Xiangqi Structurally Variable

Research on Xiangqi (Chinese chess) is exploring the game's inherent complexities, particularly its non-transitive nature where player A may beat B, B beats C, but C beats A. Current efforts focus on developing AI agents that not only master the game but also mimic human playing styles, employing diverse architectures like structurally variable networks and algorithms combining Monte Carlo Tree Search with techniques like Policy Space Response Oracles. These advancements contribute to a deeper understanding of game AI, offering insights into both optimal strategies and the unpredictable aspects of human decision-making in complex games.