Publication | Closed Access

Potential-based multiobjective reinforcement learning approaches to low-impact agents for AI safety

Artificial IntelligenceEngineeringAgent Decision-makingSafety SciencePotential-based Multiobjective ReinforcementAi SafetySystems EngineeringComputer ScienceIntelligent SystemsRobot LearningLow-impact AgentsMulti-agent LearningRoboticsMulti-agent PlanningAi Safety Education