Come hang out with me tomorrow afternoon at poster JJ5! I'll be sharing new work investigating how replay features like rate of occurrence, speed and self-avoidance change dynamically with learning at behaviorally relevant locations
My 4 year old just showed me the "book" she made at school explaining, "here's your book, I hope you like it. Don't forget to subscribe!"
I was in tears after a rough science day until my 4 year old hugs me and says "at least you have a great kid!"
Interested in the brain, circuits, and behavior?
Enjoy tinkering and asking bold questions in neuroscience?
The Forli Lab (IIT, Genova, Italy) is hiring!
🧠 🔬 🟩◻️🟥->🍕🚫🍍
Check out our website for updates:
sites.google.com/view/forli-lab
This is one of the most outstanding examples of circuit understanding I've seen in a long time. The unification of theory and experiment is beautiful.
When Malcolm presented this in my lab, the audience was cheering at the end, and one person shouted (non-ironically) "You did it!"
Congratulations Malcolm!!! Can't wait to read :)
Our paper on foraging is now published in Neuron! Read it here:
www.cell.com/neuron/fullt...
This project was co-led by Michael Bukwich (not on Bluesky) and me, with major contributions from all co-authors. Huge thanks to the whole team!
This is amazing 😅
Congratulations to all the amazing scientists named 2025 Freeman Hrabowski Scholars by HHMI. We’re immensely honored to have two #PMB faculty members, @doctheagrif.bsky.social and @elisazhang.bsky.social , among this inspiring group. An incredible and well-deserved achievement! @ucdavis.bsky.social
Thank you!!
Thanks so much Kevin!
Thank you, Dan!
Ha, it most definitely is. When you do get a chance, let me know what you think!
Thank you Anna! :)
19/19 And of course, a huge thank you to my postdoc mentor, David Foster, for his guidance and invaluable input along the journey!
I welcome your thoughts and would love to chat with anyone about our findings!
18/19 John had conceived of similar adaptation models of replay and theta sequences before. When we teamed up to look at whether a neuronal fatigue model could explain replay's perplexing ‘past-avoidance’, everything began to click into place.
17/19 This project unfolded like a fun but super challenging puzzle. For a long time we wondered, why would replay avoid the recent past? I want to give a HUGE shout out to co-author John Widloski for his tremendous contribution in helping put the final pieces together.
16/19 When we added a source of facilitation onto cells that had been activated along the animal’s prior path, we were able to subtly increase the rate of retrospective replays overall and recapitulate the eventual emergence of a past bias.
15/19 The model produced replays that both avoided each other and the animal’s recent past (for a period of time related to time constant adaptation). Notably, without an additional input (presumably from MEC), the model did not produce the eventual bias for past-going replays.
14/19 Finally, we considered whether a recurrent network endowed with neuronal fatigue could recapitulate this counterintuitive organization of replay. We found that, in our model, adaptation not only *generates* replay, but it also organizes it.
13/19 We optogenetically (or chemogenetically) shut down MEC as rats consumed reward. We indeed found that retrospective replays became less frequent with MEC suppressed, although they were not altogether abolished.
12/19 How does the prior path eventually become preferentially replayed later in the stopping period? We wondered whether the medial entorhinal cortex, a cortical region highly interconnected with hippocampus, might play a role in biasing replay for the animal’s recent past.
11/19 This made us wonder, would *replay* of a path also fatigue the participating cells and prevent a subsequent replay of the same path? Consistent with the idea, we found that when two replays occur back-to-back (within about 1 sec), the second replay avoids the first.
10/19 We hypothesized that neuronal fatigue effectively takes recently activated place cells ‘out of commission’ for a few seconds. If replay is going to occur after the animal stops, it has to recruit different cells/represent a different path from that just taken.
9/19 So, which did we observe? Clearly, the latter- when the past and future paths overlapped, replay avoided them both for the first few seconds.
8/19 But sometimes rats turned around, retracing their steps so that the future and past paths overlapped. In this particular case, future-preferring versus past-avoiding replays would point in opposite directions.
7/19 We suspected that the primary factor influencing replay’s direction in the first few seconds after stopping is ‘past-avoidance’, rather than ‘future-preference'. This can be hard to distinguish, since rats often continue moving in the same direction after pausing.
6/19 But now you can more easily see that what changes most dramatically over the stopping period is the angle of replays relative to the animal’s past path. It starts off large and decreases with a time course that is strikingly similar across subjects…
5/19 Here’s another way of visualizing the open field data, considering ALL replays. Every replay contributes 2 points to the bottom left plot: an angle relative to the rat’s future path (green) and an angle relative to the rat’s past path (purple). This tells a similar story.
