Summary: New UC Berkeley research reveals how the brain replays and organizes memories, compressing long experiences into short, fixed-length neural sequences. By recording hundreds of hippocampal neurons in freely flying bats, scientists observed neural replay and theta-like sequences outside the rodent model, offering fresh insights into memory formation, navigation, and planning.
The team found that replay events often occur minutes after an experience, are compressed to a consistent duration regardless of flight length, and align with wingbeat rhythms rather than continuous hippocampal oscillations. These results shed light on core mechanisms the brain may use to encode, store, and anticipate behavior across species.
Key points
- Neural replay compresses both short and long flight trajectories into equal-length events.
- Theta-like sequential activity in bats is tied to wingbeat cycles, not continuous hippocampal theta oscillations.
- Replay events most often appear minutes after the original experience and frequently occur at places different from where the experience took place.
Source: UC Berkeley
Every day, the brain converts countless transient experiences—brief walks, conversations, or tasks—into lasting memories. How it does this is still being uncovered, but one crucial mechanism appears to be neural replay, where networks of neurons briefly recreate the activity patterns that occurred during an original experience.

Neural replay can occur both before and after an event, implying roles in consolidating memories and in planning future actions. To study these dynamics in a naturally behaving mammal, researchers at UC Berkeley recorded activity from hundreds of neurons at once in freely flying Egyptian fruit bats—a first for ensemble recordings in a flying animal.
Using high-density silicon electrode arrays and wireless recording technology, the team simultaneously monitored large neural populations and local field potentials while bats foraged in a spacious flight room. This allowed the scientists to study place-cell sequences, replay events during rest, and fast representational sweeps during flight under ethologically relevant conditions.
Place cells are hippocampal neurons that fire when an animal occupies a particular location. By mapping which cells corresponded to specific locations and tracking many cells together, the researchers reconstructed flight trajectories and identified when those same sequences reappeared during rest—replay events.
Unlike many rodent studies that create artificial separations between active and inactive states, the bat experiments captured multiple natural cycles of activity and rest within the same session. This revealed several surprising patterns. Replays typically occurred minutes after the original behavior and often at locations far from where the flight occurred. Most strikingly, replay duration was invariant: sequences corresponding to short and long flights were compressed to the same fixed time window.
This consistent duration suggests the brain packages spatial and temporal information into fixed-size packets during replay. From a computational perspective, such fixed “chunks” of information can simplify downstream reading and integration, making memory transmission more efficient.
The researchers also examined theta sequences—rapid sequential activation of place cells believed to represent a brief look-ahead during movement. In rodents, these sequences are linked to continuous theta oscillations in the hippocampus. Bats and humans, however, often lack sustained hippocampal theta rhythms. The study found fast representational sweeps in bats during flight that resembled theta sequences but were not tied to hippocampal theta oscillations. Instead, these sweeps were phase-locked to the bat’s wingbeat cycle at around 8 Hz.
The alignment of fast neural sequences with an overt motor rhythm suggests that behaviorally relevant sensorimotor rhythms—such as wingbeats, whisking, or syllabic rates of speech—can structure hippocampal ensemble dynamics. Because many species exhibit rhythmic behaviors in the ~8 Hz range, the findings point to a potentially conserved mechanism linking movement, perception, and internal neural representations.
Together, the results challenge existing models built primarily on rodent recordings and underscore the value of comparative, naturalistic studies. By demonstrating that replay and fast sequential dynamics can have different timing, locations, and ties to motor rhythms across species, the work broadens our understanding of how hippocampal ensembles support navigation, memory consolidation, and planning.
Funding: Research support included grants from the Air Force Office of Scientific Research, the National Institute of Neurological Disorders and Stroke, the Office of Naval Research, the New York Stem Cell Foundation, the Vallee Foundation, and the Howard Hughes Medical Institute.
About this memory research news
Author: Kara Manke
Source: UC Berkeley
Contact: Kara Manke, UC Berkeley
Image credit: Neuroscience News
Original research (open access): “Replay and representation dynamics in the hippocampus of freely flying bats,” by Michael Yartsev et al., published in Nature. The study reports time-compressed forward and reverse replay events associated with sharp-wave ripples, and fast representational sweeps during flight that are phase-locked to wingbeat cycles rather than continuous theta oscillations.
Abstract (summary)
Navigation and memory rely on emergent properties of hippocampal ensembles, including neural replay and sequential representational sweeps. This study wirelessly recorded large neural populations and local field potentials in freely flying bats during spontaneous foraging. During rest, the hippocampus exhibited time-compressed forward and reverse replay of multiple flight trajectories that coincided with sharp-wave ripples; replays often occurred at spatially and temporally distant locations and showed speed scaling with trajectory length. During flight, ensembles produced fast representational sweeps that moved ahead of the bat’s position; unlike rodent reports, these sweeps occurred without continuous theta oscillations and were phase-locked to the bats’ wingbeat rhythm. These findings highlight that behaviorally relevant sensorimotor rhythms can shape hippocampal dynamics and emphasize the importance of comparative, ethologically realistic approaches for understanding memory and navigation.