<ul><li>A new hybrid variant of the Monte Carlo Tree Search (MCTS) algorithm, called search-contempt, has been introduced to improve the computational efficiency of training AlphaZero-like engines.</li><li>The search-contempt algorithm alters the distribution of positions generated in self-play, emphasizing more challenging positions.</li><li>It has been demonstrated to significantly enhance the strength of engines in Odds Chess.</li><li>The use of search-contempt enables the possibility of training self-play based engines with fewer training games, lowering the computational and financial costs.</li></ul>

Search-contempt: a hybrid MCTS algorithm for training AlphaZero-like engines with better computational efficiency

Discover more