A Common Recognition Architecture Across Memory Scales

Recognition old/new EEG signatures persist from 1 to 128-item set sizes, while a shared neural response shifts progressively later and weaker as memory scale increases.

• Why it matters: It remains unclear whether recognition relies on distinct neural formats across episode scales or a common mechanism that adapts with set size.
• What they did: EEG was recorded during a recognition task with set sizes from 1 to 128 items (2 to 256 recognition judgments), and multivariate classifiers plus temporal-alignment and cross-correlation analyses tested generalization and waveform similarity.
• The result: Classifiers generalized robustly across set sizes, and timing shifts propagated into motor response preparation; behavioral relevance tracked working-memory capacity at smaller scales and long-term memory performance at larger scales.

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