Enabling Adaptive Agent Training in Open-Ended Simulators by Targeting Diversity

–Neural Information Processing Systems 

[no summary]