From 76f79c886d5af5b5bdb05972d03f5b94a41b27af Mon Sep 17 00:00:00 2001 From: Pranath Reddy Kumbam Date: Fri, 27 Feb 2026 08:21:00 -0600 Subject: [PATCH 1/3] Update DeepLense Proposals - Add Test Link to all proposals - Reorder projects to match the order in the internal document - Fix affiliation --- _gsocproposals/2026/proposal_DEEPLENSE1.md | 43 +++++------ _gsocproposals/2026/proposal_DEEPLENSE2.md | 5 +- _gsocproposals/2026/proposal_DEEPLENSE3.md | 76 +++++++++++++++----- _gsocproposals/2026/proposal_DEEPLENSE4.md | 41 ++++++----- _gsocproposals/2026/proposal_DEEPLENSE5.md | 36 ++++------ _gsocproposals/2026/proposal_DEEPLENSE6.md | 33 +++++---- _gsocproposals/2026/proposal_DEEPLENSE7.md | 42 ++++------- _gsocproposals/2026/proposal_DEEPLENSE8.md | 83 ++++++---------------- 8 files changed, 172 insertions(+), 187 deletions(-) diff --git a/_gsocproposals/2026/proposal_DEEPLENSE1.md b/_gsocproposals/2026/proposal_DEEPLENSE1.md index f689ea29..3159703c 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE1.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE1.md @@ -1,5 +1,5 @@ --- -title: Foundation Model for Gravitational Lensing +title: Agentic AI for Autonomous Gravitational Lensing Simulation Workflows layout: gsoc_proposal project: DEEPLENSE project size: 350hr @@ -7,47 +7,50 @@ year: 2026 organization: - Alabama - MIT - - Florida - - PSL --- ## Description -Strong gravitational lensing is a powerful tool for studying dark matter and the large-scale structure of the universe. This project focuses on developing a vision foundation model specifically designed for lensing data, which can be fine-tuned for a variety of downstream tasks, including classification, super-resolution, regression, and lens finding. +Strong gravitational lensing is a powerful tool for studying dark matter and the large-scale structure of the universe. Current gravitational lensing simulation pipelines (e.g., [DeepLenseSim](https://github.com/mwt5345/DeepLenseSim) built on Lenstronomy) require substantial manual intervention: researchers must configure parameters, submit cluster jobs, validate outputs, retrain downstream ML models, and iterate on failures. This creates bottlenecks in large-scale dataset generation, limits exploration of parameter space, and consumes researcher time on engineering tasks rather than scientific analysis. +This project proposes the development of an Agentic AI framework to autonomously orchestrate gravitational lensing simulation workflows. Unlike traditional automation scripts that follow rigid rules, agentic systems employ LLM-powered agents that can reason about scientific objectives, adapt to failures, and coordinate complex multi-step workflows with minimal human supervision. -This project will explore different training strategies such as self-supervised learning, contrastive learning, or transformer-based models to learn meaningful representations of lensing images. By leveraging diverse datasets and training methodologies, the model will serve as a general-purpose backbone that can adapt to different astrophysical tasks while improving generalization across various observational conditions. ## Duration -Total project length: 350 hours. +Total project length: 175/350 hours. ## Difficulty level Advanced ## Task ideas - * Develop a pre-training strategy for learning robust representations of gravitational lensing data. - * Fine-tune the foundation model for multiple tasks such as classification, super-resolution, and regression. - * Evaluate the model's performance on different astrophysical datasets and benchmark against traditional methods. + * Survey agent frameworks and orchestration patterns + * Design multi-agent system with specialized agents + * Integrate with DeepLenseSim and SLURM + * Evaluate automation gains and dataset quality + ## Expected results - * A vision foundation model for gravitational lensing capable of being fine-tuned for various astrophysical tasks. - * Improved generalization and adaptability across different lensing datasets and observational setups. + * Working multi-agent system for autonomous workflow orchestration + * Significant reduction in manual intervention with adaptive behavior + * Benchmark study comparing against traditional pipelines + ## Requirements - * Python, PyTorch, experience with machine learning, and familiarity with computer vision techniques. - * Understanding of self-supervised learning, representation learning, and deep learning architectures. + * Prior experience with LLMs and local inference tools such as Llama.cpp or Ollama + * Prior experience with agentic AI frameworks such as Orchestral AI or Pydantic AI + * Prior experience with Reinforcement Learning (RL) and its application to AI agents, planning, or autonomous workflow optimization is highly preferred. + * Familiarity with gravitational lensing is preferred but not mandatory. + - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) - * [Pranath Reddy](mailto:ml4-sci@cern.ch) (University of Florida) - * [Anna Parul](mailto:ml4-sci@cern.ch) (Observatoire de Paris) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) + * [Rajat Shinde](mailto:ml4-sci@cern.ch) (University of Alabama in Huntsville) Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). @@ -56,5 +59,5 @@ Please **DO NOT** contact mentors directly by email. General questions can be di ## Links * [Paper 1](https://arxiv.org/abs/2008.12731) * [Paper 2](https://arxiv.org/abs/1909.07346) - * [Paper 3](https://arxiv.org/abs/2112.12121) - + * [Paper 3](https://arxiv.org/pdf/2512.15867) + \ No newline at end of file diff --git a/_gsocproposals/2026/proposal_DEEPLENSE2.md b/_gsocproposals/2026/proposal_DEEPLENSE2.md index 43c2b61d..6848c985 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE2.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE2.md @@ -7,7 +7,6 @@ year: 2026 organization: - Alabama - MIT - - Florida - PSL --- @@ -34,11 +33,13 @@ Advanced * Proficiency in Python & Pennylane or Qiskit. * Ability to work independently on research projects. +## Test +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) - * [Pranath Reddy](mailto:ml4-sci@cern.ch) (University of Florida) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) * [Rajat Shinde](mailto:ml4-sci@cern.ch) (University of Alabama in Huntsville) diff --git a/_gsocproposals/2026/proposal_DEEPLENSE3.md b/_gsocproposals/2026/proposal_DEEPLENSE3.md index 20a50888..c93ca906 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE3.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE3.md @@ -1,8 +1,8 @@ --- -title: Physics Guided Machine Learning on Real Lensing Images +title: Neural Operators for Fast Simulation of Strong Gravitational Lensing layout: gsoc_proposal project: DEEPLENSE -project size: 175hr/350hr +project size: 350hr year: 2026 organization: - Alabama @@ -11,7 +11,17 @@ organization: ## Description -This project focuses on developing a Physics-Informed Neural Network (PINN) framework for analyzing real strong gravitational lensing datasets to study dark matter distribution. Strong gravitational lensing, a key prediction of general relativity, occurs when a massive galaxy or cluster bends light from a background source, creating arcs or Einstein rings. Traditional algorithms struggle or fail entirely when applied to real lensing datasets due to observational complexities and noise. By leveraging PINNs, the project will integrate physical laws directly into the learning process, enhancing the accuracy and interpretability of dark matter inferences. The model will be trained on real lensing images, incorporating observational constraints to refine mass distribution estimates and improve the efficiency of dark matter studies. +Strong gravitational lensing is a powerful probe of dark matter substructure and cosmological structure formation. Current simulation pipelines (e.g., Lenstronomy-based ray tracing) require solving the lens equation repeatedly for different mass distributions and source configurations. While accurate, these simulations are computationally expensive and limit large-scale parameter sweeps, uncertainty quantification, and real-time inference. +This project proposes the development of a Neural Operator framework to learn the functional mapping between mass distributions and lensed images. Unlike traditional neural networks that operate on finite-dimensional vectors, neural operators learn mappings between infinite-dimensional function spaces. This makes them particularly well-suited for approximating solutions of physical systems governed by partial differential equations. +The goal is to train a neural operator that directly maps: + * Lens mass distribution → Lensed image + or + * (Mass distribution, source light profile) → Observed lensed image + + +Such a model would serve as a fast surrogate simulator capable of producing high-fidelity lensing outputs at a fraction of the computational cost of traditional ray-tracing solvers. +This would represent the first exploration of neural operators within the ML4SCI DeepLense ecosystem. + ## Duration @@ -19,27 +29,59 @@ Total project length: 175/350 hours. ## Difficulty level -Intermediate/Advanced +Advanced ## Task ideas - * Build various physics-informed neural network architectures that are endowed with known physics for real lensing datasets. - * Apply these models to study dark matter in strong lensing images in various contexts: classification, regression, anomaly detection, and more. - +1. Literature Study + 1. Study Fourier Neural Operators (FNO) + 2. Study DeepONets + 3. Review operator learning in physical systems +2. Neural Operator Implementation + 1. Implement Fourier Neural Operator for 2D fields + 2. Compare with DeepONet-style architectures + 3. Explore spectral vs spatial operator parameterizations +3. Evaluation + 1. Compare speed vs traditional solver + 2. Measure pixel-wise reconstruction error + 3. Evaluate preservation of physical invariants + 4. Test generalization across: + 1. Different mass profiles + 2. Different redshifts + 3. Different source morphologies +4. Extensions (if time permits) + 1. Conditional neural operators + 2. Uncertainty-aware operator learning + 3. Physics-informed operator constraints + 4. Hybrid operator + diffusion refinement + + + + ## Expected results - * A more capable architecture that can operate on a wider variety of lensing images, including lensing images created with real galaxy datasets. - * Insight into the lensing systems, and their sub-structures. + * A neural operator capable of approximating the lensing simulation operator. + * Significant speedup compared to ray-tracing pipelines. + * Demonstration of generalization across varying astrophysical conditions. + * A benchmark study comparing neural operators with CNN-based surrogate models. + + ## Requirements -Python, PyTorch, experience with machine learning, knowledge of computer vision techniques, familiarity with autoencoders. +Python, PyTorch, experience with machine learning and deep learning. +Partial understanding of: + * Representation learning + * Spectral methods + * Partial differential equations (basic familiarity) + * Scientific machine learning concepts +Familiarity with gravitational lensing is preferred but not mandatory. + - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) * [Ashutosh Ojha](mailto:ml4-sci@cern.ch) (Indian Institute of Technology (Indian School of Mines), Dhanbad) @@ -47,7 +89,7 @@ Please **DO NOT** contact mentors directly by email. General questions can be di ## Links - * [Paper 1](https://arxiv.org/abs/2008.12731) - * [Paper 2](https://arxiv.org/abs/1909.07346) - * [Paper 3](https://ml4physicalsciences.github.io/2024/files/NeurIPS_ML4PS_2024_78.pdf) - * [Paper 4](https://ml4physicalsciences.github.io/2025/files/NeurIPS_ML4PS_2025_252.pdf) + * [Paper 1](https://arxiv.org/abs/2010.08895) + * [Paper 2](https://arxiv.org/abs/2108.08481) + * [Paper 3](https://arxiv.org/abs/1910.03193) + \ No newline at end of file diff --git a/_gsocproposals/2026/proposal_DEEPLENSE4.md b/_gsocproposals/2026/proposal_DEEPLENSE4.md index 6fe6f998..9cd32162 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE4.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE4.md @@ -1,8 +1,8 @@ --- -title: Gravitational Lens Finding +title: Foundation Model for Gravitational Lensing layout: gsoc_proposal project: DEEPLENSE -project size: 175hr/350hr +project size: 350hr year: 2026 organization: - Alabama @@ -12,37 +12,38 @@ organization: ## Description -This project focuses on the task of lens finding in the currently available wide-field surveys (e.g., HSC-SSP). The expected number of strong lenses in the large surveys is significantly overpowered by the number of non-lensed objects, which leads to the high number of false positives in typical lens searches. +Strong gravitational lensing is a powerful tool for studying dark matter and the large-scale structure of the universe. This project focuses on developing a vision foundation model specifically designed for lensing data, which can be fine-tuned for a variety of downstream tasks, including classification, super-resolution, regression, and lens finding. -The goal of the project is to develop lens finding algorithms, apply them to the observational data, and assess the limitations of the algorithms (for example, analyse the properties of the identified lens population and examine the typical contaminants). +This project will explore different training strategies such as self-supervised learning, contrastive learning, or transformer-based models to learn meaningful representations of lensing images. By leveraging diverse datasets and training methodologies, the model will serve as a general-purpose backbone that can adapt to different astrophysical tasks while improving generalization across various observational conditions. ## Duration -Total project length: 175/350 hours. +Total project length: 350 hours. ## Difficulty level -Intermediate/Advanced + +Advanced ## Task ideas - * Start with architectures previously explored within the DeepLense project and optimise them for the lens finding task - * Perform the lens search in the real observational data and analyse the properties of the detected lens candidates - * Evaluate model performance on different surveys + * Develop a pre-training strategy for learning robust representations of gravitational lensing data. + * Fine-tune the foundation model for multiple tasks such as classification, super-resolution, and regression. + * Evaluate the model's performance on different astrophysical datasets and benchmark against traditional methods. ## Expected results - * Increase the number of known strong lenses. - * Insight into properties of the identified lens candidates. - + * A vision foundation model for gravitational lensing capable of being fine-tuned for various astrophysical tasks. + * Improved generalization and adaptability across different lensing datasets and observational setups. + ## Requirements -Python, PyTorch, experience with machine learning, familiarity with astrophysics datasets. + * Python, PyTorch, experience with machine learning, and familiarity with computer vision techniques. + * Understanding of self-supervised learning, representation learning, and deep learning architectures. - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors - * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) + * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) * [Anna Parul](mailto:ml4-sci@cern.ch) (Observatoire de Paris) @@ -50,5 +51,7 @@ Please **DO NOT** contact mentors directly by email. General questions can be di ## Links - * [Paper 1](https://arxiv.org/abs/1909.07346) - * [Paper 2](https://ml4physicalsciences.github.io/2024/files/NeurIPS_ML4PS_2024_107.pdf) + * [Paper 1](https://arxiv.org/abs/2008.12731) + * [Paper 2](https://arxiv.org/abs/1909.07346) + * [Paper 3](https://arxiv.org/abs/2112.12121) + diff --git a/_gsocproposals/2026/proposal_DEEPLENSE5.md b/_gsocproposals/2026/proposal_DEEPLENSE5.md index 0b2b707a..12a2290d 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE5.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE5.md @@ -1,5 +1,5 @@ --- -title: Physics-Informed Diffusion Models for Gravitational Lensing Simulation +title: Physics Guided Machine Learning on Real Lensing Images layout: gsoc_proposal project: DEEPLENSE project size: 175hr/350hr @@ -7,45 +7,40 @@ year: 2026 organization: - Alabama - MIT - - Florida --- ## Description -Strong gravitational lensing is a promising probe of the substructure of dark matter to better understand its underlying nature. Deep learning methods have the potential to accurately identify images containing substructure and differentiate WIMP particle dark matter from other well-motivated models, including axions and axion-like particles, warm dark matter, etc. -This project proposes the development of physics-informed diffusion and flow matching models for simulating strong gravitational lensing images. Gravitational lensing is governed by well-defined physical laws and symmetries. We aim to explore strategies for encoding these physical structures directly into generative models so that outputs are not merely statistically realistic but physically consistent. Possible directions include incorporating symmetry-aware architectures, parameterizing the generation process through physically meaningful intermediate representations (e.g., convergence and shear maps), enforcing governing equations as soft or hard constraints, or other approaches the contributor may propose. -The resulting model is intended to produce high-fidelity, physics-compliant lensing simulations to augment training data for substructure detection and dark matter classification, while offering stronger generalization and a narrower sim-to-real gap compared to purely data-driven generative approaches. +This project focuses on developing a Physics-Informed Neural Network (PINN) framework for analyzing real strong gravitational lensing datasets to study dark matter distribution. Strong gravitational lensing, a key prediction of general relativity, occurs when a massive galaxy or cluster bends light from a background source, creating arcs or Einstein rings. Traditional algorithms struggle or fail entirely when applied to real lensing datasets due to observational complexities and noise. By leveraging PINNs, the project will integrate physical laws directly into the learning process, enhancing the accuracy and interpretability of dark matter inferences. The model will be trained on real lensing images, incorporating observational constraints to refine mass distribution estimates and improve the efficiency of dark matter studies. ## Duration Total project length: 175/350 hours. ## Difficulty level + Intermediate/Advanced ## Task ideas - * Explore diffusion models for the generation of strong gravitational lensing images. - * Benchmark generated images against traditional simulations for physical consistency and downstream task utility. - * Create a diverse dataset of simulated gravitational lensing images under various astrophysical conditions. - - + * Build various physics-informed neural network architectures that are endowed with known physics for real lensing datasets. + * Apply these models to study dark matter in strong lensing images in various contexts: classification, regression, anomaly detection, and more. + ## Expected results - * A diffusion model capable of generating realistic, physically consistent simulations of strong gravitational lensing phenomena. + * A more capable architecture that can operate on a wider variety of lensing images, including lensing images created with real galaxy datasets. + * Insight into the lensing systems, and their sub-structures. ## Requirements - * Python, PyTorch and relevant past experience in Machine Learning. - * Familiarity with astrophysics and gravitational lensing is preferred but not required. +Python, PyTorch, experience with machine learning, knowledge of computer vision techniques, familiarity with autoencoders. - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) - * [Pranath Reddy](mailto:ml4-sci@cern.ch) (University of Florida) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) - * [Hamees Sayed](mailto:ml4-sci@cern.ch) (IIT Madras) + * [Ashutosh Ojha](mailto:ml4-sci@cern.ch) (Indian Institute of Technology (Indian School of Mines), Dhanbad) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) + Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). @@ -53,6 +48,5 @@ Please **DO NOT** contact mentors directly by email. General questions can be di ## Links * [Paper 1](https://arxiv.org/abs/2008.12731) * [Paper 2](https://arxiv.org/abs/1909.07346) - * [Paper 3](https://arxiv.org/abs/2112.12121) - * [Paper 4](https://arxiv.org/abs/2203.17003) - * [Paper 5](https://arxiv.org/abs/2211.03812) + * [Paper 3](https://ml4physicalsciences.github.io/2024/files/NeurIPS_ML4PS_2024_78.pdf) + * [Paper 4](https://ml4physicalsciences.github.io/2025/files/NeurIPS_ML4PS_2025_252.pdf) diff --git a/_gsocproposals/2026/proposal_DEEPLENSE6.md b/_gsocproposals/2026/proposal_DEEPLENSE6.md index adca2df7..02feb183 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE6.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE6.md @@ -1,53 +1,52 @@ --- -title: Data Processing Pipeline for the LSST +title: Gravitational Lens Finding layout: gsoc_proposal project: DEEPLENSE project size: 175hr/350hr year: 2026 organization: - Alabama - - Brown - MIT - PSL --- ## Description -The Rubin Observatory will provide an unprecedented volume of astronomical data, accessible via a dedicated open-source software suite consisting of data reduction pipelines and tools for interacting with calibrated images and catalogs. In order to prepare for the application of DeepLense methods on upcoming LSST data, this project focuses on developing a complementary pipeline that integrates LSST’s data access tools with DeepLense workflows. This pipeline will enable efficient data retrieval, preprocessing, and adaptation for various DeepLense applications such as lens finding, super-resolution, and classification. +This project focuses on the task of lens finding in the currently available wide-field surveys (e.g., HSC-SSP). The expected number of strong lenses in the large surveys is significantly overpowered by the number of non-lensed objects, which leads to the high number of false positives in typical lens searches. + +The goal of the project is to develop lens finding algorithms, apply them to the observational data, and assess the limitations of the algorithms (for example, analyse the properties of the identified lens population and examine the typical contaminants). ## Duration Total project length: 175/350 hours. ## Difficulty level - Intermediate/Advanced ## Task ideas - * Explore the existing LSST data access tools and design the workflow to provide the data for the DeepLense tasks. - * Test the workflow on the mock surveys provided by The Rubin Observatory. + * Start with architectures previously explored within the DeepLense project and optimise them for the lens finding task + * Perform the lens search in the real observational data and analyse the properties of the detected lens candidates + * Evaluate model performance on different surveys ## Expected results - * A functional pipeline capable of interfacing LSST data with DeepLense applications. - + * Increase the number of known strong lenses. + * Insight into properties of the identified lens candidates. + ## Requirements -Python, familiarity with astronomical data processing, and understanding of data access APIs and pipelines. +Python, PyTorch, experience with machine learning, familiarity with astrophysics datasets. - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors - * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) + * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Anna Parul](mailto:ml4-sci@cern.ch) (Observatoire de Paris) - * [Lucca Paris](mailto:ml4-sci@cern.ch) (Brown University) + Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). ## Links - * [LSST Pipeline Docs](https://pipelines.lsst.io) - * [Paper 1](https://arxiv.org/abs/2008.12731) - * [Paper 2](https://arxiv.org/abs/1909.07346) + * [Paper 1](https://arxiv.org/abs/1909.07346) + * [Paper 2](https://ml4physicalsciences.github.io/2024/files/NeurIPS_ML4PS_2024_107.pdf) diff --git a/_gsocproposals/2026/proposal_DEEPLENSE7.md b/_gsocproposals/2026/proposal_DEEPLENSE7.md index 95961cda..fa7453f9 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE7.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE7.md @@ -1,19 +1,19 @@ --- -title: Agentic AI for Autonomous Gravitational Lensing Simulation Workflows +title: Data Processing Pipeline for the LSST layout: gsoc_proposal project: DEEPLENSE -project size: 350hr +project size: 175hr/350hr year: 2026 organization: - Alabama + - Brown - MIT + - PSL --- ## Description -Strong gravitational lensing is a powerful tool for studying dark matter and the large-scale structure of the universe. Current gravitational lensing simulation pipelines (e.g., [DeepLenseSim](https://github.com/mwt5345/DeepLenseSim) built on Lenstronomy) require substantial manual intervention: researchers must configure parameters, submit cluster jobs, validate outputs, retrain downstream ML models, and iterate on failures. This creates bottlenecks in large-scale dataset generation, limits exploration of parameter space, and consumes researcher time on engineering tasks rather than scientific analysis. -This project proposes the development of an Agentic AI framework to autonomously orchestrate gravitational lensing simulation workflows. Unlike traditional automation scripts that follow rigid rules, agentic systems employ LLM-powered agents that can reason about scientific objectives, adapt to failures, and coordinate complex multi-step workflows with minimal human supervision. - +The Rubin Observatory will provide an unprecedented volume of astronomical data, accessible via a dedicated open-source software suite consisting of data reduction pipelines and tools for interacting with calibrated images and catalogs. In order to prepare for the application of DeepLense methods on upcoming LSST data, this project focuses on developing a complementary pipeline that integrates LSST’s data access tools with DeepLense workflows. This pipeline will enable efficient data retrieval, preprocessing, and adaptation for various DeepLense applications such as lens finding, super-resolution, and classification. ## Duration @@ -21,45 +21,31 @@ Total project length: 175/350 hours. ## Difficulty level -Advanced +Intermediate/Advanced ## Task ideas - * Survey agent frameworks and orchestration patterns - * Design multi-agent system with specialized agents - * Integrate with DeepLenseSim and SLURM - * Evaluate automation gains and dataset quality - + * Explore the existing LSST data access tools and design the workflow to provide the data for the DeepLense tasks. + * Test the workflow on the mock surveys provided by The Rubin Observatory. ## Expected results - * Working multi-agent system for autonomous workflow orchestration - * Significant reduction in manual intervention with adaptive behavior - * Benchmark study comparing against traditional pipelines - + * A functional pipeline capable of interfacing LSST data with DeepLense applications. ## Requirements - * Prior experience with LLMs and local inference tools such as Llama.cpp or Ollama - * Prior experience with agentic AI frameworks such as Orchestral AI or Pydantic AI - * Prior experience with Reinforcement Learning (RL) and its application to AI agents, planning, or autonomous workflow optimization is highly preferred. - * Familiarity with gravitational lensing is preferred but not mandatory. +Python, familiarity with astronomical data processing, and understanding of data access APIs and pipelines. - - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) - * [Pranath Reddy](mailto:ml4-sci@cern.ch) (University of Florida) - * [Rajat Shinde](mailto:ml4-sci@cern.ch) (University of Alabama in Huntsville) - + * [Anna Parul](mailto:ml4-sci@cern.ch) (Observatoire de Paris) + * [Lucca Paris](mailto:ml4-sci@cern.ch) (Brown University) Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). ## Links + * [LSST Pipeline Docs](https://pipelines.lsst.io) * [Paper 1](https://arxiv.org/abs/2008.12731) * [Paper 2](https://arxiv.org/abs/1909.07346) - * [Paper 3](https://arxiv.org/pdf/2512.15867) - \ No newline at end of file diff --git a/_gsocproposals/2026/proposal_DEEPLENSE8.md b/_gsocproposals/2026/proposal_DEEPLENSE8.md index b163c68b..e159b87a 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE8.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE8.md @@ -1,8 +1,8 @@ --- -title: Neural Operators for Fast Simulation of Strong Gravitational Lensing +title: Physics-Informed Diffusion Models for Gravitational Lensing Simulation layout: gsoc_proposal project: DEEPLENSE -project size: 350hr +project size: 175hr/350hr year: 2026 organization: - Alabama @@ -11,88 +11,45 @@ organization: ## Description -Strong gravitational lensing is a powerful probe of dark matter substructure and cosmological structure formation. Current simulation pipelines (e.g., Lenstronomy-based ray tracing) require solving the lens equation repeatedly for different mass distributions and source configurations. While accurate, these simulations are computationally expensive and limit large-scale parameter sweeps, uncertainty quantification, and real-time inference. -This project proposes the development of a Neural Operator framework to learn the functional mapping between mass distributions and lensed images. Unlike traditional neural networks that operate on finite-dimensional vectors, neural operators learn mappings between infinite-dimensional function spaces. This makes them particularly well-suited for approximating solutions of physical systems governed by partial differential equations. -The goal is to train a neural operator that directly maps: - * Lens mass distribution → Lensed image - or - * (Mass distribution, source light profile) → Observed lensed image - - -Such a model would serve as a fast surrogate simulator capable of producing high-fidelity lensing outputs at a fraction of the computational cost of traditional ray-tracing solvers. -This would represent the first exploration of neural operators within the ML4SCI DeepLense ecosystem. - +Strong gravitational lensing is a promising probe of the substructure of dark matter to better understand its underlying nature. Deep learning methods have the potential to accurately identify images containing substructure and differentiate WIMP particle dark matter from other well-motivated models, including axions and axion-like particles, warm dark matter, etc. +This project proposes the development of physics-informed diffusion and flow matching models for simulating strong gravitational lensing images. Gravitational lensing is governed by well-defined physical laws and symmetries. We aim to explore strategies for encoding these physical structures directly into generative models so that outputs are not merely statistically realistic but physically consistent. Possible directions include incorporating symmetry-aware architectures, parameterizing the generation process through physically meaningful intermediate representations (e.g., convergence and shear maps), enforcing governing equations as soft or hard constraints, or other approaches the contributor may propose. +The resulting model is intended to produce high-fidelity, physics-compliant lensing simulations to augment training data for substructure detection and dark matter classification, while offering stronger generalization and a narrower sim-to-real gap compared to purely data-driven generative approaches. ## Duration Total project length: 175/350 hours. ## Difficulty level - -Advanced +Intermediate/Advanced ## Task ideas -1. Literature Study - 1. Study Fourier Neural Operators (FNO) - 2. Study DeepONets - 3. Review operator learning in physical systems -2. Neural Operator Implementation - 1. Implement Fourier Neural Operator for 2D fields - 2. Compare with DeepONet-style architectures - 3. Explore spectral vs spatial operator parameterizations -3. Evaluation - 1. Compare speed vs traditional solver - 2. Measure pixel-wise reconstruction error - 3. Evaluate preservation of physical invariants - 4. Test generalization across: - 1. Different mass profiles - 2. Different redshifts - 3. Different source morphologies -4. Extensions (if time permits) - 1. Conditional neural operators - 2. Uncertainty-aware operator learning - 3. Physics-informed operator constraints - 4. Hybrid operator + diffusion refinement - - + * Explore diffusion models for the generation of strong gravitational lensing images. + * Benchmark generated images against traditional simulations for physical consistency and downstream task utility. + * Create a diverse dataset of simulated gravitational lensing images under various astrophysical conditions. ## Expected results - * A neural operator capable of approximating the lensing simulation operator. - * Significant speedup compared to ray-tracing pipelines. - * Demonstration of generalization across varying astrophysical conditions. - * A benchmark study comparing neural operators with CNN-based surrogate models. - - + * A diffusion model capable of generating realistic, physically consistent simulations of strong gravitational lensing phenomena. ## Requirements -Python, PyTorch, experience with machine learning and deep learning. -Partial understanding of: - * Representation learning - * Spectral methods - * Partial differential equations (basic familiarity) - * Scientific machine learning concepts -Familiarity with gravitational lensing is preferred but not mandatory. - + * Python, PyTorch and relevant past experience in Machine Learning. + * Familiarity with astrophysics and gravitational lensing is preferred but not required. - - +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. ## Mentors * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) - * [Pranath Reddy](mailto:ml4-sci@cern.ch) (University of Florida) - * [Ashutosh Ojha](mailto:ml4-sci@cern.ch) (Indian Institute of Technology (Indian School of Mines), Dhanbad) - + * [Hamees Sayed](mailto:ml4-sci@cern.ch) (IIT Madras) Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). ## Links - * [Paper 1](https://arxiv.org/abs/2010.08895) - * [Paper 2](https://arxiv.org/abs/2108.08481) - * [Paper 3](https://arxiv.org/abs/1910.03193) - \ No newline at end of file + * [Paper 1](https://arxiv.org/abs/2008.12731) + * [Paper 2](https://arxiv.org/abs/1909.07346) + * [Paper 3](https://arxiv.org/abs/2112.12121) + * [Paper 4](https://arxiv.org/abs/2203.17003) + * [Paper 5](https://arxiv.org/abs/2211.03812) From d08064d4a76521310ef07696ccfe382238c58d01 Mon Sep 17 00:00:00 2001 From: Pranath Reddy Kumbam Date: Fri, 27 Feb 2026 10:34:32 -0600 Subject: [PATCH 2/3] Update proposal_DEEPLENSE2.md - Add QML project tasks and expected results --- _gsocproposals/2026/proposal_DEEPLENSE2.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/_gsocproposals/2026/proposal_DEEPLENSE2.md b/_gsocproposals/2026/proposal_DEEPLENSE2.md index 6848c985..12b32815 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE2.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE2.md @@ -24,8 +24,12 @@ Total project length: 175/350 hours. Advanced ## Task ideas + * Implement hybrid quantum-classical architectures (variational quantum circuits, quantum kernels) for dark matter substructure classification from strong lensing images. + * Benchmark quantum models against classical baselines and evaluate noise robustness under realistic NISQ conditions. ## Expected results + * Trained hybrid quantum-classical model for multi-class lensing classification. + * Benchmark of quantum vs. classical performance with analysis of near-term hardware feasibility. ## Requirements * Strong background in Machine Learning & Deep Learning. From a66a743a021369f59f78a9408d58d5afc9a3a065 Mon Sep 17 00:00:00 2001 From: Pranath Reddy Kumbam Date: Fri, 27 Feb 2026 11:09:52 -0600 Subject: [PATCH 3/3] Add DeepLense Super-Resolution proposal - Add the missing super resolution proposal to the list --- _gsocproposals/2026/proposal_DEEPLENSE1.md | 1 + _gsocproposals/2026/proposal_DEEPLENSE2.md | 1 + _gsocproposals/2026/proposal_DEEPLENSE3.md | 1 + _gsocproposals/2026/proposal_DEEPLENSE5.md | 1 + _gsocproposals/2026/proposal_DEEPLENSE8.md | 1 + _gsocproposals/2026/proposal_DEEPLENSE9.md | 52 ++++++++++++++++++++++ 6 files changed, 57 insertions(+) create mode 100644 _gsocproposals/2026/proposal_DEEPLENSE9.md diff --git a/_gsocproposals/2026/proposal_DEEPLENSE1.md b/_gsocproposals/2026/proposal_DEEPLENSE1.md index 3159703c..35eca2eb 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE1.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE1.md @@ -7,6 +7,7 @@ year: 2026 organization: - Alabama - MIT + - UAH --- ## Description diff --git a/_gsocproposals/2026/proposal_DEEPLENSE2.md b/_gsocproposals/2026/proposal_DEEPLENSE2.md index 12b32815..2b95cb8d 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE2.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE2.md @@ -8,6 +8,7 @@ organization: - Alabama - MIT - PSL + - UAH --- ## Description diff --git a/_gsocproposals/2026/proposal_DEEPLENSE3.md b/_gsocproposals/2026/proposal_DEEPLENSE3.md index c93ca906..40ffef5e 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE3.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE3.md @@ -7,6 +7,7 @@ year: 2026 organization: - Alabama - MIT + - IIT(ISM) Dhanbad --- ## Description diff --git a/_gsocproposals/2026/proposal_DEEPLENSE5.md b/_gsocproposals/2026/proposal_DEEPLENSE5.md index 12a2290d..677ba826 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE5.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE5.md @@ -7,6 +7,7 @@ year: 2026 organization: - Alabama - MIT + - IIT(ISM) Dhanbad --- ## Description diff --git a/_gsocproposals/2026/proposal_DEEPLENSE8.md b/_gsocproposals/2026/proposal_DEEPLENSE8.md index e159b87a..b5be686b 100644 --- a/_gsocproposals/2026/proposal_DEEPLENSE8.md +++ b/_gsocproposals/2026/proposal_DEEPLENSE8.md @@ -7,6 +7,7 @@ year: 2026 organization: - Alabama - MIT + - IIT Madras --- ## Description diff --git a/_gsocproposals/2026/proposal_DEEPLENSE9.md b/_gsocproposals/2026/proposal_DEEPLENSE9.md new file mode 100644 index 00000000..eef9c569 --- /dev/null +++ b/_gsocproposals/2026/proposal_DEEPLENSE9.md @@ -0,0 +1,52 @@ +--- +title: Unsupervised Super-Resolution and Analysis of Real Lensing Images +layout: gsoc_proposal +project: DEEPLENSE +project size: 175hr/350hr +year: 2026 +organization: + - Alabama + - MIT + - IIT Madras +--- + +## Description + +This project’s aims are twofold: developing an unsupervised super-resolution architecture to upscale the quality of lensing images constructed using real galaxy sources, and to obtain insight about the lenses themselves. An unsupervised super-resolution technique could be very valuable for lensing studies as access to high resolution lensing images for training and study can be limited, especially given potential lensing data from upcoming surveys such as Euclid and LSST. The overall goal of this project is to develop an architecture that can better study the characteristics of the gravitational lenses and their substructure. + +## Duration + +Total project length: 175/350 hours. + +## Difficulty level +Intermediate/Advanced + +## Task ideas + * Start with unsupervised SR of simulated images and think of ways to bridge the gap to real images. + * Try then integrating into the pipeline modules that study characteristics of the lenses. + + +## Expected results + * A more capable architecture that can operate on a wider variety of lensing images, including lensing images created with real galaxy datasets. + * Insight into the lensing systems, and their sub-structures + +## Requirements + * Python, PyTorch and relevant past experience in Machine Learning. + * Familiarity with astrophysics and gravitational lensing is preferred but not required. + +## Test +Please use this [link](https://docs.google.com/document/d/10APh49fvayGoSftzO4fGXs2HP3uvYSzG-fSrq4xHL1w/edit?usp=sharing) to access the test for this project. + +## Mentors + * [Michael Toomey](mailto:ml4-sci@cern.ch) (Massachusetts Institute of Technology) + * [Pranath Reddy](mailto:ml4-sci@cern.ch) (Independent Researcher) + * [Sergei Gleyzer](mailto:ml4-sci@cern.ch) (University of Alabama) + * [Hamees Sayed](mailto:ml4-sci@cern.ch) (IIT Madras) + +Please **DO NOT** contact mentors directly by email. General questions can be directed to [ml4-sci@cern.ch](mailto:ml4-sci@cern.ch). To submit your proposal, CV, and test task solutions, please use [this Google form](https://forms.gle/SPXo8kSwHHptcBmk9). + + +## Links + * [Paper 1](https://arxiv.org/abs/2008.12731) + * [Paper 2](https://arxiv.org/abs/1909.07346) + * [Paper 3](https://iopscience.iop.org/article/10.1088/2632-2153/ad76f8/meta)