Physics-Informed Neural Networks (PINN) Solvers of (Partial) Differential Equations for Scientific Machine Learning (SciML) accelerated simulation

Pre-built implicit layer architectures with O(1) backprop, GPUs, and stiff+non-stiff DE solvers, demonstrating scientific machine learning (SciML) and physics-informed machine learning methods

Automatic Finite Difference PDE solving with Julia SciML

Build and simulate jump equations like Gillespie simulations and jump diffusions with constant and state-dependent rates and mix with differential equations and scientific machine learning (SciML)

Extension functionality which uses Stan.jl, DynamicHMC.jl, and Turing.jl to estimate the parameters to differential equations and perform Bayesian probabilistic scientific machine learning

A framework for developing multi-scale arrays for use in scientific machine learning (SciML) simulations

Taylor mode automatic differentiation (jets) in PyTorch

Code Repository for the paper "Mechanistic Neural Networks for Scientific Machine Learning", ICML 2024

Elucidating the Design Choice of Probability Paths in Flow Matching for Forecasting

DDPM-based U-Net transformer for fluid dynamics prediction, reproducing and extending DiffFluid. Validated on Navier-Stokes vorticity and Lattice Boltzmann (D2Q9) with corrected noise-prediction loss formulation.

Published PyPI package for ArXiv embedding benchmarks, retrieval evaluation, and scientific RAG experiments.

Structured ecosystem of 190+ AI systems spanning foundation models, agentic reasoning, reinforcement learning, generative architectures, scientific ML, and self-evolving neural systems. A research-driven lab exploring scalable intelligence.

Neural ODE-based State of Charge (SOC) estimation for Li-ion batteries using the NASA Battery Dataset. Built as a weekend project to explore learned dynamics for battery modeling, with visualizations designed for engineering audiences.

🔬 Predict molecular melting points with a robust machine learning pipeline that prioritizes reproducibility and efficient data handling.

Physics-informed neural network solving Maxwell's equations in ICP reactors. Hard BC ansatz, transfer learning across geometries, autograd sensitivity maps. 1700× faster than FEM. Built with NVIDIA Modulus + PyTorch.

Parametric PINN surrogate for flow over cylinders — 95% accuracy, 95%+ CFD speedup, <50ms inference

GPU-accelerated, fault-tolerant Schlieren/PIV shock tracking with interactive ROI, 1-px edges, and resumable training.

Solving 1D beam, 2D plate, and 3D solid mechanics problems using Physics-Informed Neural Networks (PINNs) with DeepXDE and TensorFlow. Results validated against ANSYS Mechanical 2025 R1. 2nd sem project

Physics-aware neural surrogate for black hole accretion flow (GRMHD-like) using Fourier Neural Operators.

Physics-Informed Neural Networks with passivity constraints and ensemble uncertainty quantification for nonlinear inverse modeling.