Today in a supermarket in France I noticed they had two separate aisles, one for patisseries and one for patisseries industrielles. I often wish I'd never left.
Dan Goodman
@neural-reckoning.org
Computational neuroscientist at Imperial College. I like spikes and making science better (Neuromatch, Brian spiking neural network simulator, SNUFA annual workshop on spiking neurons). 🧪 https://neural-reckoning.org/ 📷 https://adobe.ly/3On5B29
If you need a strategy just to take your annual leave, something is wrong.
Five ways academics can protect their annual leave
Organisation and communication will ease the way for university staff to book guilt-free time off and prevent burnout
timeshighereducation.com
This is a really great and well researched article for anyone interested in current AI and its limitations. Did you know you can replace the text in "reasoning" trace with something else without changing performance? These models don't work for the reasons we think they do.
Quanta Magazine@quantamagazine.org · 5d agoFrom this week’s Qualia column: “I’ll just say it: What the hell is going on with AI ‘reasoning’?” www.quantamagazine.org/is-ai-reason...
This is an interesting read, also the comment thread and quote posts and their threads. I'm not going to comment because I don't want to get embroiled (I'm on holiday right now), but if you're interested in politics of science reform, check it out.
Jeff Lockhart@jwlockhart.bsky.social · 5d agoI wrote about science funding, science reform, and anti-science agendas. scatter.wordpress.com/2026/07/31/s...
What does synapse position buy a neuron? We trained a dendritic network to place its own synapses on a covariance task. It clustered correlated inputs onto shared branches, and shuffling that placement destroyed performance while every weight stayed intact. Preprint: arxiv.org/abs/2607.24503 🧵-->
Synaptic clustering emerges from learning and supports covariance discrimination
Functional synapse clusters (FSCs) are synapses with correlated presynaptic activity that are colocalized on the same neuronal dendritic branch. FSCs have been observed after learning in cortical and ...
arxiv.org
My postdoc is coming to an end, so I'm looking for work! Let me know if you, or anyone you know, wants to hire a philosopher of biology/cognition with a knack for cross-disciplinary work, a focus on autonomy + agency, and experience writing for the broader public (10+ yrs in sci/tech journalism).
Since we wrote 'Larger GPU-accelerated brain simulations with procedural connectivity' (doi.org/10.1038/s435...), GPU memory capacities have 🚀. In our newish paper (doi.org/10.23919/PDP...), we explore what this means for simulating large spiking brain models using GeNN (genn-team.github.io).
doi.org
This is super cool! Just tried it on my most recent biorxiv preprint and it did a fantastic job (with just a few glitches highlighted in next post for @curvenote.com people). reader.openrxivlabs.org/content/10.1...
Neuromodulation enhances the capability and efficiency of spiking neural networks
Spiking neurons underlie the brain’s extreme energy efficiency, and therefore have great potential in neuromorphic computing, although realising this efficiency in practice has proven challenging. We ...
reader.openrxivlabs.org
openRxiv@openrxiv.bsky.social · 2w agoThe Curvenote Reader button is live on bioRxiv! 😎 Click it to explore papers in a new way—zoom figures, pull up references without leaving the page. Button appears 48 hours after papers drop. Feedback? DM us or hit the GitHub repo. openrxivlabs.org/articles/rea...
Very nice review of energy efficiency mechanisms of the brain and how and when they might apply to AI systems. I've been realising over the last year or so that this is actually the main unifying theme of my recent research (without explicitly planning it).
megan peters 🧠@meganakpeters.bsky.social · 2w agoI am *thrilled* to share this review in @currentbiology.bsky.social out today, by yours truly, @myoo.bsky.social, @michalk.bsky.social, Taro Toyoizumi, @tyrellturing.bsky.social, @taylorwwebb.bsky.social, & @hakwan.bsky.social www.sciencedirect.com/science/arti... 🧵👇
Beautiful work from the Neural Reckoning lab, Neuromodulation enhances performance and energy efficiency of Spiking Neural Nets!
Dan Goodman@neural-reckoning.org · 3w agoNew preprint (well, very updated). 🤖🧠🧪 We find that an abstract model of neuromodulation lets spiking neural networks perform much better, particularly in challenging noisy environments, using less energy. Relevant to #neuroscience and #neuromorphic computing. 🧵👇 www.biorxiv.org/content/10.1...
Everyone is praising this article, but I find it a bit disingenuous. He criticises Taylorism ("scientific management") but that's exactly what researchers are subject to, and the glam publication economy is at the heart of it. www.science.org/doi/10.1126/...
AI in scientific publishing: Slower, worse, and more expensive
There’s a saying in the management world, popularized by NASA administrator Daniel Goldin in the 1990s, that the goal of technological improvements is to make products faster, better, and cheaper. Alt...
science.org
After my lecture at the FENS-CHEN Cambridge summer school, I shared this coding tutorial (pytorch): It shows how to fit RNN (and spiking RNN) to electrophy to predict optogenetic perturbation from cortical recordings. github.com/guillaumeBel...
GitHub - guillaumeBellec/perturbation_testing_tutorial
Contribute to guillaumeBellec/perturbation_testing_tutorial development by creating an account on GitHub.
github.com
New preprint (well, very updated). 🤖🧠🧪 We find that an abstract model of neuromodulation lets spiking neural networks perform much better, particularly in challenging noisy environments, using less energy. Relevant to #neuroscience and #neuromorphic computing. 🧵👇 www.biorxiv.org/content/10.1...
Neuromodulation enhances the capability and efficiency of spiking neural networks
Spiking neurons underlie the brain’s extreme energy efficiency, and therefore have great potential in neuromorphic computing, although realising this efficiency in practice has proven challenging. We ...
biorxiv.org
This is why I’ve gone from “never want to touch a spiking network again” to “excited to do the kinds of things I always wanted to do with spiking networks” in the span of 5 years.
Dan Goodman@neural-reckoning.org · 3w agoI just gave a short tutorial on what is a spiking neural network and how to train one. From zero to a trained network in one hour. Slides and code you can run in Colab at this repo: github.com/neural-recko... 🤖🧠🧪
I just gave a short tutorial on what is a spiking neural network and how to train one. From zero to a trained network in one hour. Slides and code you can run in Colab at this repo: github.com/neural-recko... 🤖🧠🧪
GitHub - neural-reckoning/cambridge-fens-chen-summer-school-2026: Code and slides for 2026 FENS Chen Summer School on Learning with spikes
Code and slides for 2026 FENS Chen Summer School on Learning with spikes - neural-reckoning/cambridge-fens-chen-summer-school-2026
github.com
Is the brain a computer solved in 4 posts so that we never have to talk about this again (please please please). 1. It's not the same physically as a computer - it's made of meat. Hopefully this one at least is uncontroversial.
I suspect this sort of thing is inevitable. If you are part of a publicly listed company you will eventually have to make decisions in favour of profit even when they conflict with your ethics. For me, the fact that these companies all sell adverts is anyway arguably as harmful as military.
Andreas Kirsch@blackhc.bsky.social · 3w agoI work at Google DeepMind. This won't make me popular. But it's all public reporting: 2014: DeepMind reportedly sold to Google on conditions: no military use, independent oversight 2026: a Pentagon contract for "any lawful government purpose" Not one safeguard survived intact
Am I reading the UKRI strategy correctly that they want to speed up grant decision making by using AI? 😮💨 Guess it's time to do some experiments on what ChatGPT likes to see in a grant application. www.ukri.org/publications...
UK Research and Innovation Strategy 2026 to 2031
UK Research and Innovation's (UKRI) strategy for 2026 to 2031. This document supersedes the UKRI 2022 to 2027 strategy.
ukri.org
New Preprint! We gave neural networks infant-like curiosity and let them free to explore. They grew brain-like architecture, followed human synaptic development, and even started to display compositional thought. So we think curiosity doesn't just come from a complex mind... it helps us build one 👇
Curiosity shapes brain-like architectures and functions
How does complex cognition emerge from simpler underlying processes? We show that two components are sufficient: infant-like curiosity and brain-like biophysical constraints jointly drive the emergenc...
biorxiv.org
Registration is now open for the Paris Circuit Dynamics Conference & Workshop 2026! Neuromodulation of Neuronal Circuits: From Behavior to Computation Paris Brain Institute 16–17 September 2026 Registration is free (but required): www.paris-circuits.eu
Paris Circuit Dynamics
This conference will bring together leading experts to discuss how neuromodulatory systems shape neuronal circuit dynamics, behavior, and computation. The program will focus on how neuromodulators suc...
paris-circuits.eu
Dear peer reviewers everywhere: sometimes we feel pressure to find issues with a submission as a demonstration of diligence. But we have to recognise this is a mistake: picking random and sometimes incoherent holes in research add confusion, generates unnecessary work and encourages defensive
A good experiment-theory collaboration takes _________. Wrong answers only.
Marcus Ghosh@marcusghosh.bsky.social · 5mo agoHow 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...
Our next workshop will be at @imperial-ix.bsky.social -> 16.07 from 16:00 With a talk and discussion hosted by @sdrsd.bsky.social Registration is free, but required! www.eventbrite.co.uk/e/understand... See you there
Marcus Ghosh@marcusghosh.bsky.social · 5mo agoHow 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...
If you're in London and want to sit in air conditioning (so I am told) tomorrow afternoon, you can also get a free lecture out of it! Come join me to talk about consciousness, neuroscience, & adversarial collaborations tomorrow afternoon at the Crick! www.eventbrite.co.uk/e/understand...
Understanding the brain in theory and practice
A series of interactive workshops on building better experiment-theory collaborations in neuroscience.
eventbrite.co.uk
Our next workshop will be at the @crick.ac.uk -> 25.06 from 16:00 With a talk and discussion hosted by @meganakpeters.bsky.social Registration is free, but required! www.eventbrite.co.uk/e/understand... See you there
Marcus Ghosh@marcusghosh.bsky.social · 5mo agoHow 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...
New neuromorphic/SNN paper with @pengfei-sun.bsky.social Zhe Su, @achterbrain.bsky.social @giacomoi.bsky.social and @danakarca.bsky.social: www.nature.com/articles/s42... Long story short: neat trick to augment SNNs with a tiny memory buffer to improve performance at low energy/param cost. 🤖🧠🧪
Algorithm–hardware co-design of neuromorphic networks with dual memory pathways - Nature Machine Intelligence
Pengfei Sun et al. develop a spiking neural network with a dual memory pathway, co-designed with a custom neuromorphic chip. The approach delivers over 4× throughput and 5x energy efficiency gains whi...
nature.com
Interesting Transmitter article by Brian DePasquale on what will be the effect of the increasingly widespread use of LLMs to code up research ideas in neuroscience. Includes a somewhat sceptical commentary from me. 😉
The Transmitter @thetransmitter.bsky.social · 2mo agoAgentic coding makes it possible to specify a neuroscience model in hours instead of months, writes @briandepasquale.bsky.social. The field risks becoming prolific but shallow—generating models faster than we can generate insights. #neuroskyence www.thetransmitter.org/the-big-pict...
Right, time for a social media break. Not sure how long. See you on the other side (or maybe not, who knows). Might drop back to post new papers.
Here's some quick thoughts on something I've been thinking about for a while: One of the books that most influenced me about cognitive science was The Modularity of Mind by Jerry Fodor (1983). (1/N)
