Monte Carlo simulations in many different kind and varieties are used in many fields of theorethical physics, from statistichal physics to high energy physics passing through condensed matter physics. Those simulations have the aim to simulate systems on a smaller scale in order to compute observables that we could not estimate otherwise because of some mathemathical constraints. In the last few years, the increasing power and performances of generative models in the pletora of tasks common for the Machine Learning community, attracted an increasing number of people from different fields, including physicists. A lot of nice publications recently came out. In those works, researchers were trying to come up with new suitable configurations in simple toy models using generative models such as GANs, RBMs and so forth. The focus of the research was mainly on well-known simple models such as Ising in 2D or Potts and XY models for some more advanced and difficult challenges. Nevertheless, despite is not an obvious task for a generative model to learn a complicate underlying distribution of such physical systems, they show to be able to generate suitable configurations under specific settings. Of course, although the results achieved so far look promising, this is just a starting point. The challenges out there are enourmous and the efforts that several researchers from all over the world are making in order to get some well-performing algorithm, might open the doors of a truly new, ambitious and exciting field of research. Quantum Chemistry met the benefits of Machine Learning few years ago; has the time came for Physics as well? Stay tuned if you want to know where the story goes.

For the people among you which are interested in knowing more about the topic, feel free to contact me directly. More will be released soon!

Cheers,

Kim