If for all and is an admissible cylindrical potential with semiconvexity constant where , then the Gibbs measure of at temperature satisfies log-Sobolev inequalities in entropy form and in Gross's form and a Talagrand inequality in the noise Wasserstein distance, each with constant governed by .
In the setting of Gaussian Analysis Relative to a Diagonal Gaussian Reference Measure with Noise Weights: Standing Notation, with the noise Wasserstein distance on and relative entropy of Gaussian Analysis Relative to a Diagonal Gaussian Reference Measure with Noise Weights: Standing Notation §background, and and the noise gradients of Gaussian Analysis Relative to a Diagonal Gaussian Reference Measure with Noise Weights: Standing Notation §cylindrical. Let be positive with for every , let satisfy and , and let be an admissible cylindrical potential with semiconvexity constant . is the Gibbs measure of at temperature ; the relative score with respect to and finite Fisher information relative to with weights , with value , are those of the cited definitions, applicable to every of finite relative entropy with respect to by Entropy and Score Relative to the Gibbs Measure Split into Their Gaussian Parts and the Potential, with a Fisher Information Bound §domain; and is the entropy with respect to .
1. (Log-Sobolev inequality, entropy form) Let have finite relative entropy with respect to , a relative score with respect to , and finite Fisher information relative to with weights . Then
2. (Talagrand inequality) Let have finite relative entropy with respect to . Then and belong to , and
3. (Log-Sobolev inequality, Gross's form) For every , the functions and , with the function of The Function : Continuity, Young's Inequality and Lower Bounds, with the Elementary Bounds for the Exponential and the Logarithm, are bounded and Borel, hence integrable with respect to , and
Loading…
No relations recorded yet.