davidAlzate() @illker

Gemma 4 by hand ✍️ in Excel tomorrow RSVP 👉 luma.com/f0annk18 == Outline == 1. Attention 1.1 Scaled Product Dot-Product Attention (SDPA) 1.2 Global Attention 1.3 Causal Attention 1.4 Local Attention 2. KV Cache 2.1 Decoding (no caching) 2.2 Decoding with KV Cache 3. Mixture of Experts (MoE) 3.1 Feed Forward Layer 3.2 Router 3.3 Experts 3.4 Mixture 4. Per-Layer Embedding (PLE) 4.1 Input Embedding 4.2 Residual Stream 4.3 Layer Embedding 4.4 Low-Rank Layer Embedding 4.5 Dynamic Low-Rank Layer Embedding 4.6 Static + Dynamic Layer Embedding #aibyhand ---- AI Math, Algorithms, Architectures by hand ✍️ Subscribe to my 60K+ reader newsletter 👉 byhand.ai

Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: goo.gle/4bsq2qI

It’s here. A recap of my #SpotifyWrapped spotify.com/wrapped-share/…

🔥 Qwen-Image-Edit-2509 IS LIVE — and it’s a GAME CHANGER. 🔥 We didn’t just upgrade it. We rebuilt it for creators, designers, and AI tinkerers who demand pixel-perfect control. ✅ Multi-Image Editing? YES. Drag in “person + product” or “person + scene” — it blends them like magic. No more Franken-images. ✅ Single-Image? Rock-Solid Consistency. • 👤 Faces stay you — through poses, filters, and wild styles. • 🛍️ Products keep their identity — ideal for ads & posters. • ✍️ Text? Edit everything: content, font, color, even material texture. ✅ ControlNet Built-In. Depth. Edges. Keypoints. Plug & play precision. ✨ Blog: qwen.ai/blog?id=7a9009… 💬 QwenChat: chat.qwen.ai/?inputFeature=… 🐙 GitHub: github.com/QwenLM/Qwen-Im… 🤗 HuggingFace: huggingface.co/Qwen/Qwen-Imag… 🧩 ModelScope: modelscope.cn/models/Qwen/Qw…

He changed frontend engineering forever. The Satoshi of UI. No one knows his whereabouts or his real name. He didn't ship a component system, he shipped a system to ship component systems.

🚀 in beta mode

more fun than ever to be a software engineer

The wait is over: Deep Think is here. At I/O, we previewed the frontiers of Gemini’s thinking capabilities. Now, @Google AI Ultra subscribers can experience it in the Gemini app. With Deep Think, Gemini 2.5 is able to intelligently extend its "thinking time" so it can generate multiple, parallel streams of thought simultaneously. Similar to the way humans brainstorm when they need to tackle complex problems that require creativity or strategic planning. Wondering how that thinking time makes a difference? Check out these examples 🧵

Imagine if gas stations didn't tell you how many gallons you were getting. The station built your car, but you can't look under the hood. That's the inference subscription dilemma in one image, and the reality of selling commodities on subscription models. 🧵

Anthropic @AnthropicAI

10 months ago

We’re rolling out new weekly rate limits for Claude Pro and Max in late August. We estimate they’ll apply to less than 5% of subscribers based on current usage.

566 304 3K 2.3M 652

Tip: call the context file AGENTS.md and symlink to CLAUDE.md, GEMINI.md pnote.eu/notes/agents-m…

>>> Qwen3-Coder is here! ✅ We’re releasing Qwen3-Coder-480B-A35B-Instruct, our most powerful open agentic code model to date. This 480B-parameter Mixture-of-Experts model (35B active) natively supports 256K context and scales to 1M context with extrapolation. It achieves top-tier performance across multiple agentic coding benchmarks among open models, including SWE-bench-Verified!!! 🚀 Alongside the model, we're also open-sourcing a command-line tool for agentic coding: Qwen Code. Forked from Gemini Code, it includes custom prompts and function call protocols to fully unlock Qwen3-Coder’s capabilities. Qwen3-Coder works seamlessly with the community’s best developer tools. As a foundation model, we hope it can be used anywhere across the digital world — Agentic Coding in the World! 💬 Chat: chat.qwen.ai 📚 Blog: qwenlm.github.io/blog/qwen3-cod… 🤗 Model: hf.co/Qwen/Qwen3-Cod… 🤖 Qwen Code: github.com/QwenLM/qwen-co…

supervision-0.26.0 is out we finally released support for ViTPose and ViTPose++ pose estimation models from @huggingface transformers link: github.com/roboflow/super…

Next.js 15.4 • Turbopack Builds: 100% integration test compatibility for next build --turbopack • General stability and performance improvements • A preview of what's coming in Next 16 nextjs.org/blog/next-15-4

Cline doesn't index your codebase. No RAG, no embeddings, no vector databases. This isn't a limitation -- it's a deliberate design choice. As context windows increase, this approach enhances Cline's ability to understand your code. Here's why. 🧵

Open source, everything

Visual Studio Code @code

a year ago

Today, we're announcing plans to make VS Code an open source AI editor. We believe AI development should stay true to VS Code's core principles: open, collaborative, and community-driven. Let's build the future of software development together. aka.ms/open-source-ai…

369 2K 14K 1.9M 2K

Introducing AlphaEvolve: a Gemini-powered coding agent for algorithm discovery. It’s able to: 🔘 Design faster matrix multiplication algorithms 🔘 Find new solutions to open math problems 🔘 Make data centers, chip design and AI training more efficient across @Google. 🧵

@windsurf_ai > @cursor_ai

🚀 Cline 3.4 is here! We've been cooking up something special, and today we're excited to share this huge update! Here's what's new: ⭐️ MCP Marketplace: Install the best MCP servers with just one click 🧜‍♀️ Mermaid Diagrams: Visual flowcharts right in your chat We've also added: • Smarter Context: Real-time terminal awareness + Git integration • New Models: Qwen 2.5, DeepSeek, and enhanced AWS Bedrock support • Advanced Settings: More control over your development environment The best part? It's all designed to make your coding workflow smoother and more intuitive. Get started with Cline 3.4 👇

Stremio v5 for Windows is Now Open Source: blog.stremio.com/stremio-v5-for…

magnet:?xt=urn:btih:11f2d1ca613ccf5a5c60104db9f3babdfa2e6003&dn=Mistral-Small-3-Instruct&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=http%3A%2F%https://t.co/ua2yzvEYLu%3A1337%2Fannounce