last update: Nov, 2026

Jinwoo Lee

Jinwoo Lee

Doctoral Student
SciMinds (PI: Dr. Eshin Jolly)
UC San Diego, Department of Psychology


Research Interest
Theories, Methods, and Applications of affective/social neuroscience

Biography

I am the 2nd-year PhD student in UCSD SciMinds (PI: Dr. Eshin Jolly). I study the paradoxical nature of human emotion: its idiosyncrasy and commonality. Specifically, I investigate (1) which and how psychological profiles shape individual differences in emotion-related neural representations and subjective feelings, and (2) how people nevertheless come to share and align emotions. To this end, I am also developing new computational tools. Ultimately, I aim to explore implications of affective science across related fields such as mental health, aesthetics, and social robotics.

I grew up in South Korea, and received my B.A. in Psychology and Aesthetics and M.S. in Brain and Cognitive Sciences from SNU Connectome Lab (PI: Dr. Jiook Cha), Seoul National University.

News

#conference#talk

I’ll be presenting Differentiable Inter-sample Modeling (DIM) at Unifying Representations in Neural Models (UniReps), a NeurIPS workshop at Paris! While at SANS2026 we introduced DIM’s prototype architecture and its potential applications to neuroimaging data, this time we present the fully developed architecture and rigorous simulations evaluating its learning stability, parameter and model recovery, and identifiability. Preprint coming soon!

#talk

I gave a research talk at Dartmouth College’s Social Computation, Representation, and Prediction Lab (SCRAP Lab; PI: Dr. Mark Thornton). I introduced my first PhD project, Differentiable Inter-sample Modeling (DIM), and discussed how it enables new research questions beyond conventional Representational Similarity Analysis (RSA). In particular, I explored how neural representations of facial expressions in ambiguous social contexts vary in similarity depending on perceiver characteristics and the perceptual features of the facial expression images.

#talk

I gave a research talk at Sungkyunkwan University’s Human Affective Neuroscience Lab (PI: Dr. Justin Minue Kim). I presented my first PhD project, Differentiable Intersample Modeling (DIM), discussing how it complements the Anna Karenina (AnnaK) framework widely used in social and affective neuroscience. Specifically, using the lab's dataset (Kim & Kim, 2022 PNAS), we explored how the representational geometry of affective traits can align with that of emotional processing neural circuit morphologies.

#conference#talk#prize

I’m excited to share that my first PhD project, Differentiable Intersample Modeling (DIM), has been selected for a Trainee Blitz Talk at SANS2026, and I’ve also received a Travel Award! DIM is a new toolkit that expands RSA (representational similarity analysis), enabling a flexible learning of intersample geometry so that the resultant representation can be maximally aligned to the target representation.

I’d love to connect and chat at SANS2026! Feel free to reach out. See you in San Diego!

#talk

In the previous quarter, I worked as a TA for PSYC137: Social Cognition (taught by Dr. Piotr Winkielman), where I also had the chance to give a guest lecture to about 200 undergrad students. In the lecture, I introduced the ongoing debates in psychology around mixed emotions and shared some critical perspectives on why mixed emotions haven’t been widely studied as a distinct affective experience in experimental psychology so far.

Research Highlight

During my master's studies, I investigated how subjective feelings in daily life intertwine over time and shift across contexts, as well as what these complex dynamics reveal about an individual's traits and resultant emotional experience. To this end, I combined naturalistic neuroimaging with computational approaches (e.g., representational geometry and dynamics, predictive modeling). Stay tuned for my upcoming PhD work!