After several years of work, my lab is starting to put out our first papers on learning in a unicellular organism (Stentor coeruleus).
Here we show evidence for a form of associative learning in Stentor:
www.biorxiv.org/content/10.6...
How can we build better experiment-theory collaborations in neuroscience?
We're hosting a series of 6 in-person, interactive workshops to discuss this.
Come along!
www.eventbrite.co.uk/e/understand...
Can we guarantee the behavior of an RNN to generalize well for infinite time? 🧠♾️
Similar to universal approximation theorems in deep nets, for systems that forget everything eventually, there are guarantees. We prove it for multistable systems!
arxiv.org/abs/2602.08640 w/ @memming.bsky.social
(1/6)
Our paper is out in @natneuro.nature.com!
www.nature.com/articles/s41...
We develop a geometric theory of how neural populations support generalization across many tasks.
@zuckermanbrain.bsky.social
@flatironinstitute.org
@kempnerinstitute.bsky.social
1/14
neuroAI comparisons of ANNs to brains do have a range of problems. Even more than I had realized. And I was worried before: www.biorxiv.org/content/10.1...
The hippocampal map has its own attentional control signal!
Our new study reveals that theta #sweeps can be instantly biased towards behaviourally relevant locations. See 📹 in post 4/6 and preprint here 👉
www.biorxiv.org/content/10.6...
🧵(1/6)
Grumpy from Disney's Snow White and the 7 dwarves.
Let's talk about "grumpy lab person". Many labs have them. With an eye to keeping science at its most rigorous, they cross the line into criticism that's too harsh. They are the ones who risk killing your scientific spirit. They are reviewer 2. /1
Applications for 2026 entry to the Gatsby Bridging Programme (7-week maths summer school) will open on 19 Jan and close on 16 Feb. Designed for students who wish to pursue a postgrad research degree in theoretical neuroscience or foundational machine learning but whose degree programme lacks a strong maths focus. Applications from students in underrepresented groups in STEM strongly encouraged. A small number of bursaries available. Register for the information webinar on 23 Jan.
📢 Applications open on 19 Jan for the 7-week #Mathematics #SummerSchool in London. You will develop the maths skills and intuition necessary to enter the #TheoreticalNeuroscience / #MachineLearning field.
Find out more & register for the information webinar 👉 www.ucl.ac.uk/life-science...
We are very excited to announce that our new preprint with Saleh Esteki, @stefanofusi.bsky.social, and @roozbehkiani.bsky.social is now available on biorxiv! www.biorxiv.org/content/10.6.... We investigated how reward context is learned, represented, and updated to bias decisions. Thread 🧵👇! 1/13
Please spread the word🔊My lab is looking to hire two international postdocs. If you want to do comp neuro, combine machine learning and awesome math to understand neural circuit activity, then come work with us! Bonn is such a cool place for neuroscience now, you don't want to miss out.
When and why do modular representations emerge in neural networks?
@stefanofusi.bsky.social and I posted a preprint answering this question last year, and now it has been extensively revised, refocused, and generalized. Read more here: doi.org/10.1101/2024... (1/7)
Move over ReLU 🚀
Meet **ReSU** (Rectified Spectral Unit): a biologically inspired, self-supervised unit for learning from dynamical data. A backprop-free multilayer ReSU network learns predictive features and recapitulates *Drosophila* vision.
To appear at AAAI: arxiv.org/abs/2512.23146
Just published my review of neuroscience in 2025, on The Spike.
The 10th of these, would you believe?
This year we have foundation models, breakthroughs in using light to understand the brain, a gene therapy, and more
Enjoy!
medium.com/the-spike/20...
New Perspective from myself, Sarah Heilbronner and @myoo.bsky.social . “Rethinking the centrality of brain areas in understanding functional organization” in Nature Neuroscience. 🧵
rdcu.be/eVZ1A
Amid the rise of billion-parameter models, I argue that toy models, with just a few neurons, remain essential—and may be all neuroscience needs, writes @marcusghosh.bsky.social.
#neuroskyence
www.thetransmitter.org/theoretical-...
Our paper on data constrained RNN that generalize to optogenetic perturbations now citable on eLife:
doi.org/10.7554/eLif...
📢Are you a young PI or senior postdoc in neuroscience looking to embark on the challenging journey of starting your own lab?
📆Then save July 5, 2026 in your calendar and join us in Barcelona for the second Young PI Symposium.
youngpisymposium2026.com
Dimensionality reduction may be the wrong approach to understanding neural representations. Our new paper shows that across human visual cortex, dimensionality is unbounded and scales with dataset size—we show this across nearly four orders of magnitude. journals.plos.org/ploscompbiol...
I’m happy to share some recent work out in PLOS Computational Biology with @guille-martin.bsky.social and Christian Machens at @champalimaudr.bsky.social . We use neural coding and population geometry to study different perspectives on hippocampal remapping.
journals.plos.org/ploscompbiol...
That’s it, thanks for reading along. There’s a lot more details in the paper about the framework and its relationship to the multitude of experimental and computational studies of hippocampal remapping. Please let us know what you think!
Overall, all three remapping mechanisms are likely to be accurate depictions of hippocampal activity changes under different conditions. More generally, this perspective is not limited to the hippocampus and can be used to understand neural variability in other brain areas as well!
In the paper, we simulate several examples of each remapping type in RNN models and quantify the types of neural activity patterns underlying them.
Null-space (NS) remapping features changes outside of the latent space, i.e., in the null space, and relates to changes in which neurons participate in a redundant code. This relates to more subtle activity changes due to excitability, trial-to-trial variability, and even cell birth and death.
Mixed-selective (MS) remapping explains activity modulation through changes or shifts in non-spatial, cognitive axes. This relates to ideas about mixed or conjunctive representations of position with other variables such as context, behavior, or sensory info.
Using this framework, we identify three underlying coding mechanisms that explain remapping.
Encoder-decoder (ED) remapping reflects a true change in the latent space, visualized as a rotation of the latent axes. This relates to classic multi-chart and grid realignment models, among others.
This allows us to visualize hippocampal population activity along positional and cognitive latent axes, with localized “place fields” on top. From this perspective, neural trajectories are composed of a linear latent component, plus a nonlinear nullspace component.
Many theories exist for remapping, but what types of neural activity patterns are compatible with these different views? We study this question in a model with one key assumption: that hippocampal population activity reflects a linearly-decodable latent space.
I’m happy to share some recent work out in PLOS Computational Biology with @guille-martin.bsky.social and Christian Machens at @champalimaudr.bsky.social . We use neural coding and population geometry to study different perspectives on hippocampal remapping.
journals.plos.org/ploscompbiol...
Are manifolds real?
Are latent circuits real?
Tatiana @engeltatiana.bsky.social uses one, infers the other, and says yes to both.
Also, how timescales are different and the same across the entire brain...
braininspired.co/podcast/226/
Finally got the job ad—looking for 2 PhD students to start spring next year:
www.gao-unit.com/join-us/
If comp neuro, ML, and AI4Neuro is your thing, or you just nerd out over brain recordings, apply!
I'm at neurips. DM me here / on the conference app or email if you want to meet 🏖️🌮
