Some of the most important discoveries in cognitive neuroscience have come from recent innovations in experimental tools. Computational models that simulate human perception of environmental inputs have revealed the internal processes and features by which those inputs are learned and represented by the brain. We advance this line of work across two separate research studies in which we leveraged these models to both generate experimental task stimuli and make predictions about behavioral and neural responses to those stimuli. Chapter 1 details how nine language models were used to generate controversial sentence pairs for which two of the models disagreed about which sentence is more likely to occur. Human judgments about these sentence pairs were collected and compared to model preferences in order to identify model-specific pitfalls and provide a behavioral performance benchmark for future research. We found that transformer models GPT-2, RoBERTa and ELECTRA were most aligned with human judgments. Chapter 2 utilizes the GloVe model of semantic word vectors to generate a set of schematically structured poems comprising ten different topics whose specific temporal order was learned by a group of participants. The GloVe model was then used to investigate learning-induced changes in the spatial geometry of the representations of the topics across the cortex. A Hidden Markov Model was also used to measure neural event segmentation during poem listening. In both analyses we identified a consistent topography of learning-induced changes in the default mode network, which could be partially explained by the models.

Fluctuations in attention are ubiquitous. We all experience the waxing and waning of our attention, with periods of focus alternating with periods of distraction by irrelevant thoughts or external sensations. Given the pervasiveness of these fluctuations, it is important to understand how they influence both our behavior in the moment and the structure of our memory. In this dissertation, I use behavioral studies and eye tracking to measure spontaneous attentional fluctuations and examine how these fluctuations shape online behavior and subsequent memory. I test my primary hypothesis that optimal attentional states help us link experiences over time to allow our memories to be temporally organized, whereas suboptimal states disrupt the temporal structure of memory. In Chapter 1, I present four studies using a novel experimental design to connect research on sustained attention and memory. I replicate prior findings linking response-time-based measures of attention to online behavior. Surprisingly, I found that these response-time measures of attention do not predict the temporal structure of free recall. In Chapter 2, I indexed attentional fluctuations with both response times and pre-trial pupil size and demonstrated that these measures of attention predict complementary aspects of behavior. Attentional fluctuations, as indexed by pupil size, predicted the temporal organization of memory but not attentional lapses in online behavior. Conversely, response times predicted attentional lapses in the moment but did not predict the temporal organization of memory. Finally, in ongoing work in Chapter 3, I examine whether providing cues at retrieval enhances or diminishes the effects of attentional fluctuations on the temporal organization of memory. Together, my results shed light on the complex interactions between fluctuations in attention and episodic memory. Critically, I show that different measures of attention – behavioral vs. physiological approaches – capture distinct aspects of cognitive function, and suggest that the attentional states that shape online behavior and later memory are at least partly distinct.