The Kac version of the Sherrington-Kirkpatrick model at high temperatures (Q1284904)

The author is engaged in the question to what extent the results for the mean-field Sherrington-Kirkpatrick (SK) model are also relevant for its Kac version, a lattice model with finite-range interaction. He proves that for all inverse temperatures \(\beta<1\), its free energy converges to that of the SK model, as the range of the interaction tends to infinity. He then considers the fluctuations in the large volume limit. He proves that for all temperatures, for which the infinite volume Gibbs state is unique, the distribution of the free energy, scaled by the square root of the volume, converges to a normal distribution. Its variance is shown to converge to 0 for almost all \(\beta\), as the range of the interaction tends to infinity. This is in contrast to the corresponding result for the SK model, where the fluctuations have to be scaled by the volume itself. But the results show that on the scale of the square root of the volume, the variance of the Kac model converges to that of the SK model.