Achin Parashar

I am currently a research assistant working in Computational Neuroscience at The Cognition, Computation and Behavior Lab at IISc Bangalore.

I'm interested in understanding how we make decisions. What is it that makes us behave in a certain way, can we nudge it with neurofeedback? And hence a selfish desire to be a strictly better decision maker. For that we need to know how our brains work !?

With interesting performance by the modern artificial neural networks, I'm also very curious to understand what sets them apart? What makes them uniquely suited for understanding the brain's computational processes? How do external signals translate into internal representations guiding decisions and actions, and how do high-level brain regions orchestrate lower-level processes to enable cognition?

email  /  github  

I've worked on decoding non-invasive neural data (fMRI and EEG). I've trained Machine Learning models to create subject specific fMRI encoders / decoders to algorithmically design images which activate some regions of the brain more that other naturalistic images, also termed as "super-stimuli".

	Designing salient naturalistic "super-stimuli" with deep generative models Mukherjee S,   Parashar A,   Murthy C,   Sridharan D CCN, 2024 video Generated attention grabbing naturalistic images which activate some regions of the brain much more than others.
	Rescuing catastrophic referral failures during automated diagnosis of domain-shifted medical images Srivastava A,   Biswas M,   Parashar A,   Sridharan D in prep, 2025 State-of-the-art: Uncertainty Quantification & Domain Generalization.
	Human-in-the-loop synthesis of behaviorally salient “super-distractors” Sen D,   Parashar A,   Ray A,   Chourasiya S,   Sridharan D CCN, 2025 (submitted) Creating novel classes of images optimized for specific behaviors, with a human-in-the-loop paradigm.

© Tantalize Zing!

Thanks Jon for the source code