Wilson–Cowan. FitzHugh–Nagumo. Brunel. Kuramoto. The detailed models that motivated those reductions are mostly forgotten outside their original communities. The reductions are still being taught. A field note on why every mechanistic model deserves a small theoretical cousin.

Foundation models in neuroscience predict brain activity at unprecedented accuracy. But prediction ≠ understanding, and we should avoid conflating the two. New essay now out:

I tracked every keyword in 22 years of Cosyne abstracts to map how computational neuroscience evolved — from Bayesian brains to neural manifolds to LLMs — and where it's heading next.

Every computational model needs numbers. But most experimental data lives in figures, not files. On the hidden labor of turning papers into models, and why open data is infrastructure for theory.

The Chomsky hierarchy isn't a linguistic curiosity — it's a map of what kinds of memory architecture different classes of computation require. Do the boundaries between computational classes correspond to boundaries in neural architecture?

For 200 years, Weber's law was the most replicated finding in psychophysics but nobody could explain why it holds. Turns out the answer was hiding in the data everyone ignored — not how often you get it right, but how long it takes you to decide.

New Field Note: a Deep dive into the work of our guest lecturers this week. Three papers from the Dynamical Inference lab on contrastive learning for neural time series—CEBRA, DCL, xCEBRA. What identifiability actually means, when you need dynamics modeling, applications and implications.

New Journal Club: Neural manifolds are maturing from visualization trick to biological claim. But if population activity lives on low-dimensional manifolds, what constrains the geometry?

I simulated a grant application evaluation committee using Claude. The result was super interesting and useful! open.substack.com/pub/grounded...

"these structures served a purpose when disciplines were stable and problems were contained. They're actively harmful now." How traditional academia is making itself irrelevant with its outdated practices.

New Journal Club explores TARPs as multi-domain regulators that couple synaptic strength to intrinsic excitability—providing a molecular substrate for coordinated adaptation across timescales. open.substack.com/pub/grounded...

New opinion article: should we think about (declarative) memory as a continuum? open.substack.com/pub/grounded...

First journal club post: aiming for a weekly update on what I have been reading and annotating (deep dives), what new papers sparked my interest and joined the reading list (new papers on deck) and an attempt to make sense of it all. open.substack.com/pub/grounded...

Launching Grounded: biologically grounded computational neuroscience. Lab notes, Journal Clubs and Opinions made public. Networking through communication. Subscribe to follow our journey! open.substack.com/pub/grounded...

Seems accurate

This is so awesome!! With all feeds filled with politics and conflict, it's refreshing to be able to focus on science

Wow, I’ve just found a Bluesky channel that lets you see only posts with links to scientific papers and get a feed that focuses on the science discussions around them. This is transformative!

I am looking for 2 Research Assistants for 1-year positions to investigate biophysical mechanisms of learning (finally the formal procedures are open for applications)

Check out our latest pre-print: #Up-down states as a dendritic effect.

"If we don’t know what we’re looking for, we’ll never find it, no matter how sophisticated our analysis tools." Beautiful piece by Sam Gershman on how far we still are from true progress in neuroscience

🚨URGENT: The Computational Neuroscience & Neural Computation (CNNC) lab is looking to fill 2x MSc Scholarships to investigate the interplay between synaptic, dendritic and neuronal adaption. Please apply here (or forward to interested students): www.comp-neuro.org/contact

Cutting it a bit fine, but here’s my review of the year in neuroscience for 2024 The eighth of these, would you believe? We’ve got dark neurons, tiny monkeys, the most complete brain wiring diagram ever constructed, and much more… Published on The Spike Enjoy! medium.com/the-spike/20...

Out now in Nature from @behrenstimb.bsky.social and crew: www.nature.com/articles/s41... Understanding this kind of schematic pattern learning and transfer will be key, IMO, to moving towards models of what we might call "higher-order cognition" or "reasoning". 🧠📈 🧪

This can be an argument for why normative modeling (optimization) may fail beyond the sensory areas of the brain. In the case of ANN models, there won’t be a finite set of loss functions that can lead to learning the whole range of neural dynamics, due to the unboundedness of real-world scenarios.

OK If we are moving to Bluesky I am rescuing my favourite ever twitter thread (Jan 2019). The renamed: Bluesky-sized history of neuroscience (biased by my interests)

Order matters: neurons in the human brain fire in sequences that encode information www.nature.com/articles/d41...