Universal Sleep Decoder: Aligning awake and sleep neural representation across subjects

Decoding memory content from brain activity during sleep has long been a goal in neuroscience. While spontaneous reactivation of memories during sleep in rodents is known to support memory consolidation and offline learning, capturing memory replay in humans is challenging due to the absence of well-annotated sleep datasets and the substantial differences in neural patterns between wakefulness and sleep. To address these challenges, we designed a novel cognitive neuroscience experiment and collected a comprehensive, well-annotated electroencephalography (EEG) dataset from 52 subjects during both wakefulness and sleep. Leveraging this benchmark dataset, we developed the Universal Sleep Decoder (USD) to align neural representations between wakefulness and sleep across subjects. Our model achieves up to 16.6% top-1 zero-shot accuracy on unseen subjects, comparable to decoding performances using individual sleep data. Furthermore, fine-tuning USD on test subjects enhances decoding accuracy to 25.9% top-1 accuracy, a substantial improvement over the baseline chance of 6.7%. Model comparison and ablation analyses reveal that our design choices, including the use of (i) an additional contrastive objective to integrate awake and sleep neural signals and (ii) the pretrain-finetune paradigm to incorporate different subjects, significantly contribute to these performances. Collectively, our findings and methodologies represent a significant advancement in the field of sleep decoding. Sleep plays a fundamental role in memory consolidation (Klinzing et al., 2019; Brodt et al., 2023). Past memories are known to reactivate during sleep, especially during the N2/3 stage of nonrapid eye-movement (NREM) sleep (Ngo & Staresina, 2022). In rodents, hippocampal cells have been found to replay their firing patterns during sleep, recapitulating awake experiences in a timecompressed order (Wilson & McNaughton, 1994; Skaggs & McNaughton, 1996). In humans, while direct cell recordings are rare, recording scalp electroencephalograms (EEG) during sleep is possible. Recent work on human sleep decoding has identified endogenous memory reactivation during the N2/3 stage of sleep, the extent of which was positively related to subsequent memory performance (Schreiner et al., 2021).