🚀Introducing Hierarchical Reasoning Model🧠🤖 Inspired by brain's hierarchical processing, HRM delivers unprecedented reasoning power on complex tasks like ARC-AGI and expert-level Sudoku using just 1k examples, no pretraining or CoT! Unlock next AI breakthrough with

Just reread the tidybot2.github.io docs today, what an incredible tutorial for building a robot system. Honestly, you could set up an entire new robot lab just by following it, Jimmy Wu even gave you the link of the screwdriver he used 😂

🚀 Qwen3-30B-A3B Small Update: Smarter, faster, and local deployment-friendly. ✨ Key Enhancements: ✅ Enhanced reasoning, coding, and math skills ✅ Broader multilingual knowledge ✅ Improved long-context understanding (up to 256K tokens) ✅ Better alignment with user intent

Qwen Thanks for releasing a smaller model guys! 🥰 You can now run Qwen3-30B-A3B-0527 using Dynamic GGUFs. Only 33GB RAM or unified mem is needed to run the full 8-bit precision model at >6 tokens/s. huggingface.co/unsloth/Qwen3-…

Modern AI is confined to the digital world. At Skild AI, we are building towards AGI for the real world, unconstrained by robot type or task — a single, omni-bodied brain. Today, we are sharing our journey, starting with early milestones, with more to come in the weeks ahead.

Qwen just released Qwen3-30B-A3B-Instruct-2507 🔥 an upgrade to the non-thinking mode model huggingface.co/Qwen/Qwen3-30B… ✨ 30B MoE / 3.3B active - Apache 2.0 ✨ Strong gains in reasoning, math, coding, & multilingual tasks ✨ Native support for 256K long-context inputs

ollama run qwen3:30b

Qwen has just released a model on par with GPT-4o... And you can run it locally easily 🤯 Yep. GPT-4o level AI running offline on a laptop. - Fully open source - Only 3B active parameters - 262k context length natively Quick steps to run it on your machine and details below

just a quick update: we managed to replicate these claims Guan's system hits 25% on ARC-AGI 1 with 50 GPU hours I still couldn't audit the code personally though, but, unless cheating somehow, this approach seems to generalize ARC-AGI instances with relatively little compute

Ex-Alibaba CTO just made the boldest claim about AI & global power: “China is building the future of AI, not Silicon Valley.” He also revealed why AI by 2030 will look nothing like ChatGPT and how China’s approach is already decades ahead. Here are my top 7 takeaways: 🧵

Maya and Miles are learning new languages! This week, we deployed a new voice model to our web demo with improved support for Spanish, French, German, Italian, Chinese, Japanese, and Korean. Give it a try and let us know what you think: sesame.com/voicedemo

So, I did some coding this week... - Qwen3 Coder Flash (30B-A3B) - Mixture-of-Experts setup with 128 experts, 8 active per token - In pure PyTorch (optimized for human readability) - in a standalone Jupyter notebook - Runs on a single A100

I think everyone interested in AI should read the model cards for the frontier models, especially the safety sections, which give you a sense of immediate concerns: Gemini Deep Think: storage.googleapis.com/deepmind-media… Claude 4: www-cdn.anthropic.com/07b2a3f9902ee1… Grok: ???? o3: cdn.openai.com/pdf/2221c875-0…

🚀 𝐄𝐱𝐜𝐢𝐭𝐢𝐧𝐠 𝐍𝐞𝐰𝐬! 𝐏𝐲𝐏𝐈 𝐚𝐧𝐝 𝐆𝐢𝐭𝐇𝐮𝐛 𝐑𝐞𝐥𝐞𝐚𝐬𝐞 𝐟𝐨𝐫 𝐋𝐞𝐑𝐨𝐛𝐨𝐭 🚀 We’re thrilled to announce that we have reached a new major milestone, the first 𝐨𝐟𝐟𝐢𝐜𝐢𝐚𝐥 𝐏𝐲𝐏𝐈 and 𝐆𝐢𝐭𝐇𝐮𝐛 𝐑𝐞𝐥𝐞𝐚𝐬𝐞! 🎉 * You can now install LeRobot with a

Every robot you see is a data firehose generating terabytes of chaos. This hidden crisis is the #1 reason robots fail, and it's costing the industry billions. You see hardware, but not the data swamp drowning engineers. In 2025, a quiet revolution is fixing it. Here’s how. 🧵

OpenMind raised $20M for building OM1, a software layer for humanoids that acts as their OS The company also unveiled FABRIC, a protocol that allows robots to verify identity and share context and information with other robots

EngineAI released a clip highlighting RL-based walking, running, and standing up movements of its PM01 humanoid The company also teased SA02, its budget-focused humanoid with 26DoF that will start at $5,400, taking on Unitree's new R1

At the World Robot Conference 2025, UBTech is showing its new and updated humanoids, including Walker S2, Cruzr S2, Walker S1, Walker C, and Walker Teinkung The company is also demonstrating production line-focused swarm intelligence with these robots

Fourier officially unveiled GR-3, a "caring and capable" soft-touch bot that stands at 5'4" and weighs 55 kg It integrates vision, audio, and tactile sensors with expressive features such as blinking and eye tracking to ensure friendly interactions