Dan Goodman

There's no better way to learn than to teach! Help make NMA a resounding success!

If you take a paper and get an LLM to review it, then get the LLM to rewrite the paper and write a reply to the reviewer, and then repeat, what happens? Convergence to something better? Cycles of arbitrary change that never converge? Descent into meaningless drivel?

Is it worth noting that your article only describes new experimental techniques and not any actual new understanding of what the brain does? That actually feels like quite an accurate description of neuroscience in recent decades.

GitHub - kaurarmanjot445-sys/snn-mnist-conversion: Time-to-first-spike SNN conversion for MNIST - exact gradient equivalence implementation Time-to-first-spike SNN conversion for MNIST - exact gradient equivalence implementation - kaurarmanjot445-sys/snn-mnist-conversion

Very refreshing interaction with the undergrad student form India @kaurarmanjot445.bsky.social ! 🧪 🧠

She was interested in our math equivalence between spike timing and deep relu nets

With a short call and few emails she reproduced the work and open sourced a turorial
github.com/kaurarmanjot...

What a wonderful election result to wake up to. Greens not only win in Manchester but they win by a huge amount. Not even close. 💚

Gives me a little hope for the future.

PDFs are pretty enshittified these days. Adobe reader is awful (although yes there are alternatives).

Compact deep neural network models of the visual cortex Nature - Parsimonious deep neural network models can be used for prediction of visual neuron responses.

DNN models of the brain are getting bigger. Are we replacing one complicated system in vivo with another in silico?

In new work, we seek the *smallest* DNN models of visual cortex, balancing prediction with parsimony.

It turns out these compact models are surprisingly small!

rdcu.be/e5H8G

The indexing is a major feature holding back much needed innovation in my opinion.

Really looking forward to participating in this! What a great way to start my next professional chapter embedded in #london #neuroscience: talks on how to get experiment and theory working together better, extended networking/group discussion, and food!

Come join us!

Absolutely this! Discoverability is critical. A lot of the data is already out there.

I agree with others that living articles seems like a maintenance nightmare, if they were an obligation. On the other hand, I've very often wanted the option to add a minor new result to an existing paper without having to write a whole new paper. I do think there's potential to do better than now.

PDFs are pretty awful tbh but I'm fine with papers. Would be nice to have some modern conveniences like adapting to screen size, automatically showing related content so you don't have to flick back and forth, etc. Lots of scope to do better, no need to make it worse along the way!

That's the plan! 🤞

I think the time is right for us to do this seriously. The growth of "NeuroAI", large scale experimental projects and a steady stream of papers showing that earlier papers in both experiment and theory made overly strong conclusions because they didn't do this right.

London #neuroscience people you may like this. We're hosting a series of talks at Imperial & Crick on how to get experiment and theory working together better. Each session will have a talk around this and extended networking / group discussion on the questions raised. Plus, free food!

🤖🧠🧪

Pace of ecology drives the tempo of visual perception across the animal kingdom Nature Ecology & Evolution - Using phylogenetic comparative methods across 237 species from disparate phyla, the authors show that species with fast-paced ecologies have higher temporal...

Our new paper is now out showing how time perception in animals is linked to their ecology. Using data from 237 species we show temporal perception is faster in species that fly and pursuit predators www.nature.com/articles/s41... 🌐

Image description: Graph of median pay relative to 2018 adjusted for inflation shows it decreasing almost every year to 2025, where it is nearly 10% lower than in 2018. Picture of Unison, Unite, and UCU union members on picket lines with banners.

We can win a better pay offer in 2026!
We've suffered a 9% pay cut in real terms since 2018.

Management imposed another real-terms cut last year - insisting that a 2% increase was all they could afford - and then refused to negotiate with the unions. 2026 pay negotiations begin in April.

At least this one was obvious! I'm finding it harder and harder to tell these days.

This would be a good thing. Those evaluation metrics are the source of a lot (possibly the majority) of problems in academia, even before AI.

This ought to be obvious. If the metrics go up when people use AI despite the papers not being insightful, then it was never measuring what you wanted.

Also, is this extra money or is it taking away from the research budgets of other fields? If so, it's very problematic.

I'm somewhat persuaded that it's worth developing a "sovereign AI" that isn't dependent on the US, especially given developments over the last year. But if you're doing that you need to be much more ambitious than this don't you?

So what does everyone think of this? Only had a quick glance but there are some good things, like a focus on aspects that might be less well developed by big companies (eg sustainable AI). On the other hand, the total 4 year budget is about 1 months spend by just one big AI company.

Neuronal spiking in the mammalian forebrain is dominated by a heterogeneous ground state Neuronal firing patterns have significant spatiotemporal variability with no agreed-upon theoretical framework. Using a combined experimental and mode…

With this one in print, I think I finally earned that PhD... 😅
Presented for the first time at the cosyne when the world ended (March 2020). I'll bring over a summary thread from twitter when it was still twitter...

www.sciencedirect.com/science/arti...

I hope you added more points than you deducted? 😉

Maybe we should sue the universities too? They're certainly treating their staff at least as badly as the students.

This is the plot of Caprica, the prequel series of Battlestar Galactica. It ends with the robots murdering almost the entire human race. 😉

We just bought a new washing machine and it was really hard to find one that didn't have an app. Sigh. 😞

Yeah I agree. I guess I wouldn't take it personally though, they're just reacting to an injustice in the only way available to them. It sucks for everyone.

Certainly the students missed out on a key part of the university experience (the intense in person aspect). So I think there's space to recognise that we did incredibly well but that still the experience was less good for students. Didn't read the article though, so not sure if this is off base.

I sympathise because I also near broke myself updating all my teaching during COVID, and actually I think it's better than what I was doing before so I've kept the flipped classroom element. However, most people went back to the old ways, suggesting they do think it's better.