Training Substrate

All research › LLM Architecture & Training Design

Reinforcement-learning-driven control of the training loop, self-discovering architectures, and the layer-role taxonomy that surfaces when those mechanisms are applied at scale.

Papers

Listed in reading order.

01 Research Posture: Why We Let the Model Choose Why every hyperparameter, threshold, and gate in a training stack should be learned, not picked.
02 Reward-Driven Training Control A small reinforcement-learning policy on the training-loop control surface beat a hand-tuned baseline by 42% at matched compute.
03 Self-Discovering Architectures The same primitive applied one level up the stack: letting the model discover its own per-layer width profile mid-training.
04 Emergent Layer Roles and Functional Specialization A four-role taxonomy of layers in trained transformers (hub, damper, passthrough, specialist), computable from any finished checkpoint.