A Neurocomputational Model of Automatic Sequence Production

Abstract

Most behaviors unfold in time and include a sequence of submovements or cognitive activities. In addition, most behaviors are automatic and repeated daily throughout life. Yet, relatively little is known about the neurobiology of automatic sequence production. Past research suggests a gradual transfer from the associative striatum to the sensorimotor striatum, but a number of more recent studies challenge this role of the BG in automatic sequence production. In this article, we propose a new neurocomputational model of automatic sequence production in which the main role of the BG is to train cortical-cortical connections within the premotor areas that are responsible for automatic sequence production. The new model is used to simulate four different data sets from human and nonhuman animals, including (1) behavioral data (e.g., RTs), (2) electrophysiology data (e.g., single-neuron recordings), (3) macrostructure data (e.g., TMS), and (4) neurological circuit data (e.g., inactivation studies). We conclude with a comparison of the new model with existing models of automatic sequence production and discuss a possible new role for the BG in automaticity and its implication for Parkinson's disease.

Serial order in perception, memory, and action.

“…Theories of serial order take one of three approaches to representation of order: item coding, position coding, and noisy coding. Item coding theories represent order in terms of associations between items (Abrahamse, Jiménez, Verwey, & Clegg, 2010; Ebenholtz, 1963; Ebbinghaus, 1885; Grainger & van Heuven, 2004; Helie, Roeder, Vucovich, Rünger, & Ashby, 2015; Hull, 1943; Lewandowsky & Murdock, 1989; Murdock, 1995; Solway et al, 2012; Whitney, 2001). Chaining theories are the simplest versions of item coding: each item is associated with its successor.…”

Habits without values.

“…These models propose that novel behaviors are first acquired via a dopamine-dependent plasticity mechanism within the basal ganglia, and that with consistent performance, control of behavior is transferred to cortex via a Hebbian cortico–cortical plasticity mechanism. This idea has been applied to categorization learning and instrumental conditioning (Ashby et al, 2007), sequence learning (Hélie, Roeder, Vucovich, Rünger, & Ashby, 2015) and to action selection in probabilistic environments (Topalidou, Kase, Boraud, & Rougier, 2017), and it has been suggested to describe how basal-ganglia-dependent behavior becomes automatic in general (Hélie, Ell, & Ashby, 2015). Though the cortical module of these models is conceptually similar to our habitual controller, the basal ganglia module is different from our goal-directed controller in important ways: its learning rule instantiates a version of model-free RL, which tends to repeat actions in situations where they have led to reinforcement in the past, but does not learn about the particular outcomes that are expected to follow each action.…”

Automatic control: How experts act without thinking.

“…Theories of sequence learning generally endorse chaining (Abrahamse, Jiménez, Verwey, & Clegg, 2010; Helie, Roeder, Vucovich, Rünger, & Ashby, 2015; Keele et al, 2003). Participants typically learn one repeating sequence, and associations between items may be useful when a single sequence repeats.…”