SOURCE-LINKED INTELLIGENCE

How Model Growth, Recursion, and Boundary Operators Influence Scaling Exponents

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

Scaling laws predict how loss decreases with increases in computation. We show, contrary to conventional wisdom, that architectural interventions can modify scaling exponents in pre-training, leading to power-law improvements in performance as computation increases. As an anchoring point, we consider the architectural formulation of looped transformers. Although not typically used in this way, looping, also known as recursive depth, provides a mechanism for model growth, by increasing the number of loops during training. Model growth, with and without shared weights, provides the biggest chang

Read original source ↗ Open in workspace

recordType

paper

region

Global