Kushin Mukherjee
kushinm at stanford dot edu

I am a postdoctoral scholar in the Cognitive Tools Lab at Stanford University. I'm broadly interested in the human ability to use and understand visualizations (charts, graphs, drawings) in service of communication and discovery. My research focuses on developing computational cognitive models of visualization understanding to both (1) better characterize human cognition and (2) bridge the gap between modern AI systems and human-like understanding of visual concepts.
I completed my PhD in Psychology at UW-Madison while based in the Knowledge and Concepts Lab and the Schloss Visual Reasoning Lab. I also interned at Apple AIML working with the Vis team on chart understanding in vision-language models. I was also fortunate to be a Kohler Fellow at the Wisconsin Institute for Discovery, which has let me think about how science and art can intersect to give rise to fun new ideas. Much of my work has been the product of engaging and thoughtful collaborations with members of the Social Interaction Lab (SoIL) at Stanford, the Visual Intelligence and Technological Advances Lab at York, and the Neuroscience of Cognitive Control Lab at Princeton. Prior to grad school, I received my BA in Cognitive Science and Japanese from Vassar College, where I was advised by Ken Livingston and Josh de Leeuw. I also spent a summer at the Computation and Cognition Lab at Stanford as a CSLI intern, working with Judy Fan and Robert Hawkins (and continue to today!).
news
Jun 5, 2025 | Our paper AI-enhanced semantic feature norms for 786 concepts was accepted to both the BiAlign Workshop at ICLR, where it won the Best Paper Award 🥇, and to CogSci 2025 where it won the Computaional Modeling Prize in Applied Cognition 🏆! |
---|---|
Mar 21, 2025 | I defended my dissertation and am now a postdoctoral scholar at Stanford in the Cognitive Tools Lab! |
Mar 5, 2024 | Two papers accepted to the ICLR 2024 Workshop on Representational Alignment! Stay tuned for preprints! |
Feb 1, 2024 | I will be interning at Apple Machine Learning Research, Cambridge this summer! Excited to be back on the east coast for a bit~ |
Oct 6, 2023 | Our paper SEVA: Leveraging sketches to evaluate alignment between human and machine visual abstraction was accepted to the NeurIPS Datasets and Benchmarks track! See publications for more! |