EUVIMEDEUVIMED entdecken
Crossref

Competition Between Memory Updating and Differentiation Emerges From Intrinsic Network Dynamics

Julia Pronoza, Nina Liedtke, Marius Boeltzig, Ricarda I Schubotz, Sen Cheng

Neural Computation · 02.06.2026 · 38 (7) · 1233-1260

DOI: 10.1162/neco.a.1535 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neural Computation, online erschienen am 02.06.2026. Verfasst von Julia Pronoza und 4 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Neural Computation, Faculty for Computer Science, Ruhr University Bochum, 44801 Bochum, Germany julia.pronoza@ruhr-uni-bochum.de. Hauptthemen laut MEDLINE: Recognition, Psychology, Memory, Mental Recall, Models, Neurological. Die Arbeit nennt 81 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

When an event occurs that is similar to a previous experience, the original episodic memory can be modified with new information (updating) or a new memory can be encoded separately (differentiation). Prediction errors, the deviation between expected and actual stimuli, are believed to mediate the competition between updating and differentiation, but the underlying mechanisms remain unclear. Here, we present a new analysis of experiments examining recognition memory and cued recall of similar conversations. We find that the original version is recognized more confidently than the modified version, the recognition confidence for modified versions shows a U-shaped dependence on the prediction error, and larger prediction errors lead to participants’ more frequently recalling two versions. These results can be accounted for by a computational model based on a modified Hopfield network, which encodes the original and modified versions sequentially and weights the encoding of new patterns based on the prediction error. The model shows that (1) similar new memories interfere with previous ones (updating), while dissimilar ones are stored separately (differentiation); (2) interference from similar representations leads to reduced memory accuracy and lower-confidence recognition; and (3) modulation of the encoding strength is critical for accounting for the experimental data. Our modeling results show that prediction-error-driven competition between updating and differentiation can emerge from intrinsic network dynamics alone.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Recognition, PsychologyMemoryMental RecallModels, NeurologicalNeural Networks, ComputerHumansCuesComputer SimulationHopfield Neural Networks

Autor:innen und Institutionen

  • Institute for Neural Computation, Faculty for Computer Science, Ruhr University Bochum, 44801 Bochum, Germany julia.pronoza@ruhr-uni-bochum.de.
  • Department of Psychology, University of Münster, 48149 Münster, Germany nina.liedtke@uni-muenster.de.
  • Department of Psychology, University of Münster, 48149 Münster, Germany marius.boeltzig@uni-muenster.de.
  • Department of Psychology, University of Münster, 48149 Münster, Germany rschubotz@uni-muenster.de.
  • Institute for Neural Computation, Faculty for Computer Science, Ruhr University Bochum, 44801 Bochum, Germany sen.cheng@rub.de.

Zitieren

APA

Pronoza, J., Liedtke, N., Boeltzig, M., Schubotz, R. I., & Cheng, S. (2026). Competition Between Memory Updating and Differentiation Emerges From Intrinsic Network Dynamics. Neural Computation, 38(7), 1233-1260. https://doi.org/10.1162/neco.a.1535

Vancouver

Pronoza J, Liedtke N, Boeltzig M, Schubotz RI, Cheng S. Competition Between Memory Updating and Differentiation Emerges From Intrinsic Network Dynamics. Neural Computation. 2026;38(7):1233-1260. doi:10.1162/neco.a.1535

BibTeX
@article{Pronoza2026,
  author = {Pronoza, Julia and Liedtke, Nina and Boeltzig, Marius and Schubotz, Ricarda I and Cheng, Sen},
  title = {Competition Between Memory Updating and Differentiation Emerges From Intrinsic Network Dynamics},
  journal = {Neural Computation},
  year = {2026},
  volume = {38},
  number = {7},
  pages = {1233--1260},
  doi = {10.1162/neco.a.1535}
}
Kostenlos starten →