On large deviation regimes for random media models (Q1024906)

The paper analyzes large deviations of random functionals above the average and below the average in random media models. Two models are considered: a first model which is an extension of the continuous time parabolic Andersons', and the second model which is the standard first passage percolation. One considers successively the analysis far from the starting point, and the analysis near the starting point in the Gaussian case and the sub-Gaussian case respectively. It is shown that the large deviations above and below the limiting value are significantly different.