What will happen where and when could be predicted by the sequential reactivation of place cells that occurs while an animal pauses, suggesting that the replay is linked to mental time travel.

Neurons in the hippocampus modulate the rates at which they fire, and the locations in which they fire, so as to encode the information that is central to forming memories about personal experiences.

A neural network showed better prediction of upcoming states when it was selective in when it encoded and retrieved episodic memories, thereby explaining why humans show this selectivity in studies of naturalistic memory.

People use uncertainty for effective arbitration between episodic memory and incremental learning.

A provoked awakening protocol during sleep reveals that selective bilateral hippocampal damage in humans is associated with reduced frequency, quality, and content of dreaming.

A cortical model inspired by cognitive theories and deep learning demonstrates how the combination of wakefulness, non-rapid eye movement (NREM), and REM dreams leads to the emergence of robust and semantically organized neuronal representations in higher cortical areas.

Combining fMRI with continuous model-based measures of retrieval enables the behavioral and neural dissociation of multiple components of episodic memory.

Four types of declarative memory rely on different weightings of similar component processes, according to fMRI and subjective ratings.

Human hippocampal cornu ammonis 3 damage impairs both recent and remote autobiographical episodic memory, and disrupts functional integration in medial temporal lobe subsystem regions of the default network.

Behavioral and neural analyses unravel the angular gyrus’ key role in the prioritization of integrated narratives in memory and the orchestration of theta-frequency reconfigurations underpinning the process of memory integration.