Universal Spin Models are Universal Approximators in Machine Learning (2507.07669v1)

Abstract: One of the theoretical pillars that sustain certain machine learning models are universal approximation theorems, which prove that they can approximate all functions from a function class to arbitrary precision. Independently, classical spin models are termed universal if they can reproduce the behavior of any other spin model in their low energy sector. Universal spin models have been characterized via sufficient and necessary conditions, showing that simple models such as the 2d Ising with fields are universal. In this work, we prove that universal spin models are universal approximators of probability distributions. As a consequence, the characterization of the former gives rise to universal approximation theorems for the latter. This allows us to derive new proofs of universal approximation theorems for restricted and deep Boltzmann machines, as well as deep belief networks. This work illustrates that independently discovered universality statements may be intimately related, enabling the transfer of results.