Event Segmentation In Episodic Memory: The Role Of Prediction Errors, Contextual Transitions, And Affective Modulation

Episodic memory allows us to structure continuous experience into discrete, meaningfulevents, yet the mechanisms that shape segmentation and how different factorsinfluence this process are not fully understood. This thesis investigates how predictionerrors, contextual stability, and affective processes influence event segmentation.Across a series of behavioral experiments, I show that stable contextual informationplays a more critical role in triggering event boundaries than prediction errorsalone. Furthermore, while high-reward associations during the ongoing experiencecan enhance memory, they do not consistently alter the segmented structure of theexperience. Finally, I demonstrate that retrospective cognitive reappraisal of emotionalevents can reorganize how past experiences are segmented and remembered.Together, these findings challenge traditional models that emphasize prediction errorsand highlight the flexible, dynamic nature of memory organization shaped byboth bottom-up and top-down influences.