Bayesian Surprise: Predicting Event Segmentation in Story Listening

Published on October 24, 2023

Imagine listening to a captivating story. As you follow along, your brain naturally divides the story into different events or scenes. So, how does your brain determine when one event ends and another begins? Scientists conducted an experiment using a deep learning language model called GPT-2 to investigate this phenomenon. They found that event boundaries in stories coincide with sudden increases in Bayesian surprise, a measure of prediction error. It’s like when you’re watching a movie and an unexpected plot twist occurs, catching you off guard. In contrast, a simpler measure of prediction error called ‘surprisal’ didn’t show the same relationship with event boundaries. This suggests that our brains use Bayesian surprise as a control mechanism for segmenting events in storytelling. The study also highlights the importance of using accurate measures of prediction error when studying cognitive processes. Want to dive deeper into the research? Check out the full article!

Abstract
Event segmentation theory posits that people segment continuous experience into discrete events and that event boundaries occur when there are large transient increases in prediction error. Here, we set out to test this theory in the context of story listening, by using a deep learning language model (GPT-2) to compute the predicted probability distribution of the next word, at each point in the story. For three stories, we used the probability distributions generated by GPT-2 to compute the time series of prediction error. We also asked participants to listen to these stories while marking event boundaries. We used regression models to relate the GPT-2 measures to the human segmentation data. We found that event boundaries are associated with transient increases in Bayesian surprise but not with a simpler measure of prediction error (surprisal) that tracks, for each word in the story, how strongly that word was predicted at the previous time point. These results support the hypothesis that prediction error serves as a control mechanism governing event segmentation and point to important differences between operational definitions of prediction error.

Read Full Article (External Site)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>