Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Neural correlates of prediction during naturalistic event perception

Yuye Huang¹, Anna Hu, Buddhika Bellana, Janice Chen; ¹Johns Hopkins University

Presenter: Yuye Huang

Prediction is a critical ability for human beings and other intelligent agents. Classic reinforcement learning studies have extensively examined neural signals relating to prediction error in constrained laboratory settings, but much less is known about how people make predictions in the course of real-life events. In real-life events, predictions are complex: they combine information from multiple sources, have multiple possible outcomes, and range over a longer timescale. In this study, we leveraged narratives to study real-world prediction processes. Participants in our study made natural-language predictions during movie watching. We quantified these predictions in three different ways (accuracy, variety, and confidence), and used these variables to model functional MRI data collected as other participants watched the same movies. We found that fluctuations in prediction accuracy were tracked by activity in the default mode network, a system responsible for retrieving and integrating past information to simulate potential futures.

Topic Area: Predictive Processing & Cognitive Control

Extended Abstract: Full Text PDF