EUVIMEDSkopri EUVIMED
Open access · CC BY-NC 4.0

Bayesian Surprise Predicts Human Event Segmentation in Story Listening

Manoj Kumar iD, Ariel Goldstein, Sebastian Michelmann, Jeffrey M Zacks, Uri Hasson, Kenneth A Norman

Cognitive Science · 2023-10 · 47 (10)

DOI: 10.1111/cogs.13343 ↗

At a glance

Journal article in Cognitive Science, published online on 2023-10-23. Written by Manoj Kumar and 5 co-authors. Institution: Princeton Neuroscience Institute, Princeton University. Major MEDLINE topics: Language. Cited 40 times so far (Crossref); 103 references. The full text is freely available (open access, CC BY-NC 4.0). Funded by: NICHD NIH HHS, NIA NIH HHS, Office of Naval Research Global.

AstrattSubjectsDettalji biblijografiċiCiteRelated studies

Abstract (original language)

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.

Subjects

MeSH terms (MEDLINE)

LanguageHumansBayes TheoremProbability

Author keywords

Bayesian surpriseEntropyEvent segmentationGPT-2NarrativesSurprise

Authors and institutions

  • Princeton Neuroscience Institute, Princeton University.
  • Department of Cognitive and Brain Sciences and Business School, Hebrew University.
  • Google Research, Tel-Aviv.
  • Department of Psychological & Brain Sciences, Washington University in St. Louis.
  • Department of Psychology, Princeton University.

ORCID profiles: Manoj Kumar

Funding

  • NICHD NIH HHS
  • NIA NIH HHS
  • Office of Naval Research Global
  • National Science Foundation

Cite

APA

Kumar, M., Goldstein, A., Michelmann, S., Zacks, J. M., Hasson, U., & Norman, K. A. (2023). Bayesian Surprise Predicts Human Event Segmentation in Story Listening. Cognitive Science, 47(10). https://doi.org/10.1111/cogs.13343

Vancouver

Kumar M, Goldstein A, Michelmann S, Zacks JM, Hasson U, Norman KA. Bayesian Surprise Predicts Human Event Segmentation in Story Listening. Cognitive Science. 2023;47(10). doi:10.1111/cogs.13343

BibTeX
@article{Kumar2023,
  author = {Kumar, Manoj and Goldstein, Ariel and Michelmann, Sebastian and Zacks, Jeffrey M and Hasson, Uri and Norman, Kenneth A},
  title = {Bayesian Surprise Predicts Human Event Segmentation in Story Listening},
  journal = {Cognitive Science},
  year = {2023},
  volume = {47},
  number = {10},
  doi = {10.1111/cogs.13343}
}
Start for free →