I joined the AI residency right after graduating from IIT Kanpur, with a major in Electrical engineering and minors in Machine Learning and Linguistics. During my undergraduate, I spent most of time working on probabilistic models, in particular, making mixture-of-experts based models faster through latent-variable augmentation techniques. Also, I spent some time in MILA working on designing better gradient estimators for discrete latent variables. AI Residency is a very unique opportunity: One gets to collaborate and interact with incredibly talented colleagues, the flexibility to choose the problems of their interest and access to an incredible amount of compute! I am interested in Machine Learning, both theoretical and practical aspects of it. Some particular goals of research that broadly interest me are (in no particular order) (1) a better understanding of Deep Learning (2) improving the Deep Reinforcement Learning systems to learn from less instances while maintaining/improving their performance (3) integrating probabilistic techniques with Deep Learning. Machine Learning in all its diversity interests me, but, more recently I have been focussing on Deep Reinforcement Learning. In particular, I am currently researching on making Deep Reinforcement Learning sample-efficient by leveraging information from multiple tasks and/or using objectives augmented with auxiliary rewards.