Search Self-Play (SSP) is a reinforcement learning framework designed for training adversarial self-play agents with integrated search capabilities—enabling both proposer and solver agents to conduct ...