Finite-Size Gradient Transport in Large Language Model Pretraining: From Cascade Size to Intensive Transport Efficiency

Ping Wang, Yan-Qi Du

We introduce a finite-size gradient-transport framework for real language-model training, based on five observables $(D,z,{\hat{I}}^2,\hat{I}´,v_{\mathrm{rel}})$ that separate cascade size, durat…