Machine Learning for Science (ML4SCI)
Machine learning applications in science
GSoC Participation History
Technologies
Topics
Past Projects
Foundation models for End-to-End event reconstruction
Current deep learning models for CMS particle reconstruction are often task-specific. This project aims to develop versatile foundation models by...
Unsupervised Super-Resolution and Analysis of Real Lensing Images
This project develops an open-source, unsupervised super-resolution (SR) pipeline for enhancing the resolution of gravitational lensing images,...
Quantum Particle transformer for High Energy Physics Analysis at the LHC
This proposal aims to build upon and expand the progress made in recent years on Quantum Vision Transformers (QViTs) for High Energy Physics (HEP)...
Discovering Hidden Symmetries in CMS Calorimetric Data via Semi-Supervised Learning
This project tackles the challenge of learning hidden symmetries in complex, high-dimensional physics data using modern machine learning techniques....
Unsupervised super-resolution and analysis of observed lensing images
Machine learning based analyses are proving to be very potent in understanding complex physical systems. This project aims to develop a reinforcement...
Q-MAML for Variational Quantum Algorithms for High Energy Physics Analysis at the LHC
This project focuses on developing a new machine learning approach called Quantum Model-Agnostic Meta-Learning (Q-MAML) specifically for high-energy...
Physics Guided Machine Learning on Real Lensing Images
The project aims to develop a physics-informed neural network framework to infer the properties and distribution of dark matter in the lensing galaxy...
Discovery of hidden symmetries and conservation laws
Because of its cylindrical shape, the Compact Muon Solenoid (CMS) detector has intrinsic rotational symmetries. Data analysis can be greatly improved...
Physics-Informed Performer for Symbolic Squared Amplitudes in HEP
In high-energy physics (HEP), automating the symbolic computation of squared amplitudes is essential for predicting cross-sections and validating...
A Diffusion-Based Deep Learning Framework for Denoising Protoplanetary Disk Observations
This project proposes ProtoDiff, a diffusion-based deep learning model designed to denoise astronomical observations of protoplanetary disks from...
Continuous learning for high-energy physics data quality monitoring
Continuous learning serves as a powerful approach for maintaining reliable data quality monitoring in high-energy physics experiments facing evolving...
DeepLense: Gravitational Lens Finding Project
This project aims to develop and optimize deep learning algorithms for identifying strong gravitational lenses within wide-field surveys such as the...
Exoplanet Atmosphere Characterization
This project intends to develop cutting-edge machine-learning tools for spectral analysis to characterize the atmospheres of exoplanets. The project...
Discovery of hidden symmetries and conservation laws
This project aims to uncover fundamental symmetries and conserved laws hidden within complex high-energy physics data (specifically CMS datasets). It...
Diffusion Models for Gravitational Lensing Simulation
My project focuses on developing diffusion models for gravitational lensing simulations. I will implement and evaluate various diffusion-based...
Foundation Model for Gravitational Lensing
This project aims to develop a vision foundation model for strong gravitational lensing by comparing various self-supervised learning techniques,...
Latent Neural Signatures in Clinical vs. Neurotypical Dyads: A CEBRA Pipeline
This project aims to develop a cutting-edge computational pipeline—"Latent Neural Signatures in Clinical vs. Neurotypical Dyads: A CEBRA Pipeline"—to...
Data Processing Pipeline for the LSST
The project, "Data Processing Pipeline for the LSST," is designed to bridge the gap between the LSST data ecosystem and DeepLense's deep learning...
Foundation models for symbolic regression tasks
Problem: Symbolic regression is crucial for discovering underlying physical laws from data, but traditional methods are often computationally...
Neural Harmony – Decoding Social Interactions with CEBRA-based framework for analysing EEG data
This project aims to decode the neural dynamics of social interactions by adapting the CEBRA framework to analyze dyadic EEG data. Focusing on...
Quantum Diffusion Model for HEP
Diffusion models have experienced rapid growth in usability, availability, and research. The classical algorithm consists of two main parts: a...
Implementation of Quantum Generative Adversarial Networks to Perform HEP Analysis at the LHC
This project aims to implement a Quantum Generative Adversarial Network (QGAN) using the Pennylane framework to explore the advantages of quantum...
Building a Foundational Model for Symbolic Regression in High Energy Physics
This project proposes the development of a foundational model for symbolic regression tailored to high energy physics (HEP). Symbolic regression can...
State-space models for squared amplitude calculation in high-energy physics
One of the most important physical quantities in particle physics is the cross section, or a probability that a particular process takes place in the...
Quantum Kolmogorov-Arnold Networks for High Energy Physics Analysis at the LHC
This project explores Quantum Kolmogorov–Arnold Networks (QKANs) as a novel and interpretable architecture for analyzing collider data from the...
Next-Generation Transformer Models for Symbolic Calculations of Squared Amplitudes in HEP
In particle physics, a cross section is a measure of the likelihood that particles will interact or scatter with one another when they collide. It is...
Graph Representation Learning for Fast Detector Simulation
High-fidelity detector simulations are critical for accurate analysis in particle physics, but traditional Monte Carlo-based methods are...
Physics informed neural network diffusion equation
This project aims at incorporating Physics informed neural network (PINN) based ODE solvers into the Diffusion probabilistic models(DPM) , to build a...
Quantum Machine Learning For Exoplanet Characterization
This project aims to develop quantum machine learning models for exoplanet atmosphere characterization by leveraging the power of quantum neural...
Foundation Models for Exoplanet Characterization
The project aims at building foundational models suitable for characterizing the vast astronomical data and emphasizing their use case for various...
A Self-Supervised, Physics-Informed Hybrid Transformer Framework for Multi-Tasks in HEP
Accurately classifying particle collisions—whether distinguishing quark- from gluon-initiated jets or isolating Higgs events from complex...
Frequently Asked Questions
Machine Learning for Science (ML4SCI) | GSoC Org Profile & Stats - Learn about Machine Learning for Science (ML4SCI)'s involvement in Google Summer of Code (GSoC), their technologies, detailed reports.
Participation
Projects
Top Programming Languages
Project Difficulty Distribution
No difficulty data available
Built with ❤️ for the Google Summer of Code (GSoC) open source community | © 2026 GSoCOrganizationsGuide