Hello. I'm Shweta Prasad.
CV // Github // Twitter //
currently: shweta[dot]prasad@student[dot]uni-tuebingen[dot]de
always: shweta[dot]prasad[dot]402@gmail.com
I am from Thrissur, Kerala in India. Pictured behind me are coconut trees, which I cannot climb (yet). Leslie Lamport and an education in computer science have clouded my vision, as a result of which I am unable to read things that are not LaTeX formatted.
I am an IMPRS MMFD scholar pursuing an M.Sc in Neural and Behavioral Sciences at the Graduate Training Center for Neuroscience, University of Tübingen. I primarily work to understand corvid cognition at Andreas Nieder's lab in Tübingen. I also work as a research intern at Ziad Hafed's Active Vision Lab at the Werner Reichardt Center for Integrative Neuroscience.
I did my undergrad at Ashoka University, where I worked with Mahavir Jhawar and Subhashish Banerjee for my bachelor thesis on compression schemes that can incorporate deep probabilistic models of image data. While there, I was fortunate to work with Bittu Rajaraman and his neuroethology lab, where I studied call pattern production and recognition in bushcrickets endemic to the Western Ghats of India.
While my ultimate research interests are at the intersection of information theory, dynamical systems, self-organization, neuroscience and behaviour, I am currently exploring how the crow brain - fundamentally so different to the primate brain in its organisation - is able to produce comparable cognitive capabilities.
Here are some things that I've done :
-
how may serotonin affect patience? an average reward RL model for an interval timing task, based on a series of very cool papers from Miyazaki et. al
// report // slides // write me for the code -
simulating forest fires
// code -
an Asymmetrical Numeral Systems based compression codec, following this paper by Jarek Duda, et. al
// code P.S: this stuff is SOTA, my implementation is naïve but gives a good proof of concept, and is written to be super intuitive (hence Python) (the choice of language had nothing to do with last minute cramming and having barely a day to implement from scratch). -
the majority problem in the emergent phenomenon of stomatal patchiness
// webpage // report // code -
an interactive deep map of gendered migration in time during the India-Pakistan partition
// webpage -
some fun stuff to read, on
// the statistical theory of machine learning // Alice Gridlocked