Return to Article Details Safe Learning of Multi-Agent Action Models from Concurrent Joint Action Observations Download Download PDF