In essence the NN evaluations are acting as a PRNG with values heavily biased by an established heuristic. So while it's very much a specialization of MCTS and

In essence the NN evaluations are acting as a PRNG with values heavily biased by an established heuristic. So while it's very much a specialization of MCTS and probably should go by a different name, one could argue it's still using monte carlo methods.