AI RESEARCH

A Model-Free Universal AI

arXiv CS.AI

ArXi:2602.23242v2 Announce Type: replace In general reinforcement learning, all established optimal agents, including AIXI, are model-based, explicitly maintaining and using environment models. This paper