Differentiability and overlap concentration in optimal Bayesian inference

Hong-Bin Chen, Victor Issa

In this short note, we consider models of optimal Bayesian inference of finite-rank tensor products. We add to the model a linear channel parametrized by $h$. We show that at every…