Souvik Kundu

Linearizing Models for Efficient yet Robust Private Inference

2024 · arXiv (Cornell University)

The growing concern about data privacy has led to the development of private inference (PI) frameworks in client-server applications which protects both data privacy and model IP. However, the cryptographic primitives required yield significant latency …

Analyzing Adversarial Vulnerabilities of Graph Lottery Tickets

2024

Graph neural networks (GNNs) have displayed significant potential in various graph-based learning tasks. However, the computational demands of deploying GNNs on large-scale graphs can grow exponentially. A recent method, termed unified graph sparsification (UGS), shows …