I work with state-space models, linear attention mechanisms, and spiking neural networks to build stable and scalable representations of data. My work spans long-horizon sequence tasks and real-world time-series applications. I am also interested in continual learning, particularly parameter-efficient techniques that enable adaptive and scalable modeling.
“Curiosity kills a cat, not a Research Scholar.”
I’m Kartikay — endlessly curious about how the brain processes information and how those mechanisms can inspire efficient machine learning systems. I enjoy translating neuroscience-inspired ideas into mathematically grounded algorithms. After all, what’s an algorithm without math standing over it with a stick, demanding it justify itself?
When I’m not wrangling models or debugging code, you’ll probably find me on a stage — in theatre, in an amphitheatre, or with a guitar in hand. I try to balance logic with a little bit of drama. Please feel free to reach out to me.
