Dynamical Mechanisms for Coordinating Long-term Working Memory Based on the Precision of Spike-timing in Cortical Neurons

Dynamical Mechanisms for Coordinating Long-term Working Memory
Based on the Precision of Spike-timing in Cortical Neurons

Terrence J. Sejnowski1,2

1 Computational Neurobiology Laboratory, Salk Institute for Biological Sciences, La Jolla, 92037 CA, USA 2 Department of Neurobiology, University of California, San Diego, La Jolla, 92093 CA, USA

Abstract

In the last century, most sensorimotor studies of cortical neurons relied on average firing rates.
Rate coding is efficient for fast sensorimotor processing that occurs within a few seconds. Much less is known about long-term working memory with a time scale of hours (Ericsson and Kintsch, 1995). The discovery of the millisecond precision of spike initiation in cortical neurons was unexpected (Mainen and Sejnowski, 1995). Even more striking was the precision of spiking in vivo, in response to rapidly fluctuating sensory inputs, suggesting that neural circuits could, in principle, preserve and manipulate sensory information through spike timing. High temporal resolution enables a broader range of neural codes. It could also support spike-timing-dependent plasticity (STDP), which is triggered by the relative timing of spikes between presynaptic and postsynaptic neurons in the millisecond range. What spike-timing mechanisms could regulate STDP in vivo? Cortical traveling waves have been observed across many frequency bands with high temporal precision. Traveling waves have wave fronts that could link spike timing to STDP. As a wave front passes through a cortical column, excitatory synapses on the dendrites of both pyramidal and basket cells are synchronously stimulated. Inhibitory basket cells form a calyx on pyramidal cell bodies, and inhibitory rebound following a strong transient hyperpolarization can trigger a backpropagating action potential, which arrives shortly after the excitatory inputs on pyramidal dendrites. STDP activated in this way could persist for hours, creating a second-tier network. This temporary network could support long-term working memory, a cognitive network riding above the long-term sensorimotor network. On their own, traveling waves and STDP have not yet yielded new insights into cortical function. Together, they could be responsible for how we think (Sejnowski, 2025).

2

Introduction
We are at a crossroads in systems neuroscience. In the 20th century, progress was made in characterizing the response properties of single neurons using sharp microelectrodes during behavioral tasks. The cortical responses of sensory neurons correlated with sensory inputs, those of motor neurons with movements, and those of neurons in association areas with higher cognitive functions. The tasks typically lasted a few seconds and required extensive training.
However, in the wild, behavior is self-generated and is coordinated over much longer time spans.
Recordings from freely moving rodents, for example, revealed that hippocampal neurons responded selectively to places in the environment, which would have been missed in head-fixed experiments. Self-generated cognition that occurs without any sensory inputs or motor outputs over minutes and hours, as in remembering and planning, is much more difficult to study than tasks that require working memory over seconds.
During the 1980s, I focused on network models of vision (Ballard, Hinton, and Sejnowski, 1983), grounded in rich psychophysical research and neural recordings from the visual cortex. I was inspired by pioneering network models, such as the Marr and Poggio (1976) model of stereopsis. Vision research in that era focused on images and object recognition. I knew that the visual system integrated information across eye movements and was curious how it was done. For example, as you read this article, your eyes make fast, saccadic movements across the page, taking in small groups of words in your fovea three times per second. Each saccade is a snapshot that must be integrated with previous words to build a conceptual understanding of what is being conveyed.
Psychologists call this long-term working memory (Ericsson and Kintsch, 1995). After reading this article, your brain will think about it in the context of experiences and thoughts previously stored in long-term memory. After listening and watching a lecture for an hour, you can retain enough details that you heard and the slides you saw to ask a relevant question. During a concert, recurring themes are expected and variations detected. Long-term working memory has a time scale of hours, much longer than short-term working memory, which lasts seconds and minutes, such as remembering a phone number by rehearsal.

3

We can now record distributed neural activity from many thousands of neurons simultaneously throughout the cortex. We need a new conceptual framework for how cognition arises from global activity on long time scale