Measuring Maximum Activations in Open Large Language Models Paper โข 2605.15572 โข Published May 15 โข 18
EndPrompt: Efficient Long-Context Extension via Terminal Anchoring Paper โข 2605.14589 โข Published May 14 โข 17
Running on CPU Upgrade Featured 3.22k The Smol Training Playbook ๐ 3.22k The secrets to building world-class LLMs