TY - RPRT TI - Reinforcing the World's Edge: A Continual Learning Problem in the Multi-Agent-World Boundary AU - Dane Malenfant PY - 2026 UR - https://arxiv.org/abs/2603.06813 ID - 2603.06813 ER -