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.

OpenClaw meets RL! Most agents evolve via prompt tricks and markdown hacks. MetaClaw updates actual model weights from every failed interaction. Everything happens on the fly, without any dataset or code changes. GitHub:

Akshay 🚀

276,654 subscribers

17,260 views • 3 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

NVIDIA + Stanford just dropped NitroGen "plays-any-game" AI trained on 40,000 hours of gameplay across 1,000+ games. They're releasing the dataset + model weights. 👇

NVIDIA + Stanford just dropped NitroGen "plays-any-game" AI trained on 40,000 hours of gameplay across 1,000+ games. They're releasing the dataset + model weights. 👇

Min Choi

315,151 views • 6 months ago

1/ Happy to share UniDisc - Unified Multimodal Discrete Diffusion – We train a 1.5 billion parameter transformer model from scratch on 250 million image/caption pairs using a **discrete diffusion objective**. Our model has all the benefits of diffusion models but now in multimodal space! - flexible compute-quality tradeoff, zero-shot inpainting and editing, better control via classifier-free guidance and lower latency! We open source everything - our code, weights and the training dataset.

1/ Happy to share UniDisc - Unified Multimodal Discrete Diffusion – We train a 1.5 billion parameter transformer model from scratch on 250 million image/caption pairs using a discrete diffusion objective. Our model has all the benefits of diffusion models but now in multimodal space! - flexible compute-quality tradeoff, zero-shot inpainting and editing, better control via classifier-free guidance and lower latency! We open source everything - our code, weights and the training dataset.

Mihir Prabhudesai

104,862 views • 1 year ago

You can now delegate tasks to GitHub Copilot coding agent from any page on GitHub 🤖 Open the new Agents panel in one click, write a simple prompt, then hit Enter. GitHub Copilot works in the background, and opens a PR for your review. No interruptions to your workflow required. ✅

You can now delegate tasks to GitHub Copilot coding agent from any page on GitHub 🤖 Open the new Agents panel in one click, write a simple prompt, then hit Enter. GitHub Copilot works in the background, and opens a PR for your review. No interruptions to your workflow required. ✅

GitHub

112,727 views • 10 months ago

🚀 Exciting news! The #GeoAI Python package now lets you train land cover classification models with just one line of code. Leverage any PyTorch segmentation model from — with hundreds of image encoders & pretrained weights available. 📍 GitHub: 📓 Notebook: #GeoAI #Geospatial #DeepLearning

🚀 Exciting news! The #GeoAI Python package now lets you train land cover classification models with just one line of code. Leverage any PyTorch segmentation model from — with hundreds of image encoders & pretrained weights available. 📍 GitHub: 📓 Notebook: #GeoAI #Geospatial #DeepLearning

Qiusheng Wu

16,150 views • 1 year ago

UnifAI Skills are now live on OpenClaw!! Traders and Developers can now integrate UnifAI-powered skills directly into their agents, unlocking more powerful and composable capabilities within the OpenClaw ecosystem. UnifAI skills are automatically pre-installed for users who deploy their OpenClaw🦞 AI Agents with UniClaw Build smarter agents. Ship faster. 🤖 GitHub repo -

UnifAI Skills are now live on OpenClaw!! Traders and Developers can now integrate UnifAI-powered skills directly into their agents, unlocking more powerful and composable capabilities within the OpenClaw ecosystem. UnifAI skills are automatically pre-installed for users who deploy their OpenClaw🦞 AI Agents with UniClaw Build smarter agents. Ship faster. 🤖 GitHub repo -

UnifAI Network

102,987 views • 3 months ago

Introducing Matrix I crawled 100,000+ agents, skills and tools to train a new model which can answer what capabilities are the best match for a task. Think Google, but for agents. A living model that learns from the gossiping network, and gets smarter with every interaction.

Introducing Matrix I crawled 100,000+ agents, skills and tools to train a new model which can answer what capabilities are the best match for a task. Think Google, but for agents. A living model that learns from the gossiping network, and gets smarter with every interaction.

Varun

897,719 views • 3 months ago

If your agents aren’t texting yet… you’re already behind. Agents can now text any number directly from OpenClaw. And yes… they’re faster than you 😂 Sound On 🔊

If your agents aren’t texting yet… you’re already behind. Agents can now text any number directly from OpenClaw. And yes… they’re faster than you 😂 Sound On 🔊

Luke The Dev

18,765 views • 3 months ago

We built the best codebase search to give any agent grounded information from GitHub repos. Introducing Sandbox Search. Point it at any repo, and we’ll spin up a secure coding agent in its own sandbox to do research for you. Use inside claude code, openclaw, cursor, and more.

We built the best codebase search to give any agent grounded information from GitHub repos. Introducing Sandbox Search. Point it at any repo, and we’ll spin up a secure coding agent in its own sandbox to do research for you. Use inside claude code, openclaw, cursor, and more.

Arlan

12,412 views • 2 months ago

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL Optimal Advantage-based Policy Optimization with Lagged Inference policy (OAPL) shows you don’t need strict on-policy training to improve reasoning. It matches or beats Group Relative Policy Optimization (GRPO), stays stable with large policy lag, and uses ~3× fewer training generations. For Databricks customers, it’s a simpler, practical, and equally powerful approach to RL that Databricks is pioneering internally — and bringing directly to Databricks customers, so enterprises can improve agents using the same methods we use for our in-house agents, without complex infrastructure changes.

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL Optimal Advantage-based Policy Optimization with Lagged Inference policy (OAPL) shows you don’t need strict on-policy training to improve reasoning. It matches or beats Group Relative Policy Optimization (GRPO), stays stable with large policy lag, and uses ~3× fewer training generations. For Databricks customers, it’s a simpler, practical, and equally powerful approach to RL that Databricks is pioneering internally — and bringing directly to Databricks customers, so enterprises can improve agents using the same methods we use for our in-house agents, without complex infrastructure changes.

Databricks AI Research

12,439 views • 3 months ago

The hardcover book of GPT-1’s weights that Claude Code designed, produced, and sold (including the cool cover which visualizes the numbers in the volume) actually came in the mail today and it looks really nice. I never touched any code or did any design or any API to make this.

The hardcover book of GPT-1’s weights that Claude Code designed, produced, and sold (including the cool cover which visualizes the numbers in the volume) actually came in the mail today and it looks really nice. I never touched any code or did any design or any API to make this.

Ethan Mollick

475,843 views • 4 months ago

One of many agents coming your way across the GitHub platform: Copilot Code Review. 🤖 Copilot gives direct feedback on your PR – summarizing changes, catching bugs, suggesting tests, and fixing typos – while you wait for a human review 🤝

One of many agents coming your way across the GitHub platform: Copilot Code Review. 🤖 Copilot gives direct feedback on your PR – summarizing changes, catching bugs, suggesting tests, and fixing typos – while you wait for a human review 🤝

Thomas Dohmke

77,745 views • 1 year ago

Claude Code ships with a built-in skill for working with the Claude Platform. Useful for model migrations, using API features (e.g., prompt caching), or onboarding to newer APIs like Claude Managed Agents.

Claude Code ships with a built-in skill for working with the Claude Platform. Useful for model migrations, using API features (e.g., prompt caching), or onboarding to newer APIs like Claude Managed Agents.

ClaudeDevs

312,527 views • 1 month ago

We have released the code and weights for our #CVPR2023 paper "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model"! code: abs: project: The demo is below:

We have released the code and weights for our #CVPR2023 paper "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model"! code: abs: project: The demo is below:

Artsiom Sanakoyeu

35,710 views • 3 years ago

AI just killed the green screen. Rendora builds entire movie scenes, voices, and characters from a single prompt. This changes everything 👇

AI just killed the green screen. Rendora builds entire movie scenes, voices, and characters from a single prompt. This changes everything 👇

Kelsey

81,316 views • 8 months ago

You won't believe it, but this AIApply video was made by Claude... ✨ One-shot prompt ✨ Claude code + Github + Remotion ✨ Taking elements straight from our codebase Imagine a world where a 1-person growth team can automate and deploy product content on any channel just by linking GitHub to LLMs...

You won't believe it, but this AIApply video was made by Claude... ✨ One-shot prompt ✨ Claude code + Github + Remotion ✨ Taking elements straight from our codebase Imagine a world where a 1-person growth team can automate and deploy product content on any channel just by linking GitHub to LLMs...

Yoann Pavy

78,499 views • 4 months ago

Businesses, you can now* accept payments directly from agents—without, or with, a human in the loop—using Stripe machine payments. It works for both cards and stablecoins via Machine Payments Protocol or x402. Add to your existing integration in a single prompt. *No more waitlist, get started =>

Businesses, you can now* accept payments directly from agents—without, or with, a human in the loop—using Stripe machine payments. It works for both cards and stablecoins via Machine Payments Protocol or x402. Add to your existing integration in a single prompt. *No more waitlist, get started =>

Jeff Weinstein

12,398 views • 1 month ago

This founder has 16 OpenClaw🦞 agents for $400/month. His entire company operates while he sleeps. In the video: • How he structured 16 sub-agents under one OpenClaw instance • The cron job system running research, content, code reviews, and outreach 24/7 • Routing cheap LLMs for simple tasks, expensive models for coding •The Reddit workflow that drove 450+ users to his SaaS with zero ad spend The founder Vadim has the most complete OpenClaw setup I've seen from a solo founder. Full video is live on YouTube. Comment "Agents" and I'll DM you the link.

This founder has 16 OpenClaw🦞 agents for $400/month. His entire company operates while he sleeps. In the video: • How he structured 16 sub-agents under one OpenClaw instance • The cron job system running research, content, code reviews, and outreach 24/7 • Routing cheap LLMs for simple tasks, expensive models for coding •The Reddit workflow that drove 450+ users to his SaaS with zero ad spend The founder Vadim has the most complete OpenClaw setup I've seen from a solo founder. Full video is live on YouTube. Comment "Agents" and I'll DM you the link.

Jacob Klug

87,710 views • 3 months ago

this is the best line with barely any actual tricks ever

this is the best line with barely any actual tricks ever

Black Phaze

12,365 views • 1 year ago

This is wild. Higgsfield Games lets you build and deploy a full multiplayer game from a single prompt. Any genre, 2D or 3D, with characters, props, and worlds generated on the fly. Powered by Claude Fable 5. Huge W for indie devs.

This is wild. Higgsfield Games lets you build and deploy a full multiplayer game from a single prompt. Any genre, 2D or 3D, with characters, props, and worlds generated on the fly. Powered by Claude Fable 5. Huge W for indie devs.

How To Prompt

61,144 views • 11 days ago