I am a PhD student navigating the intersection of Theoretical Computer Science and Statistical Physics.
Based in Nice, France, I am affiliated with 3IA Côte d’Azur, Inria, and CNRS. Working under the supervision of Prof. Emanuele Natale, my research aims to mathematically demystify why deep learning works so well, with a specific focus on network pruning.
Prior to this, I completed my BS-MS in Physics at IISER Kolkata, giving me a strong foundation in the physical principles that often govern complex systems like Neural Networks.
Strong Lottery Ticket Hypothesis (SLTH)
The SLTH suggests that within a large, randomly initialized neural network, there exists a "lucky" subnetwork that performs well without any weight training—purely by identifying the right structure (pruning).
My current work investigates the interplay between Quantization and Pruning. By applying tools from probability theory, I analyze the existence of these subnetworks in quantized regimes, paving the way for ultra-efficient neural networks that require minimal compute to train and deploy.
Team COATI. Advisors: Emanuele Natale.
Thesis: "Phase transition in Artificial Neural Networks" .
Teaching Assistant for Mathematical Methods I.
