Disentanglement of persistent and dynamic neural substrates of working memory

In macaque match/non-match working memory tasks, cue stimuli are decoded at empirical or above-chance levels from persistent neurons only when correlations of activity patterns across epochs are constrained.

• Why it matters: Competing theories disagree on whether working memory relies mainly on persistent activity, dynamic activity, or both, leaving the relative contributions unresolved.
• What they did: The study analyzed macaque neuronal population recordings, extracted persistent and dynamic signatures using cue-specific temporal means, temporal principal components, and cross-epoch pattern correlations, then generated synthetic data constrained by these signatures to decode cues separately from persistent versus non-persistent neurons.
• The result: High decoding depended on a nuanced interplay: persistent neurons supported decoding across all tasks, but only after enforcing cross-epoch correlation constraints, providing a computational controlled framework to adjudicate competing models.

AI summary of the abstract · may contain errors · not medical advice · read the paper.