Motor skills and word pairs respond differently to the same stimulation
The clearest direct comparison comes from a study that delivered brief bursts of pink noise just after a sleep spindle—a short burst of brain activity during non-REM sleep—and then tested two different memories after a nap. Relative to a nap without stimulation, this post-spindle stimulation reduced the post-nap drop in motor sequence accuracy, but it did not significantly change word-pair recall [1]. In other words, the same stimulation protocol helped one kind of memory and left the other essentially unchanged.
That split is consistent with a broader thesis that tested closed-loop auditory stimulation across several memory types, including a motor sequence learning task, a pattern separation task, and three declarative tasks. Across those tests, stimulation did not produce measurable behavioral improvements on the declarative or pattern-separation tasks; in fact, one night of stimulation was associated with a decline in pattern separation performance [5]. The thesis concluded that stimulation can affect the brain during sleep in a way that interacts with memory tasks, but does not always produce a measurable performance change [5].
Even within motor memory, the benefit is not guaranteed
A separate study using closed-loop auditory stimulation timed to slow oscillations—the large, slow brain waves of deep sleep—found that stimulation strongly altered sleep physiology, evoking slow oscillations, spindles, and coupled slow-oscillation–spindle events. Yet participants showed similar post-nap improvement on a motor sequence task whether or not they received stimulation [2]. That result sits in tension with the motor benefit reported in the post-spindle study [1], and the difference may lie in timing: the post-spindle protocol deliberately avoided triggering new spindles, whereas the slow-oscillation-timed protocol evoked them [1][2].
The post-spindle study also found that the motor benefit was greatest in participants who had lower spindle density during a baseline adaptation nap [1]. That suggests the benefit may depend on an individual's starting sleep physiology, not just on the task label. So "motor tasks benefit" is too strong a summary; the evidence supports "motor tasks can benefit under some timing and participant conditions."
Declarative memory has repeatedly come up empty
Two other studies that used word-based declarative tasks also failed to find a retention benefit from closed-loop stimulation. One used a paired-associates word task and compared verum stimulation nights with sham nights; linear mixed models found no significant difference in retention scores or subjective sleep quality [3]. The other, the post-spindle experiment described above, likewise found no significant effect on word-pair recall [1].
These null results do not prove that declarative memory can never be helped—they show that the protocols tested so far have not produced a measurable gain. The paired-associates study did find that less noise in the slow-oscillation signal was correlated with greater detection and stimulation of slow oscillations, alongside a trending relationship to higher retention [3], which hints that technical precision may matter more than task type alone.
Why the task might change the outcome
One plausible explanation is that different memory tasks depend on different sleep rhythms. The slow-oscillation-timed study found that post-nap motor improvement correlated with the density of slow-oscillation-coupled spindles across conditions [2], suggesting that motor consolidation rides on a particular coupling pattern that stimulation can either support or disrupt. The post-spindle study, by contrast, elicited widespread slow-oscillation activity without triggering new spindles and still saw a motor benefit [1], which suggests the refractory period after a spindle is not functionally silent and can support motor memory even without fresh spindles.
A thesis that examined brain activity during recall also found that stimulation changed brain activity during memory task recall, even when behavioral performance did not change [5]. That dissociation—brain response without behavioral gain—is a reminder that a null behavioral result does not mean the stimulation did nothing. It may mean the task was not sensitive to the particular rhythm the stimulation altered, or that the effect was too small to detect in the samples studied.
Sources used in this answer
Closed-loop auditory stimulation during sleep shapes cortical responses and memory consolidation depending on spindle timing
Post-spindle closed-loop auditory stimulation reduced post-nap deterioration in motor sequence accuracy but did not significantly affect word-pair recall, with the motor benefit greatest in participants with lower baseline spindle density.
The effects of closed-loop auditory stimulation on sleep oscillatory dynamics in relation to motor procedural memory consolidation
Closed-loop auditory stimulation timed to slow-oscillation upstates strongly altered sleep physiology but did not enhance motor procedural memory consolidation relative to no stimulation.
THE RELATIONSHIP BETWEEN CLOSED-LOOP AUDITORY STIMULATION DURING SLEEP AND EEG ON MEMORY CONSOLIDATION
Verum closed-loop auditory stimulation produced no significant difference in paired-associates word-task retention or sleep quality compared with sham nights, though less noisy slow-oscillation signals correlated with greater stimulation detection and a trending retention relationship.
The application of closed-loop targeted memory reactivation during slow wave sleep
Closed-loop targeted memory reactivation during slow-wave sleep was examined using a transitive inference task over three weeks, but the abstract does not report a clear memory outcome.
The effects of closed loop auditory stimulation on brain and behaviour in the short and long-term
Across motor sequence, pattern separation, and declarative tasks, closed-loop auditory stimulation changed brain activity during recall but did not produce measurable behavioral improvement on most tasks, and one night of stimulation was associated with a decline in pattern separation performance.