Inkling-Small Brings Frontier MoE AI to a Single GPU
Inkling-Small Brings Frontier MoE AI to a Single GPU Thinking Machines Lab has just released Inkling-Small, an open-weights Mixture-of-Experts (MoE)
Read more
Neural Network
Why Default Optimizer Settings Are Sabotaging Your Machine Learning Models
Machine learning practitioners have a dirty little secret. Most of us never touch our optimizer settings. We import Adam or
Read more
Apple Makes RNN Training Practical at Billions of Parameters for the First Time
Apple researchers made recurrent neural network training practical at billions of parameters for the first time, reopening a previously abandoned architectural direction more memory-efficient than transformers. Here’s what this breakthrough, plus new cancer prediction and neuroimaging foundation models, reveal about where machine learning research is headed.
Read more