GLM 5.2 is now on DeepSWE as the top open-source model on our leaderboard.
With a pass@1 score of 44% at max effort, GLM 5.2 is indisputable #1 open-source model besting Kimi K2.7 Code by 17%.
Introducing FrontierCode: a coding eval that raises the bar for difficulty & quality. Each task took 40+ hrs of work by leading open-source maintainers.
Models write sloppy code that works but isn’t maintainable. Our eval is first to measure: would you actually merge this code?
Composer 2.5 is now available inside Grok Build.
Composer 2.5 is a fast, highly intelligent model that excels on long-running tasks and following complex instructions.
We built a bipedal robot for about $2,500.
A real, mostly 3D-printed robot you can build, repair, simulate, train, and control.
Today we’re releasing LeRobot Humanoid: an open robot-learning platform with hardware, runtime, identification tools, and training environments.
Blog post: https://t.co/zu2etb1NZo
Repo: https://t.co/4myLRUtZ3W
actually, much easier to solve CAD engineering than it seems, because we have complete loop of Reinforcement Learning with Verifiable Rewards using FEA, CFD, and other type of engineering simulations
Welcome @anthropicai as a Corporate Patron sponsor of the Blender Foundation’s Development Fund! This support will be dedicated towards general Blender core development. More details at https://t.co/oiQeATUVW3 #b3d#DevFund
Happy to share that TriAttention now supports Apple Silicon and AMD chips 🚀
Both work seamlessly with TurboQuant👏.
Huge thanks to @DeadByDawn1O1 for the Apple Silicon contribution and @domvox for the AMD via llama.cpp!
Please check it out on GitHub https://t.co/CC1jjV837W
Today we're releasing WildDet3D—an open model for monocular 3D object detection in the wild.
It works with text, clicks, or 2D boxes, and on zero-shot evals it nearly doubles the best prior scores. 🧵
Ngl everyone using Chinese open weight models makes me anxious about sleeper agents.
We need American open source models to provide a similar performance alternative