Interdisciplinary StudiesarXiv
Analyzing Memory Effects in Large Language Models through the Lens of Cognitive Psychology
Z. Cao, L. Schooler, et al.
Memory is adaptive but fallible — and this study finds that cutting-edge language models echo many human memory quirks. Using classic human-memory paradigms, the authors show LLMs display list-length and list-strength effects, associative interference, DRM-style false recognitions, and cross-domain generalization, while differing in order sensitivity and resilience to nonsense. This research was conducted by Zhaoyang Cao, Lael Schooler, and Reza Zafarani.
Related Publications
Explore these studies to deepen your understanding
Adjacent work that informs or extends this paper's methodology and findings.
The Arts
Seeing disasters through the eyes of children: a critical reading of One Snowy Night and The Storm in the Barn through the lens of critical disaster studies
H. M. Bayoumy
Medicine and Health
The implication of the illness metaphors of *In America*: Mitigating negative effects of metaphors through illness narratives in the postpandemic era of COVID-19
R. Dai and S. Peng
Psychology
Understanding the Role of Large Language Models in Personalizing and Scaffolding Strategies to Combat Academic Procrastination
A. Bhattacharjee, Y. Zeng, et al.
Computer Science
Sentiment Analysis in the Era of Large Language Models: A Reality Check
W. Zhang, Y. Deng, et al.
