Ongoing
Constrained Lagrangian Neural Networks with Learned Physical Constraints
Boğaziçi University
Developing a neural framework that learns both system dynamics and hidden physical constraints directly from data.
Project detailsScientific work
Research
Current and completed work in scientific machine learning and theoretical physics.
Projects
Completed
Multi-Task Symbolic Regression for Gravitational Wave Law Discovery
Radboud University · High Energy Physics
Developed transformer-based symbolic regression models using contrastive learning to uncover physical laws from gravitational wave data.
Project detailsOngoing
Blackbox Hamiltonian Learning from Quantum Measurements
Boğaziçi University · Computer Engineering
Deep learning models for predicting quantum system dimensions and Hamiltonian parameters from minimal measurement data
Project detailsCompleted
Mutual Information Estimation in Quantum Many-Body Systems using Transformers
Forschungszentrum Jülich · PGI-8
Investigating mutual information as a proxy to predict the difficulty of learning wave functions using Neural Quantum States in quantum many-body systems
Project detailsCompleted
Lagrangian Neural Networks and Symbolic Regression
Boğaziçi University · Physics
Implementing physics-informed neural networks to learn theoretical quantities like the Lagrangian from trajectory data using symbolic regression
Project detailsResearch vision
My goal is to build tools that help theoretical scientists analyze natural-science data and express its structure mathematically and consistently. As experiments grow more complex, AI can help identify patterns that are difficult to recognize directly and make the process of scientific discovery more effective.
I am particularly interested in the mathematical side of scientific machine learning: using artificial intelligence to construct interpretable theoretical structures and move closer to fundamental descriptions of nature.
I believe scientists also have a responsibility to direct their work toward improving human wellbeing. The value and consequences of a project should remain part of how we choose what to pursue.