Building
Intelligent
Systems.
I am Vedant, a Computer Engineering student researching machine learning behavior, building deep browser simulations, and engineering low-level software utilities.
Explainable
Mechanics.
My work begins at the intersection of mathematical theory and systems implementation. I prefer projects where mechanics are visible, behavior is explainable, and architecture is intentional.
Currently focusing on symbolic regression, operation neural networks, and spatial simulations.
Glassbox is a neuro-symbolic search engine that uses operation neural networks to discover mathematical equations directly from data. By enforcing algebraic structure during training, it ensures that output models are inherently interpretable rather than black-box approximations.
View RepositoryA browser-based gravitational simulation focused on visible behavior and explainable dynamics. It processes hundreds of orbital bodies in real-time, leveraging Web Workers for physics calculations and a minimalist interface for interaction.
View RepositoryA minimalist utility built in Go engineered for practical, daily developer use. It focuses on low latency, rigorous error handling, and a zero-dependency deployment model to solve immediate terminal-based workflows.
View RepositoryBased in the intersection of theory and deployment. Open for discussion on systems programming, neuro-symbolic research, and well-designed tooling.