Coverage 1 publisher
Articles stay on their publishers’ sites; each link opens the original.
New
Mellum releases Mellum2.1, an open 12B mixture-of-experts model designed for coding agents and fast sub-agents running on local hardware. It uses the same architecture as version 2, with 2.5B active parameters, and is trained with reinforcement learning in real environments.
Articles stay on their publishers’ sites; each link opens the original.