Robert McCarthy

steganographic collusion (2) llm safety (2) multi-agent system (2) reinforcement learning (2) positive unlabeled learning (1) robot learning (1) offline learning (1) language model (1) foundation model (1) robotics manipulation (1) physical dynamics (1) agent system (1) information hiding (1) large language model (1) behavioural cloning (1) internet video (1) reward incentive (1) communication paraphrasing (1) imitation learning (1) transfer learning (1)