yangdf

@Zyyon_ Please don’t tell anyone: we aren’t just open sourcing all the models. We are doing the unspeakable: open sourcing all the gpu kernels too. Making them run on multivendor consumer hardware, and opening the door to folks who can beat our work. Plz keep it quiet, ok? 😉

Zhilin's full GTC 2026 keynote is here. If you're curious about the "how" behind scaling Kimi’s latest models, this is the session you can't miss. :)

I just launched /office-hours skill with gstack. Working on a new idea? GStack will help you think about it the way we do at YC. (It's only a 10% strength version of what a real YC partner can do for you, but I assure you that is quite powerful as it is.)

Qwen3.5-4B searched 20+ websites, cited its sources, and found the best answer! 🔥 Try this locally with just 4GB RAM via Unsloth Studio. The 4B model did this by executing tool calls + web search directly during its thinking trace.

🙌 Andrej Karpathy’s lab has received the first DGX Station GB300 -- a Dell Pro Max with GB300. 💚 We can't wait to see what you’ll create @karpathy! 🔗 #dgx-station" target="_blank" rel="nofollow noopener">blogs.nvidia.com/blog/gtc-2026-… @DellTech

OpenAI just dropped a training challenge: Train a <16MB language model in 10 minutes on 8×H100s and minimize held-out loss on a fixed FineWeb dataset. Basically NanoGPT Speedrun. They’re sponsoring $1M in compute. I can summon my autoresearch army to win it… if I have time.

Apple's M4 chip engineers earn $450k+ Intel's CPU architects earn $400k+ NVIDIA's hardware engineers power every AI model on the planet They all understand one thing almost no software dev ever studies: How a computer actually works at the hardware level "ETH Zurich – Digital Design & Computer Architecture" by Onur Mutlu Free on YouTube. 30+ full lectures. Spring 2023 By a professor who holds joint appointments at both ETH Zurich and Carnegie Mellon Starts from a single transistor. Ends with a complete CPU you understand entirely: • Logic gates – how electricity becomes computation at the most fundamental level • Instruction Set Architecture – the contract between software and hardware every dev ignores • Pipelining – how your CPU executes multiple instructions simultaneously without you knowing • Out-of-order execution – why your CPU secretly reorders your code to run faster • Memory hierarchy – the design decision that determines the speed of every program ever written Every line of code you've ever written ran on hardware you don't understand The engineers who built that hardware earn $450k Now you know where to start

I (finally) put together a new LLM Architecture Gallery that collects the architecture figures all in one place! sebastianraschka.com/llm-architectu…

🚨 Want to parse complex PDFs with SOTA accuracy, 100% locally? 📄🔍 At just 0.9B parameters, you can drop GLM-OCR straight into LM Studio and run it on almost any machine! 🥔 🧠 0.9B total parameters 💾 Runs on < 1.5GB VRAM (or ~1GB quantized!) 💸 Zero API costs 🔒 Total data privacy Desktop document AI is officially here. 💻⚡

🎾Introducing LATENT: Learning Athletic Humanoid Tennis Skills from Imperfect Human Motion Data Dynamic movements, agile whole-body coordination, and rapid reactions. A step toward athletic humanoid sports skills. Project: zzk273.github.io/LATENT/ Code: github.com/GalaxyGeneralR…

@cloudwu @cyberlancer 我的经验证明确实如此。如果一直在疲于应付不同的项目，就会难以反思积累和沉淀。而《人月神话》的叙事，在LLM时代似乎要引来巨大的挑战了！

我的工作是不是很好笑

@cloudwu @geniusvczh 青山遮不住

@geniusvczh 现实是：只要用 AI 做素材就打差评，打折并不能改好评。

游戏行业有特殊问题，玩家觉得凡是用了AI做素材的都必须打三折，因为你花钱少了🤪

We're sharing a new method for scoring models on agentic coding tasks. Here's how models in Cursor compare on intelligence and efficiency:

last week we got 1M views and 100s of death threats for giving Openclaw access to drones, humanoids, quadrupeds, and other physical hardware. Now we’re releasing EVERYTHING open-source. Dimensional gives agents access to the physical world. Join us. Repo 👇🏽👇🏽👇🏽

Introducing the new /crawl endpoint - one API call and an entire site crawled. No scripts. No browser management. Just the content in HTML, Markdown, or JSON.

Training runs slow, memory fills up, and multi-GPU scaling behaves strangely... That's usually about when ML engineers realize the GPU can't stay a black box. In CUDA for Deep Learning, @elliotarledge explains what's actually happening inside the GPU and how understanding architecture, memory behavior, and parallelism can unlock major performance gains. Hear him expand on it here: hubs.la/Q0462hgB0

Everyone is misreading this chart. At first glance it looks scary for Software Engineers. According to Anthropic’s data, 96% of software development tasks are exposed to being replaced by AI. That’s the highest of any profession. - Higher than finance. - Higher than legal. - Higher than management. If you stop reading there, the conclusion seems obvious: - Software Engineers are the first to be replaced. But look closer. Actual observed usage is only 32%. And more importantly, ask the second question: - Who is building the automation for every other industry? Software Engineers! AI does not eliminate software. It makes software dramatically cheaper to produce. And when something becomes cheaper to produce, demand explodes. This is the Jevons paradox of software. As developers become AI-augmented, they do not disappear. They build: - AI systems for finance - automation for legal workflows - decision engines for healthcare - optimization tools for logistics Every industry in that chart becomes programmable. Software Engineers may be the first profession heavily automated by AI. But they are also the ones automating the rest of the economy. And that is why software demand keeps rising. (thanks Peter Walker for making the radar chart a bar charts, much easier to read)

Just found out eon open sourced their code here: github.com/eonsystemspbc/… Will review it later but seems like my method is very similar

We've uploaded a fruit fly. We took the @FlyWireNews connectome of the fruit fly brain, applied a simple neuron model (@Philip_Shiu Nature 2024) and used it to control a MuJoCo physics-simulated body, closing the loop from neural activation to action. A few things I want to say about what this means and where we're going at @eonsys. 🧵