-JEPA 2, a new world model with state-of-the-art performance in visual understanding and prediction. V-JEPA 2 is a 1.2 billion-parameter model, trained on video, that can enable zero-shot planning in robots—allowing them to plan and execute tasks in unfamiliar environments. Learn more about V-JEPA 2 ➡️ As we continue working toward our goal of achieving advanced machine intelligence (AMI), we’re also releasing three new benchmarks for evaluating how well existing models can reason about the physical world from video. Learn more and download the new benchmarks ➡️

Our vision is for AI that uses world models to adapt in new and dynamic environments and efficiently learn new skills. We’re sharing V-JEPA 2, a new world model with state-of-the-art performance in visual understanding and prediction. V-JEPA 2 is a 1.2 billion-parameter model, trained on video, that can enable zero-shot planning in robots—allowing them to plan and execute tasks in unfamiliar environments. Learn more about V-JEPA 2 ➡️ As we continue working toward our goal of achieving advanced machine intelligence (AMI), we’re also releasing three new benchmarks for evaluating how well existing models can reason about the physical world from video. Learn more and download the new benchmarks ➡️

AI at Meta

310,281 次观看 • 1 年前

Robots need to be able to apply pressure and make contact with objects as needed in order to accomplish their tasks. From compliance to working safely around humans to whole-body manipulation of heavy objects, combining force and position control can dramatically expand the capabilities of robots. This is especially true for legged robots, which have so much ability to exert forces on the world around them. But how do we train robots which can do this? Baoxiong Jia tells us more in our discussion of his team’s recent, Best Paper Award winning work on learning a unified policy for position and force control, called UniFP. To learn more, watch Episode #49 of RoboPapers, hosted by Michael Cho - Rbt/Acc and Chris Paxton.

Robots need to be able to apply pressure and make contact with objects as needed in order to accomplish their tasks. From compliance to working safely around humans to whole-body manipulation of heavy objects, combining force and position control can dramatically expand the capabilities of robots. This is especially true for legged robots, which have so much ability to exert forces on the world around them. But how do we train robots which can do this? Baoxiong Jia tells us more in our discussion of his team’s recent, Best Paper Award winning work on learning a unified policy for position and force control, called UniFP. To learn more, watch Episode #49 of RoboPapers, hosted by Michael Cho - Rbt/Acc and Chris Paxton.

RoboPapers

44,803 次观看 • 8 个月前

Excited to present the LLM-Council skill. Initial idea by Karpathy. I just packaged it as a skill. You can easily spin up a council of LLMs or agents via Fireworks. Watch how the new GLM-5 model "deliberates" on other LLMs' thoughts on the big question, "Can LLMs reason?" Things worth paying attention to: New open models like GLM-5 have surprisingly improved on complex reasoning and long-running agentic tasks. The AskUserQuestion tool in Claude Code came in handy to select the council and chairperson. As Andrej Karpathy puts it, it's a really interesting way to get different perspectives from LLMs, which can lead to better decision-making on whatever task you are working on. You can use it for other agentic coding use cases, like evaluation, tool building, designing, and research.

Excited to present the LLM-Council skill. Initial idea by Karpathy. I just packaged it as a skill. You can easily spin up a council of LLMs or agents via Fireworks. Watch how the new GLM-5 model "deliberates" on other LLMs' thoughts on the big question, "Can LLMs reason?" Things worth paying attention to: New open models like GLM-5 have surprisingly improved on complex reasoning and long-running agentic tasks. The AskUserQuestion tool in Claude Code came in handy to select the council and chairperson. As Andrej Karpathy puts it, it's a really interesting way to get different perspectives from LLMs, which can lead to better decision-making on whatever task you are working on. You can use it for other agentic coding use cases, like evaluation, tool building, designing, and research.

elvis

39,593 次观看 • 5 个月前

Calling all mathematicians! We need novel advancements for science and security. We recently announced an Advanced Research Concepts (ARC) topic called COMPASS — Critical Orientation of Mathematics to Produce Advancements in Science and Security.

Calling all mathematicians! We need novel advancements for science and security. We recently announced an Advanced Research Concepts (ARC) topic called COMPASS — Critical Orientation of Mathematics to Produce Advancements in Science and Security.

DARPA

14,121 次观看 • 1 年前

Legged Locomotion… meets Skateboarding [Paper ⬇️] Most robot movement models either rely on fixed patterns or struggle to handle complex changes. DHAL (Discrete-time Hybrid Automata Learning) takes a different approach: using reinforcement learning to teach robots when and how to switch movements in real-time: ✅ Learns when to switch between different motions without pre-labeled data ✅ Handles complex, high-dimensional movements like a quadrupedal robot on a skateboard ✅ Uses a multi-critic architecture to improve contact-based motion control ✅ Works in both simulation and real-world environments with strong results It proves that robots can learn movement transitions on their own, without predefined rules. Paper: Thanks to Hang Liu for bringing this to my attention!

Legged Locomotion… meets Skateboarding [Paper ⬇️] Most robot movement models either rely on fixed patterns or struggle to handle complex changes. DHAL (Discrete-time Hybrid Automata Learning) takes a different approach: using reinforcement learning to teach robots when and how to switch movements in real-time: ✅ Learns when to switch between different motions without pre-labeled data ✅ Handles complex, high-dimensional movements like a quadrupedal robot on a skateboard ✅ Uses a multi-critic architecture to improve contact-based motion control ✅ Works in both simulation and real-world environments with strong results It proves that robots can learn movement transitions on their own, without predefined rules. Paper: Thanks to Hang Liu for bringing this to my attention!

Ilir Aliu - eu/acc

41,894 次观看 • 1 年前

Math is the source of enormous technological innovation and surprise, but it can take decades to make breakthroughs. Patrick Shafto is leading our new Exponentiating Mathematics program that will explore how AI can radically accelerate the rate of progress in mathematics.

Math is the source of enormous technological innovation and surprise, but it can take decades to make breakthroughs. Patrick Shafto is leading our new Exponentiating Mathematics program that will explore how AI can radically accelerate the rate of progress in mathematics.

DARPA

67,004 次观看 • 1 年前

The holy grail of robotics is to be able to perform previously-unseen, out-of-distribution manipulation tasks “zero shot” in a new environment. NovaFlow proposes an approach which (1) generates a video, (2) computes predicted flow — how points move through the scene — and (3) uses this flow as an objective to generate a motion. Using this procedure, NovaFlow generates motions in unseen scenes, for unseen tasks, and can transfer across embodiments. To learn more, we are joined by Hongyu Li and Jiahui Fu from RAI. Watch Episode #63 of RoboPapers with Chris Paxton and Michael Cho - Rbt/Acc now to learn more!

The holy grail of robotics is to be able to perform previously-unseen, out-of-distribution manipulation tasks “zero shot” in a new environment. NovaFlow proposes an approach which (1) generates a video, (2) computes predicted flow — how points move through the scene — and (3) uses this flow as an objective to generate a motion. Using this procedure, NovaFlow generates motions in unseen scenes, for unseen tasks, and can transfer across embodiments. To learn more, we are joined by Hongyu Li and Jiahui Fu from RAI. Watch Episode #63 of RoboPapers with Chris Paxton and Michael Cho - Rbt/Acc now to learn more!

RoboPapers

21,423 次观看 • 5 个月前

Excited to finally share Generative Value Learning (GVL), my Google DeepMind project on extracting universal value functions from long-context VLMs via in-context learning! We discovered a simple method to generate zero-shot and few-shot values for 300+ robot tasks and 50+ datasets using SOTA VLMs like Gemini (Try out the demo on our website on your robot video today!) I worked a lot on leveraging foundation models as guidance for robots in my PhD, and to me, this result forges a new frontier in how we can use foundation models for robot learning, given its broad applicability independent of embodiment and task types. Quite excited about how we can build on this work as a community!

Excited to finally share Generative Value Learning (GVL), my Google DeepMind project on extracting universal value functions from long-context VLMs via in-context learning! We discovered a simple method to generate zero-shot and few-shot values for 300+ robot tasks and 50+ datasets using SOTA VLMs like Gemini (Try out the demo on our website on your robot video today!) I worked a lot on leveraging foundation models as guidance for robots in my PhD, and to me, this result forges a new frontier in how we can use foundation models for robot learning, given its broad applicability independent of embodiment and task types. Quite excited about how we can build on this work as a community!

Jason Ma

98,090 次观看 • 1 年前

For decades, we’ve dreamed of robots that can seamlessly step into our world and lend a hand. Today, we take a major stride toward making that dream a reality: Introducing Gemini Robotics 2 from Google DeepMind, the intelligence layer powering the next generation of truly adaptable robots. This major advance unlocks intelligent whole-body control, advanced dexterity, and even multi-robot collaboration 🤯. Ok but... how does a robot actually "think"? Real-world tasks take time and planning. To manage that complexity, our new embodied reasoning model, Gemini Robotics ER 2, acts as the robot’s high-level brain, enhancing the robot’s capabilities to: — Observe the environment — Reason about the actions needed to complete the task — Coordinate with the vision-language-action model to carry out actions — Track progress until the job is done This setup allows robots to execute complex multi-step workflows, self-correct if a step fails, and adapt to completely novel situations. Learn more about Gemini Robotics ER 2 (and our two other brand new models) here:

Google AI

395,347 次观看 • 11 天前

I like to think of evals as something active, not passive -- it's a North Star that steers LLMs toward higher intelligence. Evals should drive your RL/SFT/post-training decisions. Internal evals at frontier labs make a huge difference -- and you can see it in how models behave differently (GPT seems better at one-shot tasks, Claude at multi-turn). If you want to learn more about building evals that actually improve your model in post-training, check out our AMD x DeeplearningAI course "Fine-tuning & RL for LLMs: Intro to Post-training" (content is free):

I like to think of evals as something active, not passive -- it's a North Star that steers LLMs toward higher intelligence. Evals should drive your RL/SFT/post-training decisions. Internal evals at frontier labs make a huge difference -- and you can see it in how models behave differently (GPT seems better at one-shot tasks, Claude at multi-turn). If you want to learn more about building evals that actually improve your model in post-training, check out our AMD x DeeplearningAI course "Fine-tuning & RL for LLMs: Intro to Post-training" (content is free):

Sharon Zhou

16,836 次观看 • 6 个月前

Karol Hausman is the co-founder and CEO of Physical Intelligence, a robotics company building a general-purpose “AI brain for the physical world.” The company has raised more than $1 billion in funding to develop foundation models that allow robots to operate across many machines, environments, and tasks rather than being programmed for a single purpose. In our conversation, we explore: • The moment a lecture from Sergey Levine convinced him to abandon his PhD research direction and pivot fully to deep learning • The case for building a general “AI brain” for the physical world rather than a single specialized robot • The role of real-world data in training robots, the limits of simulation, and how deployment could create a powerful data flywheel • The unique challenges of physical intelligence and why robots must operate with far higher reliability than language models Thank you to the partners who make this possible - Brex: The intelligent finance platform: - Granola: The app that might actually make you love meetings: Timestamps (00:00) Intro (04:05) Karol’s early fascination with robots (18:21) Karol’s entry point to robotics and PhD program (25:49) Combining robotics with LLMs: The Taylor Swift demo (30:48) The 1970s SHRDLU AI experiment (39:40) How research shapes what Physical Intelligence builds (49:07) The return of reinforcement learning in robotics (1:00:00) NVIDIA’s simulation engines (1:07:31) Compensating for missing senses

Karol Hausman is the co-founder and CEO of Physical Intelligence, a robotics company building a general-purpose “AI brain for the physical world.” The company has raised more than $1 billion in funding to develop foundation models that allow robots to operate across many machines, environments, and tasks rather than being programmed for a single purpose. In our conversation, we explore: • The moment a lecture from Sergey Levine convinced him to abandon his PhD research direction and pivot fully to deep learning • The case for building a general “AI brain” for the physical world rather than a single specialized robot • The role of real-world data in training robots, the limits of simulation, and how deployment could create a powerful data flywheel • The unique challenges of physical intelligence and why robots must operate with far higher reliability than language models Thank you to the partners who make this possible - Brex: The intelligent finance platform: - Granola: The app that might actually make you love meetings: Timestamps (00:00) Intro (04:05) Karol’s early fascination with robots (18:21) Karol’s entry point to robotics and PhD program (25:49) Combining robotics with LLMs: The Taylor Swift demo (30:48) The 1970s SHRDLU AI experiment (39:40) How research shapes what Physical Intelligence builds (49:07) The return of reinforcement learning in robotics (1:00:00) NVIDIA’s simulation engines (1:07:31) Compensating for missing senses

Mario Gabriele 🦊

27,871 次观看 • 4 个月前

A virtual being arrived in my home today. Within minutes my wife was talking to it like she talks to anyone else. Jeremy Bailenson, who runs VR/AR research at Stanford University told me their research showed the same: that humans treat virtual beings as real very fast. Thanks Illusion Of Life for bringing us Spark. The magic dog. Out of all the consumer electronics I have gotten or sold this is the most interesting of them all. Kiana will never know life without them. And Mehul we attached it to our Matic Robots and were driving it around. Virtual beings and robots. How much more futuristic can we get? On Thanksgiving we will introduce Spark to the family. And share that with you.

A virtual being arrived in my home today. Within minutes my wife was talking to it like she talks to anyone else. Jeremy Bailenson, who runs VR/AR research at Stanford University told me their research showed the same: that humans treat virtual beings as real very fast. Thanks Illusion Of Life for bringing us Spark. The magic dog. Out of all the consumer electronics I have gotten or sold this is the most interesting of them all. Kiana will never know life without them. And Mehul we attached it to our Matic Robots and were driving it around. Virtual beings and robots. How much more futuristic can we get? On Thanksgiving we will introduce Spark to the family. And share that with you.

Robert Scoble

37,577 次观看 • 8 个月前

Tencent announces AppAgent Multimodal Agents as Smartphone Users paper page: Recent advancements in large language models (LLMs) have led to the creation of intelligent agents capable of performing complex tasks. This paper introduces a novel LLM-based multimodal agent framework designed to operate smartphone applications. Our framework enables the agent to operate smartphone applications through a simplified action space, mimicking human-like interactions such as tapping and swiping. This novel approach bypasses the need for system back-end access, thereby broadening its applicability across diverse apps. Central to our agent's functionality is its innovative learning method. The agent learns to navigate and use new apps either through autonomous exploration or by observing human demonstrations. This process generates a knowledge base that the agent refers to for executing complex tasks across different applications. To demonstrate the practicality of our agent, we conducted extensive testing over 50 tasks in 10 different applications, including social media, email, maps, shopping, and sophisticated image editing tools. The results affirm our agent's proficiency in handling a diverse array of high-level tasks.

Tencent announces AppAgent Multimodal Agents as Smartphone Users paper page: Recent advancements in large language models (LLMs) have led to the creation of intelligent agents capable of performing complex tasks. This paper introduces a novel LLM-based multimodal agent framework designed to operate smartphone applications. Our framework enables the agent to operate smartphone applications through a simplified action space, mimicking human-like interactions such as tapping and swiping. This novel approach bypasses the need for system back-end access, thereby broadening its applicability across diverse apps. Central to our agent's functionality is its innovative learning method. The agent learns to navigate and use new apps either through autonomous exploration or by observing human demonstrations. This process generates a knowledge base that the agent refers to for executing complex tasks across different applications. To demonstrate the practicality of our agent, we conducted extensive testing over 50 tasks in 10 different applications, including social media, email, maps, shopping, and sophisticated image editing tools. The results affirm our agent's proficiency in handling a diverse array of high-level tasks.

AK

343,919 次观看 • 2 年前

A few weeks ago, we shared our progress on articulated objects and long-horizon tasks. Here are two representative examples: - We've been steadily expanding our asset library to cover more articulated objects. Articulated objects have always been a challenging asset class to handle in simulation. Interacting with them requires robots to master atomic skills such as pushing, pulling, opening, and closing, and to understand part structure, interaction constraints, and how the object moves. - Long-horizon tasks can now be generated at scale. Long-horizon tasks are the other hard category: they require chaining multiple sub-goals in sequence. A failure early in the task can cascade and make the rest unrecoverable. Axis is scaling along three dimensions at once: data volume, data quality, and task difficulty.

A few weeks ago, we shared our progress on articulated objects and long-horizon tasks. Here are two representative examples: - We've been steadily expanding our asset library to cover more articulated objects. Articulated objects have always been a challenging asset class to handle in simulation. Interacting with them requires robots to master atomic skills such as pushing, pulling, opening, and closing, and to understand part structure, interaction constraints, and how the object moves. - Long-horizon tasks can now be generated at scale. Long-horizon tasks are the other hard category: they require chaining multiple sub-goals in sequence. A failure early in the task can cascade and make the rest unrecoverable. Axis is scaling along three dimensions at once: data volume, data quality, and task difficulty.

Axis Robotics

13,700 次观看 • 1 个月前

A new robot policy just cleaned up a kitchen it had never seen before [watch what happens. paper included ⬇️] Pi-0.5 builds on top of Pi-0 and shows how smart co-training with diverse data can unlock real generalization in the home. It doesn’t just learn from one setup but adapts to many, including homes it’s never seen. What it does ✅ Handles new homes without training in them ✅ Follows complex language instructions ✅ Cleans, places dishes, handles spills ✅ Matches in-home training models using cross-embodiment and web data Robots that understand tasks and adapt to new spaces are finally within reach. More in the blog: Read the paper ⬇️ Physical Intelligence, co-founded by UC Berkeley professor Sergey Levine, is a robotics startup developing general-purpose AI foundation models that enable robots to perform a wide variety of real-world tasks with human-like adaptability, recently raising $400 million to advance this vision.

A new robot policy just cleaned up a kitchen it had never seen before [watch what happens. paper included ⬇️] Pi-0.5 builds on top of Pi-0 and shows how smart co-training with diverse data can unlock real generalization in the home. It doesn’t just learn from one setup but adapts to many, including homes it’s never seen. What it does ✅ Handles new homes without training in them ✅ Follows complex language instructions ✅ Cleans, places dishes, handles spills ✅ Matches in-home training models using cross-embodiment and web data Robots that understand tasks and adapt to new spaces are finally within reach. More in the blog: Read the paper ⬇️ Physical Intelligence, co-founded by UC Berkeley professor Sergey Levine, is a robotics startup developing general-purpose AI foundation models that enable robots to perform a wide variety of real-world tasks with human-like adaptability, recently raising $400 million to advance this vision.

Ilir Aliu - eu/acc

18,120 次观看 • 1 年前