@dotey Remotion + Claude Code，只是不知道为什么 CC 把这个视频当作一个游戏了 :)

motioncanvas 这个也挺强的 https://t.co/ANO11d9TvG

I spent my summer building TinyTPU : An open source ML inference and training chip. it can do end to end inference + training ENTIRELY on chip. here's how I did it👇:

Sub-agents in Claude Code, clearly explained:

@immccn123 并非第一（ 参见

If you found it insightful, reshare with your network. Find me → @akshay_pachaar ✔️ For more insights and tutorials on LLMs, AI Agents, and Machine Learning!

I have been training neural networks for 9 years now. Here are 16 ways I actively use to optimize model training:

How to use Kimi K2 in Claude Code: 1. Create an account at @OpenRouterAI 2. npm install -g @anthropic-ai/claude-code 3. npm install -g @musistudio/claude-code-router 4. Add the following lines to your ~/.claude-code-router/config.json (update with your OpenRouter API key) 5. ccr

We made step-by-step guides to Fine-tune & Run every single LLM! 🦥 What you'll learn: • Technical analysis + Bug fixes explained for each model • Best practices & optimal settings • How to fine-tune with our notebooks • Directory of model variants 🔗 https://t.co/7dC0UtRocu

Here's a cool paper using LLMs for lossless text compression, in what they call LLMZip, which outperforms SOTA text compression methods The idea is very intuitive Given a sentence to compress, like "My first attempt", they feed the first 2 tokens ("My" and " first") to... [1/6]

Announcing DeepSWE 🤖: our fully open-sourced, SOTA software engineering agent trained purely with RL on top of Qwen3-32B. DeepSWE achieves 59% on SWEBench-Verified with test-time scaling (and 42.2% Pass@1), topping the SWEBench leaderboard for open-weight models. Built in

Two years ago, agents were just a research experiment. Now, many devs use them daily. For us and the teams we work with, there was a LOT of trial & error along the way. We're sharing our main lessons in "Agents 101" https://t.co/rS33VlFe5i

Aider v0.85.0 is out. - Support for Responses API models like o3-pro and o1-pro. - New Gemini 2.5 Pro models. - Updated costs for o3. - Repo-map & linting support for Clojure and MATLAB. - Aider wrote 21% of the code in this release. Full release notes: https://t.co/GIYaqnaSYq

this project stores millions of text chunks inside a video file (mp4) then runs sub-second semantic search on it - no vector DB, no servers - uses 10x less RAM & storage - no internet required it's called Memvid and it just broke my brain

Shocker! Claude 4 system prompt was leaked, and it's a goldmine! The Claude system prompt incorporates several identifiable agentic AI patterns as described in "A Pattern Language For Agentic AI." Here's an analysis of the key patterns used: Run-Loop Prompting: Claude

很多朋友关注 Cherry Studio 移动端的计划，跟大家大致同步一下。设计由 Siin 完成，开发已经启动。定位还是 PC 端的配套，所以功能上不会追求大而全，像知识库和 mcp，前期都不会提供，以简单易用为主。开发框架是 expo，会同时支持 Android 和 iOS。iOS 会上架 App Store，Android 前期只提供 apk。

After more than a month of work and after spending about $250 on LLMs and cloud GPUs, I finalized the long-awaited tutorial on how to train a language model-based document classifier based on a real-world taxonomy and using an ensemble of LLMs as training data labelers. Most

🔥 Excited to share our work "Efficient Agent Training for Computer Use" Q: Do computer use agents need massive data or complex RL to excel? A: No, with just 312 high-quality trajectories, Qwen2.5-VL can outperform Claude 3.7, setting a new SOTA for Windows computer use. 1/6

This Cursor Extension is awesome Accurate tweaking of UI was always a struggle, But @stagewise_io allows you to bring full context to Cursor, just point and command: 1. Directly choose specific elements in browser 2. Send to Cursor with full context And it's open source

The DeepWiki MCP server is live! How to use it + what’s inside 🧵👇