Entropy Bounds for Gibbs Measures with Non-Gaussian Tails

We introduce and study relative entropy bounds which (under some mixing conditions) allow us to classify the distribution tails in the infinite dimensional setting.