@ahalawa.bsky.social
Neuroscience PhD student at McGill Co-supervised by Adrien Peyrache & Blake Richards
For those interested in open neuroscience learning tools, check out the preprint for “RetINaBox: A hands-on tool for experimental neuroscience" that a couple students in my lab worked on in collaboration with the Trenholm lab:
www.biorxiv.org/content/10.1...
🧠📈 🧪
Stoked to see this paper finally out!
It answers two big questions: where visual objects are encoded in the brain, and how head-direction cells get oriented using visual landmarks.
Super fun collaboration with @mace-lab.bsky.social and Stuart Trenholm.
www.science.org/doi/10.1126/...
Thrilled to share this work, long time in the making! Carried through creatively by @siddjakes.bsky.social after initial design & piloting by @trackingskills.bsky.social, with help from @trackingactions.bsky.social. Modeling in collaboration with Massimo Vergassola & Nicola Rigolli.
01.09.2025 19:06 — 👍 125 🔁 42 💬 1 📌 2NWB just announced that they’re heading for a fiscal cliff next year. 😔
It feels like NWB was really just taking off in terms of data reuse — efforts like these take time and investment.
If you want to help push back their cliff, reach out to @bendichter.com
![What do representations tell us about a system? Image of a mouse with a scope showing a vector of activity patterns, and a neural network with a vector of unit activity patterns
Common analyses of neural representations: Encoding models (relating activity to task features) drawing of an arrow from a trace saying [on_____on____] to a neuron and spike train. Comparing models via neural predictivity: comparing two neural networks by their R^2 to mouse brain activity. RSA: assessing brain-brain or model-brain correspondence using representational dissimilarity matrices](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:e6ewzleebkdi2y2bxhjxoknt/bafkreiav2io2ska33o4kizf57co5bboqyyfdpnozo2gxsicrfr5l7qzjcq@jpeg)
What do representations tell us about a system? Image of a mouse with a scope showing a vector of activity patterns, and a neural network with a vector of unit activity patterns Common analyses of neural representations: Encoding models (relating activity to task features) drawing of an arrow from a trace saying [on_____on____] to a neuron and spike train. Comparing models via neural predictivity: comparing two neural networks by their R^2 to mouse brain activity. RSA: assessing brain-brain or model-brain correspondence using representational dissimilarity matrices
In neuroscience, we often try to understand systems by analyzing their representations — using tools like regression or RSA. But are these analyses biased towards discovering a subset of what a system represents? If you're interested in this question, check out our new commentary! Thread:
05.08.2025 14:36 — 👍 163 🔁 53 💬 5 📌 0
Check out our new review/perspective (w/ @juangallego.bsky.social & Devika Narain) on neural manifolds in the brain! It was a lot of fun to think through these ideas over the past couple of years, and I'm excited it's finally out in the world!
🔗: www.nature.com/articles/s41...
📄: rdcu.be/ex8hW
Attractors are usually not mechanisms - new blog post: open.substack.com/pub/kording/...
08.07.2025 14:40 — 👍 151 🔁 33 💬 20 📌 9Self supervised learning of spatial representations from episodic memories. Led by undergraduate student in my lab. Very proud!
28.06.2025 11:19 — 👍 8 🔁 1 💬 0 📌 0What's the best neural evidence that the brain does in-context learning? In other words, learning through activity dynamics rather than through synaptic plasticity.
27.06.2025 15:59 — 👍 42 🔁 6 💬 12 📌 0Thrilled to announce I'll be starting my own neuro-theory lab, as an Assistant Professor at @yaleneuro.bsky.social @wutsaiyale.bsky.social this Fall!
My group will study offline learning in the sleeping brain: how neural activity self-organizes during sleep and the computations it performs. 🧵
New preprint! 🧠🤖
How do we build neural decoders that are:
⚡️ fast enough for real-time use
🎯 accurate across diverse tasks
🌍 generalizable to new sessions, subjects, and even species?
We present POSSM, a hybrid SSM architecture that optimizes for all three of these axes!
🧵1/7
Manitokan are images set up where one can bring a gift or receive a gift. 1930s Rocky Boy Reservation, Montana, Montana State University photograph. Colourized with AI
Preprint Alert 🚀
Multi-agent reinforcement learning (MARL) often assumes that agents know when other agents cooperate with them. But for humans, this isn’t always the case. For example, plains indigenous groups used to leave resources for others to use at effigies called Manitokan.
1/8
I don’t think this is a misapplication. MI calculations assume ideal observers ; they’re agnostic about which decoder is actually used.
…..that makes mutual information a not-so-useful metric in neuroscience. Agreed on that. But using diff decoders than the brain isnt a misapplication of the term
This is the best one yet
19.05.2025 16:30 — 👍 5 🔁 0 💬 0 📌 0Let’s talk more about it offline sometime
15.05.2025 16:16 — 👍 2 🔁 0 💬 0 📌 0There is a famous paper by physicist Seth Lloyd who argues that there is a finite amount of information in the universe: 10^90
It’s based on the maximum entropy that the observable universe can contain, given its energy and volume.
Mutual Info is relative, but info/entropy isn’t. It’s super cool
My first-ever 1st author paper is now on bioRxiv 🙂 www.biorxiv.org/content/10.1...
Special thanks to my masters supervisor Steve Prescott!
Also with @lauramedlock.bsky.social , @cdedek.bsky.social & more
Many recent posts on free energy. Here is a summary from my class “Statistical mechanics of learning and computation” on the many relations between free energy, KL divergence, large deviation theory, entropy, Boltzmann distribution, cumulants, Legendre duality, saddle points, fluctuation-response…
02.05.2025 19:22 — 👍 63 🔁 9 💬 1 📌 0@apeyrache.bsky.social maybe one day if everything goes well 😉
29.04.2025 14:22 — 👍 2 🔁 0 💬 0 📌 0
Neural manifolds: more than the sum of their neurons — a Journal Club article by Juan Alvaro Gallego
www.nature.com/articles/s41...
@juangallego.bsky.social
#neuroscience #neuroskyence
Glad to see that it’s out!! Looking forward to reading
22.04.2025 21:23 — 👍 1 🔁 0 💬 0 📌 01/7: Super excited to share our new paper! This one should be of interest to neuroscientists and deep learning theory folks. This paper was a collaboration with Alexandre Payeur, @averyryoo.bsky.social, Thomas Jiralerspong, @mattperich.bsky.social, Luca Mazzucato, @glajoie.bsky.social
25.03.2025 15:59 — 👍 69 🔁 14 💬 6 📌 2
Basic pain researchers Steven Prescott and Stéphanie Ratté critique the clinical relevance of preclinical studies in the field and highlight areas for improvement.
By @sydneywyatt.bsky.social
#neuroskyence
www.thetransmitter.org/pain/basic-p...
One of our lab trainees, Nour Chahine, just dropped a track on Spotify, and it's 🔥 .
I wish I had done the same back in grad school… if I had any talent.
open.spotify.com/track/4t6E5u...
I joined this podcast to talk about sleep, dreams, and James Joyce. Yes, all in one conversation. It was a lot of fun!
🔊 Heads up: the audio is a bit rough at the start, but it clears up quickly.
Listen here: podcasts.apple.com/ca/podcast/b...
Top-down feedback is ubiquitous in the brain and computationally distinct, but rarely modeled in deep neural networks. What happens when a DNN has biologically-inspired top-down feedback? 🧠📈
Our new paper explores this: elifesciences.org/reviewed-pre...
One of the most surprising findings is that inhibitory cells are highly selective and even work together to suppress the activity of other brain cells.
www.nature.com/articles/s41...
You already know the Ship of Theseus, but do you also know the Ship of Thesis? It was constructed out of 3 loosely related papers which tell the tales of heroic quests and mythic psychodrama. It was only constructed once and then allowed to sink into oblivion.
31.03.2025 18:10 — 👍 59 🔁 8 💬 0 📌 0
