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。

Full run and a segment of a task. 100% from 16 examples. To play with continuous learning, your base model needs to be data-efficient and stable, which we tested here, because any irrelevant fluctuations can compound over time.

Igor Kulakov (MicroFactory)

9,377 subscribers

24,486 次观看 • 1 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

Introducing Compound Skills 🪄 You can now train the AI analyst in your way of working, just like you would a human. Ask Compound to create a Skill for any task (e.g. DCFs, waterfall modeling). Compound remembers your preferences and will apply it the next time it works on a similar task. You can also ask Compound to update skills, if you’ve provided more feedback during a task, or simply if you want to iterate on a skill. With Skills, Compound has become more powerful than ever - continually learning and adapting to your preferences. Try it now today at P.S For new users, get a one-time signup bonus of 40K free credits. You can also refer a friend (check settings for a link), and you’ll both get 10K free credits.

Introducing Compound Skills 🪄 You can now train the AI analyst in your way of working, just like you would a human. Ask Compound to create a Skill for any task (e.g. DCFs, waterfall modeling). Compound remembers your preferences and will apply it the next time it works on a similar task. You can also ask Compound to update skills, if you’ve provided more feedback during a task, or simply if you want to iterate on a skill. With Skills, Compound has become more powerful than ever - continually learning and adapting to your preferences. Try it now today at P.S For new users, get a one-time signup bonus of 40K free credits. You can also refer a friend (check settings for a link), and you’ll both get 10K free credits.

Compound AI

16,746 次观看 • 7 个月前

Learn to train an LLM with distributed data while ensuring privacy using federated learning in a new two-part short course, Intro to Federated Learning and Federated Fine-tuning of LLMs with Private Data, created with Flower and taught by Daniel J. Beutel and Nicholas Lane. Federated learning allows a single model to be trained across multiple devices, such as phones, or multiple organizations, such as hospitals, without the need to share data to a central server. This two-part course gives you an introduction to federated learning, and then teaches you how to fine-tune your large language model with distributed data using Flower Lab’s open source federated learning framework. You’ll learn: - How to use federated learning to train a variety of models, ranging from speech and vision models to LLMs, across distributed data while offering data privacy options to users and organizations. - Privacy Enhancing Technologies like differential privacy (DP), which obscures individual data by adding calibrated noise to query results. - Two variants of differential privacy - Central and Local - and how to choose depending on your use case. - How to measure and decrease bandwidth usage to make federated learning more practical and efficient with techniques like using pre-trained models and Parameter-Efficient Fine-Tuning - How federated LLM fine-tuning reduces the risk of leaking training data. Sign up here!

Learn to train an LLM with distributed data while ensuring privacy using federated learning in a new two-part short course, Intro to Federated Learning and Federated Fine-tuning of LLMs with Private Data, created with Flower and taught by Daniel J. Beutel and Nicholas Lane. Federated learning allows a single model to be trained across multiple devices, such as phones, or multiple organizations, such as hospitals, without the need to share data to a central server. This two-part course gives you an introduction to federated learning, and then teaches you how to fine-tune your large language model with distributed data using Flower Lab’s open source federated learning framework. You’ll learn: - How to use federated learning to train a variety of models, ranging from speech and vision models to LLMs, across distributed data while offering data privacy options to users and organizations. - Privacy Enhancing Technologies like differential privacy (DP), which obscures individual data by adding calibrated noise to query results. - Two variants of differential privacy - Central and Local - and how to choose depending on your use case. - How to measure and decrease bandwidth usage to make federated learning more practical and efficient with techniques like using pre-trained models and Parameter-Efficient Fine-Tuning - How federated LLM fine-tuning reduces the risk of leaking training data. Sign up here!

Andrew Ng

64,609 次观看 • 2 年前

ChatGPT connecting to external data sources offers a glimpse into what AI interoperability looks like. Ask any question in ChatGPT and it can now pull data from Box. The future of AI is AI Agents that can access your systems and data to complete any task for you.

ChatGPT connecting to external data sources offers a glimpse into what AI interoperability looks like. Ask any question in ChatGPT and it can now pull data from Box. The future of AI is AI Agents that can access your systems and data to complete any task for you.

Aaron Levie

57,745 次观看 • 1 年前

China is a model of minority cultural and linguistic preservation. We need to be learning from China.

China is a model of minority cultural and linguistic preservation. We need to be learning from China.

Jason Smith - 上官杰文

28,238 次观看 • 3 个月前

Imitation learning is great, but needs us to have (near) optimal data. We throw away most other data (failures, evaluation data, suboptimal data, undirected play data), even though this data can be really useful and way cheaper! In our new work - RISE, we show a simple way to *use all of this non-optimal data to robustify imitation learning* with minimal requirements beyond BC. Key idea: use non-expert data to learn how to *recover* back to expert data with a minimal frills offline RL that works under sparse data coverage. Allows usage of *all* available data, not just expert data - never throw your data away! Paper: Website: A 🧵(1/10)

Imitation learning is great, but needs us to have (near) optimal data. We throw away most other data (failures, evaluation data, suboptimal data, undirected play data), even though this data can be really useful and way cheaper! In our new work - RISE, we show a simple way to use all of this non-optimal data to robustify imitation learning with minimal requirements beyond BC. Key idea: use non-expert data to learn how to recover back to expert data with a minimal frills offline RL that works under sparse data coverage. Allows usage of all available data, not just expert data - never throw your data away! Paper: Website: A 🧵(1/10)

Abhishek Gupta

20,635 次观看 • 10 个月前

What happens when you change the hand mid-task? We tested this by modifying the hands mid-rollout and letting the same model keep running. It perceives the new tool, and finds a new trajectory and contact strategy to complete the task. This works because training on mixed data forces the model to condition its behavior on the hand in front of it. This brings it closer to a general understanding of how shapes and contact surfaces interact with the physical world – and which actuation strategy should follow.

What happens when you change the hand mid-task? We tested this by modifying the hands mid-rollout and letting the same model keep running. It perceives the new tool, and finds a new trajectory and contact strategy to complete the task. This works because training on mixed data forces the model to condition its behavior on the hand in front of it. This brings it closer to a general understanding of how shapes and contact surfaces interact with the physical world – and which actuation strategy should follow.

Generalist

21,929 次观看 • 1 个月前

We've created a comprehensive Retrieval Harness for modern agentic retrieval in 2026. The harness provides a persistent data pipeline that can connect to a data source, index and update a large knowledge base, and expose a broad set of tools akin to filesystem operations (semantic/keyword search, regex grep, file search, read). You can plug this into any of your agents to let them autonomously crawl an arbitrary knowledge base to solve a task with any complexity. Check out our reference implementation: LlamaParse:

We've created a comprehensive Retrieval Harness for modern agentic retrieval in 2026. The harness provides a persistent data pipeline that can connect to a data source, index and update a large knowledge base, and expose a broad set of tools akin to filesystem operations (semantic/keyword search, regex grep, file search, read). You can plug this into any of your agents to let them autonomously crawl an arbitrary knowledge base to solve a task with any complexity. Check out our reference implementation: LlamaParse:

Jerry Liu

57,044 次观看 • 2 个月前

Announcing RoboFuME🤖💨, a system for autonomous & efficient real-world robot learning that 1. pre-trains a VLM reward model and a multi-task RL policy from diverse off-the-shelf demo data; 2. runs RL fine-tuning online with the VLM reward model. 🔗 🧵↓

Announcing RoboFuME🤖💨, a system for autonomous & efficient real-world robot learning that 1. pre-trains a VLM reward model and a multi-task RL policy from diverse off-the-shelf demo data; 2. runs RL fine-tuning online with the VLM reward model. 🔗 🧵↓

Jingyun Yang

88,294 次观看 • 2 年前

Announcing our new and stable release of Segment Anything for Microscopy! We introduce new features and improve the tool to segment microscopy data with a few clicks, like the mitochondria in the video, first segmented automatically and then corrected interactively.

Announcing our new and stable release of Segment Anything for Microscopy! We introduce new features and improve the tool to segment microscopy data with a few clicks, like the mitochondria in the video, first segmented automatically and then corrected interactively.

Constantin Pape

45,380 次观看 • 2 年前

Big News! Meta just released Segment Anything, a new AI model that can "cut out" any object, in any image/video, with a single click. The model is designed and trained to be promptable, so it can transfer zero-shot to new image distributions and tasks.

Big News! Meta just released Segment Anything, a new AI model that can "cut out" any object, in any image/video, with a single click. The model is designed and trained to be promptable, so it can transfer zero-shot to new image distributions and tasks.

Lior Alexander

290,330 次观看 • 3 年前

How can we build a general-purpose “foundation model” for robot motion? Zhengyi “Zen” Luo joins us to talk about SONIC, which uses motion tracking as a foundational task for humanoid robot control, and scales humanoid control training to 9k GPU hours and 100 million frames worth of data. The result: a model with a generally-useful embedding space that can be controlled by a VLA, or from human video, to perform a wide variety of humanoid whole-body-control tasks, including with zero-shot transfer to previously unseen motions. Watch episode 72 of RoboPapers, with Michael Cho - Rbt/Acc and Jiafei Duan, now!

How can we build a general-purpose “foundation model” for robot motion? Zhengyi “Zen” Luo joins us to talk about SONIC, which uses motion tracking as a foundational task for humanoid robot control, and scales humanoid control training to 9k GPU hours and 100 million frames worth of data. The result: a model with a generally-useful embedding space that can be controlled by a VLA, or from human video, to perform a wide variety of humanoid whole-body-control tasks, including with zero-shot transfer to previously unseen motions. Watch episode 72 of RoboPapers, with Michael Cho - Rbt/Acc and Jiafei Duan, now!

RoboPapers

24,802 次观看 • 4 个月前

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

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

RoboPapers

38,772 次观看 • 2 个月前

Yesterday at Lugano Plan ₿ I presented a preview of Tether's Local AI development kit, privacy focused, powered by peer-to-peer technology 🍐🍐 Write the code once, and run everywhere, from a 40$ phone, to a flagship smartphone, from a laptop to a mainframe with many H100, from a smart fridge to a car entertainment system. Based on 🐻 Terher's AI SDK is extremely modular, it uses supports any model (Marian, Llama, ...) and loads/stores weights and fine-tuning data in P2P data structures. Open source once battle tested. Demo of a sample local and 100% private translation app working on any device 👇

Yesterday at Lugano Plan ₿ I presented a preview of Tether's Local AI development kit, privacy focused, powered by peer-to-peer technology 🍐🍐 Write the code once, and run everywhere, from a 40$ phone, to a flagship smartphone, from a laptop to a mainframe with many H100, from a smart fridge to a car entertainment system. Based on 🐻 Terher's AI SDK is extremely modular, it uses supports any model (Marian, Llama, ...) and loads/stores weights and fine-tuning data in P2P data structures. Open source once battle tested. Demo of a sample local and 100% private translation app working on any device 👇

Paolo Ardoino 🤖

127,950 次观看 • 1 年前

How can you solve complex tasks using a Large Language Model? Here is a 2-minute introduction to everything you need to know to 10x the quality of your results. Let's talk about three techniques, in order of complexity, starting with the easiest one: • In-Context Learning • Indexing + In-Context Learning • Fine-tuning In-Context Learning The team that trained GPT-3 found something they couldn't explain: You can condition a model using examples of how you want it to behave. I included an example prompt in the attached video. You can "teach" the model how you want it to interpret questions, select the correct answers, and format the results by giving a few examples. You can also give specific knowledge to the model that will be helpful when formulating answers. We call this approach "grounding the model." There's another example in the video. Indexing + In-Context Learning Unfortunately, there is a limit to how much data you can include in a prompt. We call this the "context size." One version of GPT-4 supports a context of approximately 6,000 words, while the other supports 25,000 words. Although this sounds like a lot, many applications need more than that. Imagine you wrote a book and want to build an application to answer any questions about your story. What happens if your book is longer than the context? That's where Indexing comes in. Using a model, you can turn every book passage into an embedding. These are vectors, numbers that "encode" the passage's text. You can then store these embeddings in a particular database that supports fast retrieval of these vectors. You can then turn any question into an embedding and search the database for the list of passages that are similar to that query. Instead of using the entire book to ask the model, you can now use the relevant passages as in-context information, effectively working around the context size limitation. Fine-tuning Fine-tuning can give you an extra boost to get reliable outputs from your LLM. It is, however, the most complex approach on the list. There are different approaches to fine-tuning a model with your data. A popular technique is to process your data with your LLM and use the outputs to train a new classifier that solves your specific task. Notice that here you aren't modifying the LLM. Instead, you are chaining it with your trained classifier. Another approach is to modify the parameters of the LLM using your data. Think of this as "rewiring" the model in a way that solves your particular task. The results and costs will vary depending on how many layers you want to fine-tune from the original model. Many companies think that fine-tuning is the solution to their problems. In my experience, many will benefit from exploring the other two approaches. I love explaining Machine Learning and Artificial Intelligence ideas. If you enjoy in-depth content like this, follow me Santiago so you don't miss what comes next.

How can you solve complex tasks using a Large Language Model? Here is a 2-minute introduction to everything you need to know to 10x the quality of your results. Let's talk about three techniques, in order of complexity, starting with the easiest one: • In-Context Learning • Indexing + In-Context Learning • Fine-tuning In-Context Learning The team that trained GPT-3 found something they couldn't explain: You can condition a model using examples of how you want it to behave. I included an example prompt in the attached video. You can "teach" the model how you want it to interpret questions, select the correct answers, and format the results by giving a few examples. You can also give specific knowledge to the model that will be helpful when formulating answers. We call this approach "grounding the model." There's another example in the video. Indexing + In-Context Learning Unfortunately, there is a limit to how much data you can include in a prompt. We call this the "context size." One version of GPT-4 supports a context of approximately 6,000 words, while the other supports 25,000 words. Although this sounds like a lot, many applications need more than that. Imagine you wrote a book and want to build an application to answer any questions about your story. What happens if your book is longer than the context? That's where Indexing comes in. Using a model, you can turn every book passage into an embedding. These are vectors, numbers that "encode" the passage's text. You can then store these embeddings in a particular database that supports fast retrieval of these vectors. You can then turn any question into an embedding and search the database for the list of passages that are similar to that query. Instead of using the entire book to ask the model, you can now use the relevant passages as in-context information, effectively working around the context size limitation. Fine-tuning Fine-tuning can give you an extra boost to get reliable outputs from your LLM. It is, however, the most complex approach on the list. There are different approaches to fine-tuning a model with your data. A popular technique is to process your data with your LLM and use the outputs to train a new classifier that solves your specific task. Notice that here you aren't modifying the LLM. Instead, you are chaining it with your trained classifier. Another approach is to modify the parameters of the LLM using your data. Think of this as "rewiring" the model in a way that solves your particular task. The results and costs will vary depending on how many layers you want to fine-tune from the original model. Many companies think that fine-tuning is the solution to their problems. In my experience, many will benefit from exploring the other two approaches. I love explaining Machine Learning and Artificial Intelligence ideas. If you enjoy in-depth content like this, follow me Santiago so you don't miss what comes next.

Santiago

384,510 次观看 • 3 年前

You really can just do things! Use *any* Hugging Face space as a MCP server along with your Local Models! 🔥 Here in we use Qwen 3 30B A3B with ggml llama.cpp and Hugging Face tiny agents to create images via FLUX powered by ZeroGPU ⚡ It's quite a bit crazy to see local models be capable of so much and just be able to understand/ infer from tool description! There's a lot of potential here in automating video generation workflows, content curation and a lot more.. Bonus: you can plug any other Inference Provider if you don't want to run locally too! npx @ huggingface/tiny-agents run [TASK] oh, and we provide both typescript and python client! 🐐

You really can just do things! Use any Hugging Face space as a MCP server along with your Local Models! 🔥 Here in we use Qwen 3 30B A3B with ggml llama.cpp and Hugging Face tiny agents to create images via FLUX powered by ZeroGPU ⚡ It's quite a bit crazy to see local models be capable of so much and just be able to understand/ infer from tool description! There's a lot of potential here in automating video generation workflows, content curation and a lot more.. Bonus: you can plug any other Inference Provider if you don't want to run locally too! npx @ huggingface/tiny-agents run [TASK] oh, and we provide both typescript and python client! 🐐

Vaibhav (VB) Srivastav

39,113 次观看 • 1 年前

General Intuition CEO Pim de Witte, who's building foundation models trained on video game controller input data ("action-labeled gameplay clips"), says general intelligence won't "taste like an LLM": "We have a scale of data that's going to allow us to jump to the frontier in one capability — which is any system that can be controlled with a game controller (which is most robots) — and then, you can use that to create a sufficiently general intelligence." "As humans, the decision to talk or type is a very, very small subset of the actions that we can actually take." "So in order to create a sufficiently general intelligence to play 10,000+ video games, the model has to be able to predict across the entire action space of human cognition when they're interacting with these environments. Which are 2D and 3D environments, interfaces, long-horizon tasks, short-horizon tasks, [etc.]." "It has to be a sufficiently general intelligence in order to predict actions. Therefore, the type of model you get out is not going to taste like an LLM. This model is going to be incredibly good at navigating unforeseen environments. It's going to be incredibly good at zero-shotting any task that can be done with a game controller."

General Intuition CEO Pim de Witte, who's building foundation models trained on video game controller input data ("action-labeled gameplay clips"), says general intelligence won't "taste like an LLM": "We have a scale of data that's going to allow us to jump to the frontier in one capability — which is any system that can be controlled with a game controller (which is most robots) — and then, you can use that to create a sufficiently general intelligence." "As humans, the decision to talk or type is a very, very small subset of the actions that we can actually take." "So in order to create a sufficiently general intelligence to play 10,000+ video games, the model has to be able to predict across the entire action space of human cognition when they're interacting with these environments. Which are 2D and 3D environments, interfaces, long-horizon tasks, short-horizon tasks, [etc.]." "It has to be a sufficiently general intelligence in order to predict actions. Therefore, the type of model you get out is not going to taste like an LLM. This model is going to be incredibly good at navigating unforeseen environments. It's going to be incredibly good at zero-shotting any task that can be done with a game controller."

TBPN

80,838 次观看 • 2 个月前

1/ We are speed running industrial robotics. It took us just 19 days from the first day of data collection to filming a 2.5-hour continuous run of our model autonomously breaking down industrial containers — zero human intervention. The data efficiency of our DVA model is fundamentally changing how fast we bring robots out of the lab and into the factory. Autonomous operation with 3 hours of data collection at a customer factory.

1/ We are speed running industrial robotics. It took us just 19 days from the first day of data collection to filming a 2.5-hour continuous run of our model autonomously breaking down industrial containers — zero human intervention. The data efficiency of our DVA model is fundamentally changing how fast we bring robots out of the lab and into the factory. Autonomous operation with 3 hours of data collection at a customer factory.

Rhoda AI

25,764 次观看 • 5 个月前

Did you know that Facebook made $134.9 billion in 2023 from selling your data? Our society has come to the point where we accepted an unfair model in which privacy is being exploited and data gifted to big tech that is selling it for billions. We intend to break the norm of how data is shared on the internet because it is outdated and unfair. You should be empowered with more control and monetization models for your data!

Did you know that Facebook made $134.9 billion in 2023 from selling your data? Our society has come to the point where we accepted an unfair model in which privacy is being exploited and data gifted to big tech that is selling it for billions. We intend to break the norm of how data is shared on the internet because it is outdated and unfair. You should be empowered with more control and monetization models for your data!

Solana ID 🪷

30,458 次观看 • 2 年前

Sports like tennis are great examples of the sort of dynamic whole-body interaction that’s possible with humanoid robots. But capturing examples of fast, dynamic interactions from humans is really difficult. Enter LATENT, which uses lower-quality human data plus reinforcement learning to teach a robot to play tennis, able to complete back-and-forth volleys at a human level. LATENT has three steps: (1) collecting imperfect human data like a backswing, (2) using these to learn a latent action space, and (3) they train a high-level policy in simulation which can compose these actions and execute tennis skills on a robot. Haofei Lu and Yunrui Lian join us to tell us about their method. Watch Episode #80 of RoboPapers, with Chris Paxton and Jiafei Duan, now to learn more!

Sports like tennis are great examples of the sort of dynamic whole-body interaction that’s possible with humanoid robots. But capturing examples of fast, dynamic interactions from humans is really difficult. Enter LATENT, which uses lower-quality human data plus reinforcement learning to teach a robot to play tennis, able to complete back-and-forth volleys at a human level. LATENT has three steps: (1) collecting imperfect human data like a backswing, (2) using these to learn a latent action space, and (3) they train a high-level policy in simulation which can compose these actions and execute tennis skills on a robot. Haofei Lu and Yunrui Lian join us to tell us about their method. Watch Episode #80 of RoboPapers, with Chris Paxton and Jiafei Duan, now to learn more!

RoboPapers

30,067 次观看 • 3 个月前