Summary
Shanka Mondal is a Machine Learning Engineer and Princeton PhD student specializing in computer vision, deep learning, abstract reasoning, and planning within large language models, working at the intersection of cognitive neuroscience and ML. Advised by Jonathan D. Cohen, she builds neuroscience-inspired neural architectures that aim for systematic generalization in visual abstract reasoning and has extended those ideas to multi-agent LLM planning modules in consecutive Microsoft Research internships. Her background includes award-winning surgical workflow analysis, a BTech from IIT Kharagpur, and applied RL work at Adobe that led to a US patent, reflecting a blend of strong theoretical grounding and practical systems impact. Shanka’s work uniquely bridges cognitive models of prefrontal cortex function with scalable ML agents, and she has practical experience evaluating finetuning and transfer to smaller LLMs for robust reasoning. Based in New Jersey, she brings nine years of research and engineering experience focused on making AI systems generalize more like humans.
9 years of coding experience
Bachelor of Technology, Electrical, Electronics and Communications Engineering, Bachelor of Technology, Electrical, Electronics and Communications Engineering at Indian Institute of Technology, Kharagpur
Doctor of Philosophy - PhD, Electrical and Computer Engineering, Doctor of Philosophy - PhD, Electrical and Computer Engineering at Princeton University
Hindi, Bengali, French, English