PROJECT SUMMARY An astonishing, but underappreciated, aspect of AD and related dementias (ADRD) is the moment-to-moment fluctuation in cognitive health. Moments of higher cognitive and sensorimotor performance, including ‘episodes of lucidity’ or ‘lucid intervals’ occur, though are rare, unpredictable and yet undeniably precious. Sensory cues and contexts may play a critical role in triggering periods of improved cognition. For example, a nostalgic smell or a wedding song can elicit periods of lucidity even for patients in advanced stages of disease. One study concludes that “environmental context” is key to such episodes. Memories and other cognitive capacities may thus exist in the AD brain, but for the most part remain inaccessible. Can these “hidden memories” be unlocked? The underlying neural basis of cognitive fluctuations in AD in general, and lucid intervals in particular, remains completely unknown. Severe dementia is generally considered irreversible, yet even patients with advanced AD appear to be able to tap into intrinsic mechanisms that transiently improve cognition. To identify these intrinsic mechanisms, we propose to focus on the auditory cortex (AC). The AC serves as integrative hub for feedforward sensory information, top-down signals, and ascending neuromodulation. The pathological hallmarks of AD, including amyloid-beta (Aβ) plaques and neurofibrillary tangles, accumulate in the AC. Neurons in the cholinergic basal forebrain (CBF), which degenerate early in AD, strongly innervate the AC. A large body of literature, including our own previous work, demonstrates that optimal CBF neuromodulation of inhibitory networks in the AC is critical for audiomotor task performance. Our overarching hypothesis is that the accumulation of Aβ disrupts the integration of state-dependent neuromodulation into cortical networks. In a default ‘pathological’ state due to Aβ accumulation, sub-optimal CBF tone interferes with inhibitory gating of auditory cues and impairs behavior. We propose that transient improvements in performance (‘lucid-like states’) (Aim 1) arise when inhibitory networks are engaged (Aim 2) by endogenous CBF tone that is spontaneously tuned to an optimal level (Aim 3) thereby enabling task-relevant processing of auditory cues. In this supplement, we propose to perform foundational experiments achievable within the 1-year period of the award and that build on the behavioral and neural approaches from the active award (R01).