PinnedSaurav JhaNeural Processes for Uncertainty-Aware Continual LearningWith this blog post, I intend to explain our NeurIPS paper (going by the same title) in simple terms. This blog does not cover all the…Oct 31, 2023Oct 31, 2023
Saurav JhaOn generalization capability of randomly initialized vs pre-trained weightsTLDR; Q: “Can training neural networks from randomly initialized weights lead to better generalization than from pre-trained weights?” A…Feb 11, 2022Feb 11, 2022
Saurav JhaMy European Master’s: An EMA-zing journeyUpdate: A republication of this blog post is now available on Erasmus Mundus’s website!Aug 30, 2021Aug 30, 2021
Saurav JhainTowards Data ScienceContinual learning — where are we?As the deep learning community aims to bridge the gap between human and machine intelligence, the need for agents that can adapt to…Sep 16, 2020Sep 16, 2020
Saurav JhainTowards Data ScienceVisualizing the Scimago journal ranking database with PandasSCImago Journal Rank (SJR indicator) is a measure of scientific influence of scholarly journals that accounts for both the number of…Feb 22, 2019Feb 22, 2019
Saurav JhainTowards Data ScienceHierarchical RNNs, training bottlenecks and the future.As we know, the standard backpropagation algorithm is the most efficient procedure to compute the exact gradients of a loss function in a…Feb 15, 20191Feb 15, 20191