Amazing day for my first tweet. Super excited to share my PhD research, published today @Nature. We find that contrary to popular belief, cortical reactivations predict the future rather than reflect the past. 1/10.

RT @KempnerInst: NEW: Mark Andermann (@andermann_mark) of @harvardmed and @BIDMChealth brings experimental data and computational modeling….

Also thank you to @RichieHakim and @andy_ahn_ye.

Our story is also summarized here:.

Also thank you to my PhD funding, NDSEG @NDSEGOfficial and HHMI Gilliam @HHMINEWS.

Thank you to all the were involved, especially my mentor Mark Andermann @andermann_mark. @andrewlutas @Oren_Amsalem @richlyn_jesseba @JustinMcMahon_ @JDimidschstein @blsabatini. Thank you also to the Andermann Lab and others for their feedback! . 10/10.

In summary, our study indicates that offline reactivations during quiet-waking may actively reorganize sensory-evoked response patterns to enhance the separability of population responses during distinct experiences and support future behavioral performance. 9/10

Amazingly, we can accurately model future stimulus-evoked responses using only the rate and content of reactivations and the actual stimulus response on trial 1. This model also captured the observed pattern separation of the two stimulus responses across the day. 8/10.

Neurons that will increase/decrease their stimulus activity across the day already have relatively increased/decreased participation in early stimulus reactivations. Just by looking at early reactivations, we can predict changes in neuronal responses to future stimuli. 7/10.

Even on the very first trial of the day, and the very first time the mouse sees the stimulus in its life, subsequent stimulus reactivations in the following minute predicted future stimulus responses hours later that day and in the following days. 6/10

How do reactivations relate to drifting cortical responses? We had guessed that reactivations reflect, and are a copy of, the previous experience. Therefore, as stimulus representations drift, stimulus reactivations should also follow this drift. But we were wrong. 5/10.

Representational drift is a prevalent property of cortex. We find that responses to stimuli drifted both within and across days, with no behavioral changes. Interestingly, this drift was not random as the representations of the two stimulus responses became more different. 4/10

Reactivations in visual cortex (similar to hippocampus), occurred during quiet-waking, decayed with trial novelty, and were biased in rate towards the most recently presented stimulus. Although coupled to hippocampal SWRs, cortical reactivations were locally generated. 3/10

We showed visual stimuli to mice, separated by minute-long ITIs where stimulus reactivations occurred. To overcome statistical barriers when comparing the content of stimulus-evoked responses and reactivations, we simultaneously imaged from ~7k neurons in visual cortex. 2/10

RT @yaochen20: Last May, we had excellent speakers from around the world: @mariaahmedch @ambrosipri @davidhbrann @salwan_butrus @alexbchen….

RT @yaghoubi_mh: Main @CosyneMeeting is over and 3 (among many) posters I learned from:.1- Cortical reactivations predict future sensory re….

RT @andermann_mark: Excited to share Nghia Nguyen’s graduate work @biorxiv_neursci! . How do offline stimulus reactivations in cortex relat….

RT @MytienTNguyen: In an analysis of @NIH F31 fellowships from 2001-2020, we found that that there has been a disproportionate growth in ge….

RT @CellCellPress: Nghia Nguyen @NghiaDNguyen #FacesOfCell worked on much of the research in this @CellCellPress paper when he was still an….

RT @oilychickenleg: Great day for my first tweet. So happy that >90% of the work I’ve done in the last 6 years is published today @CellCell….