Keynotes | K&Ts | GACs | Talks | Posters | Search

Poster D in Poster Session D: Wednesday, August 5, 2:00 – 3:45 pm, Kimmel Center, Shorin & Rosenthal Rooms

Representational Geometry Reflects the Dynamics of Experience: Evidence from Relational Learning of Emotion Concepts

Yumeng Ma¹, Philip A. Kragel¹; ¹Emory University

Presenter: Yumeng Ma

Knowledge about emotional events is organized in a low-dimensional space defined by valence and arousal dimensions. The geometry of this space varies systematically across individuals, with differentiation of similarly valenced emotions emerging across development. Here, we test the hypothesis that distortions in representational geometry arise from variation in transition statistics of emotional experiences. Using the Tolman-Eichenbaum machine, a computational model of relational learning, we simulated agents that experienced different transition dynamics. Comparing model-derived representations of emotion concepts with human similarity judgments from an archival dataset (n = 92), we find that representations learned under reduced sampling of the arousal dimension better capture younger participants’ similarity judgments, whereas uniform sampling better captures judgments of older participants. These findings suggest that transition statistics may play an important role in shaping how the brain organizes emotion knowledge to enable flexible behavior in complex environments.

Topic Area: Memory, Learning & Knowledge Structures