Mahendra KariyaMeasuring Availability of a Service, MeaningfullyDiscussion of the paper “Meaningful Availability” published by Hauer et al. from GoogleJun 18, 20201Jun 18, 20201
Mahendra KariyaDark Knowledge in Neural NetworksImplementing (in Keras) a few ideas presented in the paper “Distilling the Knowledge in a Neural Network” by Geoffrey Hinton, Oriol…Dec 31, 20183Dec 31, 20183
Mahendra KariyaDo CNNs actually learn what they should learn?Summary of the paper “Measuring the tendency of CNNs to Learn Surface Statistical Regularities” by Jason Jo and Yoshua BengioDec 26, 2018Dec 26, 2018
Mahendra KariyaAttention in NMT — 2Part 2 of a three post series to discuss the Attention Mechanism in Neural Machine Translation systemsDec 9, 2018Dec 9, 2018
Mahendra KariyainGojek Product + TechTraining Neural nets through BatchNorm: Challenging a 2015 studyUnderstanding the what and why of Batch NormalizationDec 7, 2018Dec 7, 2018
Mahendra KariyaFederated LearningOn device inference is very common. On device training, not so much. Federated learning paves the way for doing on device training on…Nov 29, 2018Nov 29, 2018
Mahendra KariyaAttention in NMT — 1Part 1 of a three post series to discuss the Attention Mechanism in Neural Machine Translation systemsNov 21, 2018Nov 21, 2018
Mahendra KariyaPaper Discussion: Explaining and harnessing adversarial examplesDiscussion of the paper “Explaining and harnessing adversarial examples” presented at ICLR 2015 by Goodfellow et al.Nov 14, 2018Nov 14, 2018
Mahendra KariyainGojek Product + TechTips and tricks on how to clear our automated coding roundA definitive guide of our automated coding rounds — best practices, how to pass, common mistakes made and how to steer clear of themJun 21, 20181Jun 21, 20181
