Add project structure section to main README

README.md

@@ -1 +1,20 @@
 # GENIE
+
+## Project Structure
+
+This repository contains multiple experimental subprojects under the **GENIE** initiative within ML4Sci. Each subproject explores different approaches to physics-informed learning, graph neural networks, and diffusion-based modeling in high-energy physics.
+
+### Subprojects
+
+- **Graph_Representation_Learning_Rushil_Singha**  
+  Implements a JetNet Graph Diffusion Model using graph neural networks and latent diffusion for particle-level jet generation.  
+  See project README: `Graph_Representation_Learning_Rushil_Singha/README.md`
+
+- **Non_local_Jet_Classification_Tanmay_Bakshi**  
+  Explores graph neural network methods for jet classification tasks using non-local message passing architectures.
+
+- **Physics_Informed_Neural_Network_Diffusion (PINNDE)**  
+  A Physics-Informed Neural Network approach for solving reverse-time diffusion equations.  
+  Developed as part of a GSoC'25 project, PINNDE aims to build a fast and reliable sampler for complex probability distributions by combining diffusion models with physics-informed neural networks. The project demonstrates promising results in 1D, 2D, and 3D distributions and serves as a foundation for fast particle jet simulation methods.
+
+Each folder represents an independent research or experimental project. Contributors are encouraged to read the README within each subproject directory for setup instructions, usage details, and implementation information.