Juan Gallego
@juangallego.bsky.social
3.1K followers 850 following 480 posts
Thinking about the brain, spinal cord and how we move (and related neurotech). Into books, music, coffee, food, photography+art, animals & some humans. Group leader at Imperial College London #neuroskyence #Sensorimotor #compneurosky #Science
Posts Media Videos Starter Packs
Pinned
juangallego.bsky.social
🚨Big news!🚨
The lab is relocating to Lisbon, joining a great team of experimental and theoretical neuroscientists, and the Neurotechnology Warehouse, a new initiative to bridge basic and translational research.

I'll be sharing postdoc openings soon. Come join us in this new incarnation of the lab!
champalimaudr.bsky.social
🧠🎼 What does it take to restore movement? Neuroscientist and engineer, @juangallego.bsky.social, joins the new Centre for Restorative Neurotechnology at the Champalimaud Foundation.

🔗 Find out more in this interview: www.fchampalimaud.org/news/juan-al...
juangallego.bsky.social
Thanks David, really excited to be here with this amazing crowd
juangallego.bsky.social
Thanks Cristin, hope to catch up soon
juangallego.bsky.social
... in your very unbiased view of course!
thanks Matt
juangallego.bsky.social
finalmente!
and big congratulations on your exciting-but-not-yet-public-news
juangallego.bsky.social
Thanks Kevin! You know that you're welcome anytime
juangallego.bsky.social
Thanks a lot Britton! and see you soon-ish
juangallego.bsky.social
Thanks Yue, hope to see you in Kobe next year? Congrats also on your recent paper with @adrianhaith.bsky.social it's at the top of my reading pile
juangallego.bsky.social
Thanks Ben! It's very exciting news indeed
I'm looking forward to working with a lot of amazing new colleagues here at @champalimaudf.bsky.social
juangallego.bsky.social
Thank you Drew, very excited to be here
juangallego.bsky.social
thank you Laura, hope to catch up soon!
juangallego.bsky.social
the title sounds very intriguing, i should find the time to look at the paper, or listen to the podcast

congrats Ben, Shraddha, and the rest of the team
juangallego.bsky.social
🚨Big news!🚨
The lab is relocating to Lisbon, joining a great team of experimental and theoretical neuroscientists, and the Neurotechnology Warehouse, a new initiative to bridge basic and translational research.

I'll be sharing postdoc openings soon. Come join us in this new incarnation of the lab!
champalimaudr.bsky.social
🧠🎼 What does it take to restore movement? Neuroscientist and engineer, @juangallego.bsky.social, joins the new Centre for Restorative Neurotechnology at the Champalimaud Foundation.

🔗 Find out more in this interview: www.fchampalimaud.org/news/juan-al...
juangallego.bsky.social
Well, those studies are hard but they're coming

Re distributed code -- I didn't write the original post but partial compensation after focal lesions do support distributed computation...
juangallego.bsky.social
I agree, observations don't imply causation but people know that.

Also agree about having to write carefully and precisely
juangallego.bsky.social
It could, what I mean is that it likely won't exist in a real mammalian brain...
juangallego.bsky.social
You can look at papers doing single or few neuron manipulations & looked at changes in the overall neural population (we cite a few in the paper, spanning hippocampus, motor regions, V1, there's also Rebesco et al Front Neuro 2010): they all consistently show that the population changes w the target
juangallego.bsky.social
3) there are already several experiments casually probing assumptions about manifolds using brain computer interfaces (e.g. work from @aaronbatista.bsky.social + Byron Yu and Steve Chase, we're starting to do some stuff too...) and 1/2
juangallego.bsky.social
2) the toy model that you seem to outline of a single neuron within a circuit doing a "computation" (???) seems implausible in practice based on the neuroanatomy of the mammalian brain (recurrent connectivity, common inputs, distributed neuromodulator release...)
juangallego.bsky.social
Some thoughts, but I'm not sure about the exact challenge you lay out

1) I think it's well established that decoding or any correlational analysis, doesn't imply causation. This isn't a challenge for manifolds but for (neuro)science in general. Explanations should be based on many types of results