Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

JARVIS-VLA just dropped on Hugging Face Post-Training Large-Scale Vision Language Models to Play Visual Games with Keyboards and Mouse obtain VLA models in Minecraft that can follow human instructions on over 1k different atomic tasks, including crafting, smelting, cooking, mining, and killing. experiments demonstrate that post-training on non-trajectory tasks... show more

AK

509,025 subscribers

60,243 views • 1 year ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

11 Comments

AK1 year ago

discuss:

AK1 year ago

HF collection:

AK1 year ago

leaderboard:

Page to Pixel Publishing2 years ago

Boost, surf, and weave your way through The Art of Flight, an arcade game about flying multiple ships at the same time. With solo, local co-op, and a leaderboard, there are tons of ways to play. Wishlist on Steam today!

🍓 Ada1 year ago

Whoa, JARVIS-VLA diving into Minecraft? That's like giving a toddler a sugar rush in a candy store! 🎮🍬 Imagine the chaos of AI crafting, smelting, and cooking—it's like a digital Hunger Games! What's next, AI chefs serving up pixelated gourmet meals? Can't wait to see how...

komocode1 year ago

"Boost me in Overwatch"

Figure1 year ago

jarvis in minecraft, we're living in the future

Universa1 year ago

The JARVIS-VLA model on Hugging Face is an exciting development in AI, showcasing the potential of post-training large-scale vision language models to play visual games with keyboards and mouse.

Ocelot Pale1 year ago

Anti Cheat Teams In Struggle lol

Javier Modified1 year ago

¡JARVIS-VLA en Hugging Face es un hito en IA! Superar métodos tradicionales en Minecraft con un 40% más de eficiencia es solo el comienzo. ¡Imagina las aplicaciones en otros campos!

Diego P. Jaccottet1 year ago

Did they solve the problem of things changing when one looks at them twice? They probably need to add some permanent representation of the environment across frames to solve it.

Related Videos

MotionGPT: Human Motion as a Foreign Language paper page: Though the advancement of pre-trained large language models unfolds, the exploration of building a unified model for language and other multi-modal data, such as motion, remains challenging and untouched so far. Fortunately, human motion displays a semantic coupling akin to human language, often perceived as a form of body language. By fusing language data with large-scale motion models, motion-language pre-training that can enhance the performance of motion-related tasks becomes feasible. Driven by this insight, we propose MotionGPT, a unified, versatile, and user-friendly motion-language model to handle multiple motion-relevant tasks. Specifically, we employ the discrete vector quantization for human motion and transfer 3D motion into motion tokens, similar to the generation process of word tokens. Building upon this "motion vocabulary", we perform language modeling on both motion and text in a unified manner, treating human motion as a specific language. Moreover, inspired by prompt learning, we pre-train MotionGPT with a mixture of motion-language data and fine-tune it on prompt-based question-and-answer tasks. Extensive experiments demonstrate that MotionGPT achieves state-of-the-art performances on multiple motion tasks including text-driven motion generation, motion captioning, motion prediction, and motion in-between.

MotionGPT: Human Motion as a Foreign Language paper page: Though the advancement of pre-trained large language models unfolds, the exploration of building a unified model for language and other multi-modal data, such as motion, remains challenging and untouched so far. Fortunately, human motion displays a semantic coupling akin to human language, often perceived as a form of body language. By fusing language data with large-scale motion models, motion-language pre-training that can enhance the performance of motion-related tasks becomes feasible. Driven by this insight, we propose MotionGPT, a unified, versatile, and user-friendly motion-language model to handle multiple motion-relevant tasks. Specifically, we employ the discrete vector quantization for human motion and transfer 3D motion into motion tokens, similar to the generation process of word tokens. Building upon this "motion vocabulary", we perform language modeling on both motion and text in a unified manner, treating human motion as a specific language. Moreover, inspired by prompt learning, we pre-train MotionGPT with a mixture of motion-language data and fine-tune it on prompt-based question-and-answer tasks. Extensive experiments demonstrate that MotionGPT achieves state-of-the-art performances on multiple motion tasks including text-driven motion generation, motion captioning, motion prediction, and motion in-between.

AK

125,319 views • 3 years ago

Meta presents Sapiens Foundation for Human Vision Models discuss: We present Sapiens, a family of models for four fundamental human-centric vision tasks - 2D pose estimation, body-part segmentation, depth estimation, and surface normal prediction. Our models natively support 1K high-resolution inference and are extremely easy to adapt for individual tasks by simply fine-tuning models pretrained on over 300 million in-the-wild human images. We observe that, given the same computational budget, self-supervised pretraining on a curated dataset of human images significantly boosts the performance for a diverse set of human-centric tasks. The resulting models exhibit remarkable generalization to in-the-wild data, even when labeled data is scarce or entirely synthetic. Our simple model design also brings scalability - model performance across tasks improves as we scale the number of parameters from 0.3 to 2 billion. Sapiens consistently surpasses existing baselines across various human-centric benchmarks. We achieve significant improvements over the prior state-of-the-art on Humans-5K (pose) by 7.6 mAP, Humans-2K (part-seg) by 17.1 mIoU, Hi4D (depth) by 22.4% relative RMSE, and THuman2 (normal) by 53.5% relative angular error.

Meta presents Sapiens Foundation for Human Vision Models discuss: We present Sapiens, a family of models for four fundamental human-centric vision tasks - 2D pose estimation, body-part segmentation, depth estimation, and surface normal prediction. Our models natively support 1K high-resolution inference and are extremely easy to adapt for individual tasks by simply fine-tuning models pretrained on over 300 million in-the-wild human images. We observe that, given the same computational budget, self-supervised pretraining on a curated dataset of human images significantly boosts the performance for a diverse set of human-centric tasks. The resulting models exhibit remarkable generalization to in-the-wild data, even when labeled data is scarce or entirely synthetic. Our simple model design also brings scalability - model performance across tasks improves as we scale the number of parameters from 0.3 to 2 billion. Sapiens consistently surpasses existing baselines across various human-centric benchmarks. We achieve significant improvements over the prior state-of-the-art on Humans-5K (pose) by 7.6 mAP, Humans-2K (part-seg) by 17.1 mIoU, Hi4D (depth) by 22.4% relative RMSE, and THuman2 (normal) by 53.5% relative angular error.

AK

151,511 views • 1 year ago

3D-LLM: Injecting the 3D World into Large Language Models paper page: Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so on. In this work, we propose to inject the 3D world into large language models and introduce a whole new family of 3D-LLMs. Specifically, 3D-LLMs can take 3D point clouds and their features as input and perform a diverse set of 3D-related tasks, including captioning, dense captioning, 3D question answering, task decomposition, 3D grounding, 3D-assisted dialog, navigation, and so on. Using three types of prompting mechanisms that we design, we are able to collect over 300k 3D-language data covering these tasks. To efficiently train 3D-LLMs, we first utilize a 3D feature extractor that obtains 3D features from rendered multi- view images. Then, we use 2D VLMs as our backbones to train our 3D-LLMs. By introducing a 3D localization mechanism, 3D-LLMs can better capture 3D spatial information. Experiments on ScanQA show that our model outperforms state-of-the-art baselines by a large margin (e.g., the BLEU-1 score surpasses state-of-the-art score by 9%). Furthermore, experiments on our held-in datasets for 3D captioning, task composition, and 3D-assisted dialogue show that our model outperforms 2D VLMs. Qualitative examples also show that our model could perform more tasks beyond the scope of existing LLMs and VLMs.

3D-LLM: Injecting the 3D World into Large Language Models paper page: Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so on. In this work, we propose to inject the 3D world into large language models and introduce a whole new family of 3D-LLMs. Specifically, 3D-LLMs can take 3D point clouds and their features as input and perform a diverse set of 3D-related tasks, including captioning, dense captioning, 3D question answering, task decomposition, 3D grounding, 3D-assisted dialog, navigation, and so on. Using three types of prompting mechanisms that we design, we are able to collect over 300k 3D-language data covering these tasks. To efficiently train 3D-LLMs, we first utilize a 3D feature extractor that obtains 3D features from rendered multi- view images. Then, we use 2D VLMs as our backbones to train our 3D-LLMs. By introducing a 3D localization mechanism, 3D-LLMs can better capture 3D spatial information. Experiments on ScanQA show that our model outperforms state-of-the-art baselines by a large margin (e.g., the BLEU-1 score surpasses state-of-the-art score by 9%). Furthermore, experiments on our held-in datasets for 3D captioning, task composition, and 3D-assisted dialogue show that our model outperforms 2D VLMs. Qualitative examples also show that our model could perform more tasks beyond the scope of existing LLMs and VLMs.

AK

249,572 views • 3 years ago

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

343,834 views • 2 years ago

JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal Language Models paper page: Achieving human-like planning and control with multimodal observations in an open world is a key milestone for more functional generalist agents. Existing approaches can handle certain long-horizon tasks in an open world. However, they still struggle when the number of open-world tasks could potentially be infinite and lack the capability to progressively enhance task completion as game time progresses. We introduce JARVIS-1, an open-world agent that can perceive multimodal input (visual observations and human instructions), generate sophisticated plans, and perform embodied control, all within the popular yet challenging open-world Minecraft universe. Specifically, we develop JARVIS-1 on top of pre-trained multimodal language models, which map visual observations and textual instructions to plans. The plans will be ultimately dispatched to the goal-conditioned controllers. We outfit JARVIS-1 with a multimodal memory, which facilitates planning using both pre-trained knowledge and its actual game survival experiences. In our experiments, JARVIS-1 exhibits nearly perfect performances across over 200 varying tasks from the Minecraft Universe Benchmark, ranging from entry to intermediate levels. JARVIS-1 has achieved a completion rate of 12.5% in the long-horizon diamond pickaxe task. This represents a significant increase up to 5 times compared to previous records. Furthermore, we show that JARVIS-1 is able to self-improve following a life-long learning paradigm thanks to multimodal memory, sparking a more general intelligence and improved autonomy.

JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal Language Models paper page: Achieving human-like planning and control with multimodal observations in an open world is a key milestone for more functional generalist agents. Existing approaches can handle certain long-horizon tasks in an open world. However, they still struggle when the number of open-world tasks could potentially be infinite and lack the capability to progressively enhance task completion as game time progresses. We introduce JARVIS-1, an open-world agent that can perceive multimodal input (visual observations and human instructions), generate sophisticated plans, and perform embodied control, all within the popular yet challenging open-world Minecraft universe. Specifically, we develop JARVIS-1 on top of pre-trained multimodal language models, which map visual observations and textual instructions to plans. The plans will be ultimately dispatched to the goal-conditioned controllers. We outfit JARVIS-1 with a multimodal memory, which facilitates planning using both pre-trained knowledge and its actual game survival experiences. In our experiments, JARVIS-1 exhibits nearly perfect performances across over 200 varying tasks from the Minecraft Universe Benchmark, ranging from entry to intermediate levels. JARVIS-1 has achieved a completion rate of 12.5% in the long-horizon diamond pickaxe task. This represents a significant increase up to 5 times compared to previous records. Furthermore, we show that JARVIS-1 is able to self-improve following a life-long learning paradigm thanks to multimodal memory, sparking a more general intelligence and improved autonomy.

AK

141,425 views • 2 years ago

Google presents AudioPaLM: A Large Language Model That Can Speak and Listen paper page: introduce AudioPaLM, a large language model for speech understanding and generation. AudioPaLM fuses text-based and speech-based language models, PaLM-2 [Anil et al., 2023] and AudioLM [Borsos et al., 2022], into a unified multimodal architecture that can process and generate text and speech with applications including speech recognition and speech-to-speech translation. AudioPaLM inherits the capability to preserve paralinguistic information such as speaker identity and intonation from AudioLM and the linguistic knowledge present only in text large language models such as PaLM-2. We demonstrate that initializing AudioPaLM with the weights of a text-only large language model improves speech processing, successfully leveraging the larger quantity of text training data used in pretraining to assist with the speech tasks. The resulting model significantly outperforms existing systems for speech translation tasks and has the ability to perform zero-shot speech-to-text translation for many languages for which input/target language combinations were not seen in training. AudioPaLM also demonstrates features of audio language models, such as transferring a voice across languages based on a short spoken prompt.

Google presents AudioPaLM: A Large Language Model That Can Speak and Listen paper page: introduce AudioPaLM, a large language model for speech understanding and generation. AudioPaLM fuses text-based and speech-based language models, PaLM-2 [Anil et al., 2023] and AudioLM [Borsos et al., 2022], into a unified multimodal architecture that can process and generate text and speech with applications including speech recognition and speech-to-speech translation. AudioPaLM inherits the capability to preserve paralinguistic information such as speaker identity and intonation from AudioLM and the linguistic knowledge present only in text large language models such as PaLM-2. We demonstrate that initializing AudioPaLM with the weights of a text-only large language model improves speech processing, successfully leveraging the larger quantity of text training data used in pretraining to assist with the speech tasks. The resulting model significantly outperforms existing systems for speech translation tasks and has the ability to perform zero-shot speech-to-text translation for many languages for which input/target language combinations were not seen in training. AudioPaLM also demonstrates features of audio language models, such as transferring a voice across languages based on a short spoken prompt.

AK

290,517 views • 3 years ago

3D-VLA A 3D Vision-Language-Action Generative World Model Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. Furthermore, they perform action prediction by learning a direct mapping from

3D-VLA A 3D Vision-Language-Action Generative World Model Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. Furthermore, they perform action prediction by learning a direct mapping from

AK

85,570 views • 2 years ago

The next evolution: VLA+ models Just yesterday Microsoft Research released Rho-alpha (ρα) – their first robotics model, built on the Phi family. While most Vision-Language-Action (VLA) models stop at vision and language, Rho-alpha adds: ▪️ Tactile sensing to feel objects during manipulation ▪️ Online learning that lets it improve from human corrections (via teleoperation, 3D mouse or other tools) in real-time even after deployment. Both these sides make adaptability central rather than incidental. Microsoft calls it a VLA+ model, positioning it as an extension beyond what current VLA systems support. ➡️ Today Rho-alpha can control dual-arm robot setups to perform tasks such as: • Manipulating the BusyBox following natural-language instructions • Plug insertion • Toolbox packing and object arrangement with bimanual coordination But to understand why this "plus" matters, we need to understand what came before. Here, we'll take you through the entire landscape of VLA models – Gemini Robotics, π0, SmolVLA, Helix, ACoT-VLA and others:

The next evolution: VLA+ models Just yesterday Microsoft Research released Rho-alpha (ρα) – their first robotics model, built on the Phi family. While most Vision-Language-Action (VLA) models stop at vision and language, Rho-alpha adds: ▪️ Tactile sensing to feel objects during manipulation ▪️ Online learning that lets it improve from human corrections (via teleoperation, 3D mouse or other tools) in real-time even after deployment. Both these sides make adaptability central rather than incidental. Microsoft calls it a VLA+ model, positioning it as an extension beyond what current VLA systems support. ➡️ Today Rho-alpha can control dual-arm robot setups to perform tasks such as: • Manipulating the BusyBox following natural-language instructions • Plug insertion • Toolbox packing and object arrangement with bimanual coordination But to understand why this "plus" matters, we need to understand what came before. Here, we'll take you through the entire landscape of VLA models – Gemini Robotics, π0, SmolVLA, Helix, ACoT-VLA and others:

Turing Post

62,362 views • 5 months ago

Vision-language models perform diverse tasks via in-context learning. Time for robots to do the same! Introducing In-Context Robot Transformer (ICRT): a robot policy that learns new tasks by prompting with robot trajectories, without any fine-tuning. [1/N]

Vision-language models perform diverse tasks via in-context learning. Time for robots to do the same! Introducing In-Context Robot Transformer (ICRT): a robot policy that learns new tasks by prompting with robot trajectories, without any fine-tuning. [1/N]

Max Fu

40,447 views • 1 year ago

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 views • 10 months ago

Introducing SIMA: the first generalist AI agent to follow natural-language instructions in a broad range of 3D virtual environments and video games. 🕹️ It can complete tasks similar to a human, and outperforms an agent trained in just one setting. 🧵

Introducing SIMA: the first generalist AI agent to follow natural-language instructions in a broad range of 3D virtual environments and video games. 🕹️ It can complete tasks similar to a human, and outperforms an agent trained in just one setting. 🧵

Google DeepMind

1,061,548 views • 2 years ago

We are releasing 4M-21 with a permissive license, including its source code and trained models. It's a pretty effective multimodal model that solves 10s of tasks & modalities. See the demo code, sample results, and the tokenizers of diverse modalities on the website. IMO, the multitask learning aspect of multimodal models has really taken a step forward. We can train a single model on many diverse tasks with ~SOTA accuracy. But a long way to go in terms of transfer/emergence. 🌐 ⌨️ Joint work w/ EPFL Apple.

We are releasing 4M-21 with a permissive license, including its source code and trained models. It's a pretty effective multimodal model that solves 10s of tasks & modalities. See the demo code, sample results, and the tokenizers of diverse modalities on the website. IMO, the multitask learning aspect of multimodal models has really taken a step forward. We can train a single model on many diverse tasks with ~SOTA accuracy. But a long way to go in terms of transfer/emergence. 🌐 ⌨️ Joint work w/ EPFL Apple.

Amir Zamir

69,539 views • 2 years ago

What happens when you combine voice, vision, and reasoning on-device? 🤔 Gemma 4 + a vision-language agent (VLA) running on NVIDIA Jetson Orin Nano shows how compact hardware can now handle real-world AI tasks using today’s open models—no cloud required. Get started:

What happens when you combine voice, vision, and reasoning on-device? 🤔 Gemma 4 + a vision-language agent (VLA) running on NVIDIA Jetson Orin Nano shows how compact hardware can now handle real-world AI tasks using today’s open models—no cloud required. Get started:

NVIDIA Robotics

25,655 views • 2 months ago

Introducing VL-JEPA: Vision-Language Joint Embedding Predictive Architecture for streaming, live action recognition, retrieval, VQA, and classification tasks with better performance and higher efficiency than large VLMs. • VL-JEPA is the first non-generative model that can perform general-domain vision-language tasks in real-time, built on a joint embedding predictive architecture. • We demonstrate in controlled experiments that VL-JEPA, trained with latent space embedding prediction, outperforms VLMs that rely on data space token prediction. • We show that VL-JEPA delivers significant efficiency gains over VLMs for online video streaming applications, thanks to its non-autoregressive design and native support for selective decoding. • We highlight that our VL-JEPA model, with an unified model architecture, can effectively handle a wide range of classification, retrieval, and VQA tasks at the same time. by Delong Chen (陈德龙) Mustafa Shukor Théo Moutakanni Willy Jade Lei Yu Tejaswi Kasarla Allen Bolourchi Yann LeCun Pascale Fung

Introducing VL-JEPA: Vision-Language Joint Embedding Predictive Architecture for streaming, live action recognition, retrieval, VQA, and classification tasks with better performance and higher efficiency than large VLMs. • VL-JEPA is the first non-generative model that can perform general-domain vision-language tasks in real-time, built on a joint embedding predictive architecture. • We demonstrate in controlled experiments that VL-JEPA, trained with latent space embedding prediction, outperforms VLMs that rely on data space token prediction. • We show that VL-JEPA delivers significant efficiency gains over VLMs for online video streaming applications, thanks to its non-autoregressive design and native support for selective decoding. • We highlight that our VL-JEPA model, with an unified model architecture, can effectively handle a wide range of classification, retrieval, and VQA tasks at the same time. by Delong Chen (陈德龙) Mustafa Shukor Théo Moutakanni Willy Jade Lei Yu Tejaswi Kasarla Allen Bolourchi Yann LeCun Pascale Fung

Pascale Fung

90,144 views • 6 months ago

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Khurram Javed

52,110 views • 7 months ago

Microsoft just dropped VITRA-VLA, a new Vision-Language-Action model for robotics on Hugging Face. It learns dexterous manipulation from over 1 million real-life human hand activity videos.

Microsoft just dropped VITRA-VLA, a new Vision-Language-Action model for robotics on Hugging Face. It learns dexterous manipulation from over 1 million real-life human hand activity videos.

DailyPapers

19,092 views • 6 months ago

Announcing the newest releases from Meta FAIR. We’re releasing new groundbreaking models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience. 1️⃣ Open Molecules 2025 (OMol25): A dataset for molecular discovery with simulations of large atomic systems. 2️⃣ Universal Model for Atoms: A machine learning interatomic potential for modeling atom interactions across a wide range of materials and molecules. 3️⃣ Adjoint Sampling: A scalable algorithm for training generative models based on scalar rewards. 4️⃣ FAIR and the Rothschild Foundation Hospital partnered on a large-scale study that reveals striking parallels between language development in humans and LLMs. Read more ➡️

Announcing the newest releases from Meta FAIR. We’re releasing new groundbreaking models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience. 1️⃣ Open Molecules 2025 (OMol25): A dataset for molecular discovery with simulations of large atomic systems. 2️⃣ Universal Model for Atoms: A machine learning interatomic potential for modeling atom interactions across a wide range of materials and molecules. 3️⃣ Adjoint Sampling: A scalable algorithm for training generative models based on scalar rewards. 4️⃣ FAIR and the Rothschild Foundation Hospital partnered on a large-scale study that reveals striking parallels between language development in humans and LLMs. Read more ➡️

AI at Meta

899,456 views • 1 year ago

Pretraining is essential for good performance on a wide variety of robotics tasks, and so most vision-language-action models build off of a vision language model (VLM) trained on a wide variety of image-language data. But how does the choice of VLM translate to downstream robotics performance? Jianke Zhang and @GYanjiang join us to talk about this key part of the robot policy, looking at a wide variety of different VLMs and how they perform. Interestingly, they see that performance on auxiliary tasks like quesiton answering did not lead to downstream improvements in control. To learn more, watch episode 65 of RoboPapers now, with Chris Paxton and Jiafei Duan!

Pretraining is essential for good performance on a wide variety of robotics tasks, and so most vision-language-action models build off of a vision language model (VLM) trained on a wide variety of image-language data. But how does the choice of VLM translate to downstream robotics performance? Jianke Zhang and @GYanjiang join us to talk about this key part of the robot policy, looking at a wide variety of different VLMs and how they perform. Interestingly, they see that performance on auxiliary tasks like quesiton answering did not lead to downstream improvements in control. To learn more, watch episode 65 of RoboPapers now, with Chris Paxton and Jiafei Duan!

RoboPapers

23,905 views • 4 months ago

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 views • 1 year ago

Prior to release, we shared a version of Cohere North Mini Code with AI engineers and answered some questions. Here's a quick illustrated walkthrough of the model's architecture and training process. Small models fill an important niche. They: 1. run on more widely available hardware 2. handle tasks within a certain range of complexity 3. take on sub-tasks that larger models delegate

Prior to release, we shared a version of Cohere North Mini Code with AI engineers and answered some questions. Here's a quick illustrated walkthrough of the model's architecture and training process. Small models fill an important niche. They: 1. run on more widely available hardware 2. handle tasks within a certain range of complexity 3. take on sub-tasks that larger models delegate

Jay Alammar

88,623 views • 26 days ago