Working Memory Simulator: Capacity, Decay & Rehearsal Dynamics

simulator intermediate ~10 min
Loading simulation...
C ≈ 4 items — consistent with Cowan's 4-item focus of attention

With default parameters, effective working memory capacity is approximately 4 items — matching Cowan's (2001) estimate of the focus of attention, within Miller's classic 7±2 range.

Formula

C_eff = k × R × τ × (1 - A/100)
P_retention(t) = exp(-t / τ)
d = 1 / τ (decay rate in items/s)

The Bottleneck of Thought

Working memory is the cognitive workspace where we hold information temporarily while reasoning, comprehending language, or solving problems. Its limited capacity — famously described by George Miller as 'the magical number seven, plus or minus two' — represents one of the most fundamental constraints on human cognition. This simulation models how chunk size, rehearsal speed, and trace decay interact to determine how much you can hold in mind at once.

Chunking and Capacity

Chunking groups individual items into meaningful units, effectively expanding capacity. A chess grandmaster sees board positions as familiar patterns (chunks) rather than individual pieces, enabling recall of entire game states. The parameter k controls chunk size: larger chunks mean more information per slot, but each chunk still occupies one capacity unit.

The Rehearsal Loop

Baddeley's phonological loop model explains how we maintain verbal information through subvocal repetition. The rehearsal rate R determines how quickly items can be refreshed before their traces decay with time constant τ. When rehearsal is blocked (articulatory suppression), capacity drops dramatically — evidence that active maintenance is essential, not merely passive storage.

Attention as a Bottleneck

Working memory and attention are deeply intertwined. The attention load parameter A simulates dual-task conditions where cognitive resources must be shared. High load leaves fewer resources for memory maintenance, causing items to drop out of the active buffer. This interaction explains why distracted driving is dangerous and why multitasking degrades performance on all concurrent tasks.

FAQ

What is working memory capacity?

Working memory capacity is the limited amount of information that can be actively maintained and manipulated in conscious awareness. George Miller (1956) estimated it at 7±2 chunks; more recent work by Nelson Cowan suggests a core capacity of about 4 items when chunking is controlled.

How does rehearsal maintain working memory?

The phonological loop model (Baddeley, 1986) proposes that subvocal rehearsal refreshes decaying memory traces. Items must be rehearsed before their trace fades — faster rehearsal rate or slower decay extends effective capacity.

What causes working memory decay?

Memory traces in the phonological store decay over approximately 2 seconds without rehearsal. Interference from similar items and attentional distraction accelerate this decay, as demonstrated by articulatory suppression experiments.

How does attention load affect working memory?

Divided attention reduces available resources for maintenance and rehearsal. Dual-task studies show that concurrent cognitive load can reduce working memory capacity by 30-50%, supporting resource-sharing models of attention.

Sources

Embed

<iframe src="https://homo-deus.com/lab/cognitive-neuroscience/working-memory/embed" width="100%" height="400" frameborder="0"></iframe>
View source on GitHub