Alessandro Galloni

Very cool! Maybe it's just my bad intuition, but I find it surprising that weights can tolerate more extreme quantisation that delays

Exploiting heterogeneous delays for efficient computation in low-bit neural networks Neural networks rely on learning synaptic weights. However, this overlooks other neural parameters that can also be learned and may be utilized by the brain. One such parameter is the delay: the brain...

Psst - neuromorphic folks. Did you know that you can solve the SHD dataset with 90% accuracy using only 22 kb of parameter memory by quantising weights and delays? Check out our preprint with @pengfei-sun.bsky.social and @danakarca.bsky.social, or read the TLDR below. 👇🤖🧠🧪 arxiv.org/abs/2510.27434

With my great advisors and colleagues, @achterbrain.bsky.social @zhe @danakarca.bsky.social @neural-reckoning.org, we show that if heterogeneous axonal delays (imprecise) can capture the essential temporal structure of a task, spiking networks do not need precise synaptic weights to perform well.

It's been a pleasure and a privilege, going to really miss working with you and with the rest of the lab!

Really enjoyed reading this short opinion piece by Tim O'Leary. I think it echoes the classic Feynman quote "what I cannot create I do not understand". I think engineering approaches such as neuromorphic computing will prove fundamental to scientific understanding of how biological brains work

Reminder this is happening this Wed/Thu. Free spiking neural network conference - registration required (see below).

Falling asleep follows a predictable bifurcation dynamic - Nature Neuroscience Li et al. propose a conceptual framework to study the phenomenon of falling asleep based on electroencephalogram data. They show that a tipping point marks the brain’s nonlinear wake-to-sleep transiti...

My co-authors have yet to move to Bluesky, so I'm pleased to announce our latest work has just been published in @nature.com Neuroscience. Amazing work led by Junheng Li, revealing that falling asleep follows a predictable bifurcation pattern #neuroskyence #sleep
www.nature.com/articles/s41...

Microsoft Research Lab - New York City - Microsoft Research Apply for a research position at Microsoft Research New York & collaborate with academia to advance economics research, prediction markets & ML.

MSR NYC is hiring senior researchers in AI, both broadly in AI/ML & in specific areas (post-training, test-time scaling, modular transfer learning, science of deep learning).
aka.ms/msrnyc-jobs

We're reviewing on a rolling basis, interviews in Nov/Dec. Please apply here: tinyurl.com/MSRNYCjob

Ah nice, that's a good trick

Good tips! The only thing I dislike about this workflow is that the ai file with links is no longer self-contained, so harder to share with collaborators. I usually go straight for multi-panel in python (using Gridspec), and only use illustrator for adding the panel letters

Want to make publication-ready figures come straight from Python without having to do any manual editing? Are you fed up with axes labels being unreadable during your presentations? Follow this short tutorial including code examples! 👇🧵

Check out my new article: vitalstatistics.kareemcarr.com/p/a-simple-e...

Message for participants of the #SNUFA 2025 spiking neural network workshop. We got almost 60 awesome abstract submissions, and we'd now like your help to select which ones should be offered talks. Follow the "abstract voting" link at snufa.net/2025/ to take part. It should take <15m. Thanks! ❤️

#goals

A thought-provoking perspective from the visionary @giacomoi.bsky.social, calling for neuromorphic computing to return to its root: fundamental neuroscience; an inspiring vision for the future of NeuroAI 🤩

Travel awards are available for undergraduate students looking to attend #Cosyne25! The application is short, and the deadline is Nov 12. @cosynemeeting.bsky.social

Application: shorturl.at/6NEyk

More info: www.cosyne.org/travel-grants

(1/2)

I’m super excited to finally put my recent work with @behrenstimb.bsky.social on bioRxiv, where we develop a new mechanistic theory of how PFC structures adaptive behaviour using attractor dynamics in space and time!

www.biorxiv.org/content/10.1...

Q: In current clamp, does the delivered current get filtered by Rs?

A: No, the current is truly a square step with accurate amplitude and timing (as long as you don’t saturate the op-amp). The main source of error here is the current leaking out through a bad seal.

Q: In V-clamp (without compensation), does the cell membrane charge up slowly according to the membrane timeconstant tau=Rm*Cm?

A: No, the membrane charges up slowly according to tau=Rs*Cm, i.e. the speed of charging depends on *series resistance (Rs)*, not the membrane/leak resistance Rm.

Q: The pipette capacitance (Cp) causes an RC filtering/transient. Does “R” here refer to the pipette resistance? Will smaller pipette opening affect the capacitive artifact?

A: No, the R here is neither the Rs nor Rleak. It is the sum or resistances in *series* with Cp (e.g. junction resistance)

Q: Does access R affect the recorded voltage in current clamp if I'm not injecting current?

A: There is no DC error if you are not passing current, but the waveform is still affected by the pipette capacitance (tau=Rs*Cp), making fast events like APs smaller and wider

And here are some more FAQs ((I encourage students to not take my word for it, but verify these things for yourself with the simulator!):

Here are a few other great resources I came across while building this:
www.billconnelly.net?p=310
swharden.com/blog/2020-10...
www.bio.ens.psl.eu/~barbour/

Here is the link to the simulator if any of you want to play with it. Or share it with your students if they are learning about electrophysiology!
tinyurl.com/patch-clamp-...

I added sliders to let you easily play around with all the key variables (series/access R, Rs, capacitance comp, etc)

Electronics can be pretty unintuitive. I've found that being able to directly see and manipulate all the variables in a "white-box" system like this makes it much easier to grok what is going on under the hood. Active learning ftw!

On the other hand, Rs compensation in Voltage Clamp is an active circuit that directly drives the cell voltage higher (so it's very important to get it right online!)

Because of the positive feedback loop (command voltage driven by recorded current), it is prone to unstable oscillations.

For example, does bridge balance compensation affects how much current is being delivered to the cell?

Answer: no, bridge balance is a cosmetic DC correction to the recorded voltage (proportional to delivered current). It doesn't influence anything about the actual pipette/cell/injected current

Many students (and practitioners!) get confused about how series resistance, pipette capacitance, and various electronic compensation circuits affect your recordings.

Now you can see for yourself in the simulator!
tinyurl.com/patch-clamp-...

Hey ephys friends! I've been teaching electrophysiology for a few years now, and this summer while teaching at CSHL I started building an interactive simulation of a full patch-clamp amplifier circuit to help explain electronics to my students.

Sharing it here in case anyone finds it useful!

A generic non-invasive neuromotor interface for human-computer interaction - Nature A high-bandwidth neuromotor interface offers performant out-of-the-box generalization across people.

The CTRL-Labs decoding model paper is out! Saw this presented at Cosyne this year, very cool to see it out.

I would say this is the clearest demonstration of scaling laws in neural decoding to-date.

www.nature.com/articles/s41...

🧠📈 🧪