Analysis of free recall dynamics of an abstract working memory model
This paper analyzes the free recall dynamics of a working memory model. Free recalling is the reactivation of a stored pattern in the memory in the absence of the pattern. Our free recall model is based on an abstract model of a modular neural network composed on $N$ modules, hypercolumns, each of which is a bundle of minicolumns. This paper considers a network of $N$ modules, each consisting of two minicolumns, over a complete graph topology. We analyze the free recall dynamics assuming a constant, and homogeneous coupling between the network modules. We obtain a sufficient condition for synchronization of network's minicolumns whose activities are positively correlated. Furthermore, for the synchronized network, the bifurcation analysis of one module is presented. This analysis gives a necessary condition for having a stable limit cycle as the attractor of each module. The latter implies recalling a stored pattern. Numerical results are provided to verify the theoretical analysis.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Steady state behavior of the free recall dynamics of working memory
This paper studies a dynamical system that models the free recall dynamics of working memory. This model is a modular neural network with n modules, named hypercolumns, and each module consists of m minicolumns. Under mi…
Preliminary Evidence -- Diagnosed Alzheimer's Disease But Not MCI Affects Working Memory Capacity - 0.7 of 2.7 Memory Slots is Lost
Recently it was shown explicitly that free recall consists of two stages: the first few recalls empty working memory (narrowly defined) and a second stage, a reactivation stage, concludes the recall (Tarnow, 2015). It wa…
ManagementWhy Knowledge Distillation Works in Generative Models: A Minimal Working Explanation
Knowledge distillation (KD) is a core component in the training and deployment of modern generative models, particularly large language models (LLMs). While its empirical benefits are well documented--enabling smaller st…
Knowledge DistillationLanguage ModelingLanguage ModellingA Transition System Abstraction Framework for Neural Network Dynamical System Models
This paper proposes a transition system abstraction framework for neural network dynamical system models to enhance the model interpretability, with applications to complex dynamical systems such as human behavior learni…
A Bayesian Analysis of Dynamics in Free Recall
We develop a probabilistic model of human memory performance in free recall experiments. In these experiments, a subject first studies a list of words and then tries to recall them. To model these data, we draw on both …
Bayesian InferenceRetrievalTopic Models