7/n What do you think? Is that plausible? Do you have a different model for how déjà vu works?

Read the full post at our substack! open.substack.com/pub/futurene...

6/n Together, these findings suggest that déjà vu reflects a genuine memory event that fails to elicit recall due to drift in hippocampal output pathways.

5/n Using the CausalMap algorithm, we traced each hippocampal engram to the cortical subcircuits that drove it. In one example, an odor representation in olfactory cortex reactivated a hippocampal place map from a remote location that shared olfactory activity patterns.

4/n The reactivated engrams matched events encoded at least ~1 month earlier (27-415 days), yet failed to reconstruct the original whole brain pattern typically observed in successful recall.

3/n Across 7,358 identified episodes, we found that déjà vu coincided with hippocampal activity that previously corresponded to a particular memory, but had become uncoupled due to representational drift.

2/n We analyzed long-term (≥5 years) human brain recordings from brain machine interfaces and identified déjà vu episodes by decoding spontaneous reports from the natural speech of participants.