Next Gen AI!

Difference in Information Flow Among the Artificial and Biological Neurons

A Schematic for Surrogate Backpropagation Technique

My research lies at the intersection of scientific machine learning, uncertainty quantification, and energy-efficient AI, with a strong emphasis on neuroscience-inspired models and operator learning for mechanics, dynamical systems, and reliability analysis. I have developed neuroscience-inspired neural operators, graph neural networks, and physics-informed frameworks that enable sparse, event-driven computation, making them suitable for edge computing, digital twins, and neuromorphic deployment. A central theme of my work is uncertainty quantification, integrating randomized priors, Bayesian inference, and conformal prediction to deliver predictions with calibrated, trustworthy bounds. My contributions span neural operators, spiking neural networks, physics-integrated learning, and reliability analysis, with applications to PDEs, materials modeling, and nonlinear dynamics, published in leading journals including  JCP,  JMPS, Probabilistic Engineering Mechanics, and Engineering Structures. 

*Please don't steal my gifs :P Thankyou :)
**Gifs made with the help of ChatGPT and Claude