Aweakness toexisting approaches isthattheyignore thefactthatadversarial example generation is often a sequential task where multiple similar problems are being solved in a row. That is, one has access to a large number of "normal" examples each of which should be perturbed to elicit

Another where, thegoal recommend improvethe10]. Theworkof YMreceivedfunding fromthe European Research Council (ERC) underthe European Union' s Horizon 2020 research andinnovationprogram (grantagreement No. 882396), andbythe Israel Science Foundation (grant number 993/17).

In all cases, we show significant improvement over baselines across a varietyofMuJoCotasks.

Unlike AC, DAC can dynamically adjust the algorithm's configuration during