1: \begin{abstract}

2: Logarithmic Sobolev inequalities are a powerful way to estimate the

3: rate of convergence of Markov chains and to derive concentration inequalities

4: on distributions. We prove that the log-Sobolev constant of any isotropic

5: logconcave density in $\R^{n}$ with support of diameter $D$ is $\Omega(1/D)$,

6: resolving a question posed by Frieze and Kannan in 1997. This is asymptotically

7: the best possible estimate and improves on the previous bound of $\Omega(1/D^{2})$

8: by Kannan-Lov\'{a}sz-Montenegro. It follows that for any isotropic logconcave

9: density, the ball walk with step size $\delta=\Theta(1/\sqrt{n})$

10: mixes in $O\left(n^{2}D\right)$ proper steps from \emph{any }starting

11: point. This improves on the previous best bound of $O(n^{2}D^{2})$

12: and is also asymptotically tight. The new bound leads to the following

13: refined large deviation inequality for an $L$-Lipschitz function $g$

14: over an isotropic logconcave density $p$: for any $t>0$, 

15: \[

16: \P_{x\sim p}\left(\left|g(x)-\bar{g}\right|\geq c\cdot L\cdot t\right)\leq\exp(-\frac{t^{2}}{t+\sqrt{n}})

17: \]

18: where $\bar{g}$ is the median or mean of $g$ for $x\sim p$; this

19: generalizes/improves on previous bounds by Paouris and by Guedon-Milman.

20: The technique also bounds the ``small ball'' probability in terms

21: of the Cheeger constant, and recovers the current best bound. Our

22: main proof is based on stochastic localization together with a Stieltjes-type

23: barrier function.

24: \end{abstract}