OntheEffectivenessofLipschitz-DrivenRehearsal inContinualLearning-SupplementaryMaterial

Neural Information Processing Systems 

If α > β, we are overemphasizing the contribution of the first term of Eq. 9 (which brings each layer'sλk1 andck close toeach other) overthesecond one(which induces small Lipschitz targets).