12 citations · 33 across the 15 of their papers we have counts for
15 papers
Localization Without Disorder: Quantum Walks on Structured Graphs
Shyam Dhamapurkar, K. Venkata Subrahmanyam
Continuous-time quantum walks (CTQWs) exhibit localization phenomena that differ fundamentally from their classical counterparts, yet the precise relationship between network struc…
Orbit closures, stabilizer limits and intermediate -varieties
Bharat Adsul, Milind Sohoni, K V Subrahmanyam
In this paper we study the orbit closure problem for a reductive group acting on a finite dimensional vector space over $\C$. We assume that the center of $G…
Geometric Complexity Theory -- Lie Algebraic Methods for Projective Limits of Stable Points
Bharat Adsul, Milind Sohoni, K V Subrahmanyam
Let be a connected reductive group acting on a complex vector space and projective space . Let and be the Lie algebra of…
How do SGD hyperparameters in natural training affect adversarial robustness?
Sandesh Kamath, Amit Deshpande, K V Subrahmanyam
Learning rate, batch size and momentum are three important hyperparameters in the SGD algorithm. It is known from the work of Jastrzebski et al. arXiv:1711.04623 that large batch s…
On Universalized Adversarial and Invariant Perturbations
Sandesh Kamath, Amit Deshpande, K V Subrahmanyam
Convolutional neural networks or standard CNNs (StdCNNs) are translation-equivariant models that achieve translation invariance when trained on data augmented with sufficient trans…
Universalization of any adversarial attack using very few test examples
Sandesh Kamath, Amit Deshpande, K V Subrahmanyam +1
Deep learning models are known to be vulnerable not only to input-dependent adversarial attacks but also to input-agnostic or universal adversarial attacks. Dezfooli et al. \cite{D…