Thalamocortical synchrony impairment is an early neurophysiological signature of Alzheimer's disease, driven partly by NMDA receptor hypofunction on glutamatergic thalamic relay neurons. GRIN2B-selective PAMs could restore synchrony without the global excitotoxicity risk of non-selective NMDA agonists. The central challenge is demonstrating a therapeutic window: sufficient enhancement of NMDA function to restore oscillations, without pathological overactivation in already-stressed neurons. Falsifiable prediction: GRIN2B PAM GNE-8324 at 10 mg/kg should restore 40 Hz gamma power by ≥40% vs vehicle in P301S tau mice (LFP recording from primary auditory cortex during 40 Hz chirp). Morris water maze escape latency should improve by ≥25%. Safety endpoint: no increase in caspase-3 cleavage in hippocampal neurons at 4× therapeutic dose.