ProvablyEfficientAlgorithmsforMulti-ObjectiveCompetitiveRLTianchengYu1YiTian1JingzhaoZhang1SuvritSra1Abstractaveragereturntoatargetsetsmallaslongasthissetsatisfiesaconditioncalledapproachability(Bl...
AdversarialPolicyLearninginTwo-playerCompetitiveGamesWenboGuo1XianWu1SuiHuang2XinyuXing1Abstract2020),wearguethatattacksdevelopedunderthisassump-tionarenotpractical.Forexample,givenamasteragentInat...
ProvableSelf-PlayAlgorithmsforCompetitiveReinforcementLearningYuBai1ChiJin2Abstractconflictingrewards(sothattheyessentiallycompetewitheachother)yetcanbetrainedinacentralizedfashion(i.e.Self-play,wh...
ImplicitCompetitiveregularizationinGANsFlorianScha¨fer1HongkaiZheng23AnimaAnandkumar1AbstractTheminimaxinterpretation:Presently,thesuccessofGANsismostlyattributedtopropertiesofthedivergenceThesucc...
CompetitiveMulti-agentInverseReinforcementLearningwithSub-optimalDemonstrationsXingyuWang1DiegoKlabjan1Abstractoftherewardfunction,oratleastobservationsofimmediatereward.Somelearningtasks,however,p...
CompetitiveCachingwithMachineLearnedAdviceThodorisLykouris1SergeiVassilvitskii2Abstractdoingso,thesystemmayinvestitseffortsonreducingtheerroronthemajorityofinputs,attheexpenseofincreasedWedevelopaf...