Richard Gao

I'd say we hate splitters and root for the lumpers because unifying * p r i n c i p l e s *....but yeah, sadly, data don't lie

I've often had very pessimistic answers for this, but I think "new data shows that brain is much more complicated than we originally thought" is actual progress.

e.g., cortex is not homogeneous repeats of the same columns, cells are not all the same, it's not just single-neuron tuning curves, etc.

I'd have nothing to talk about on this episode if not for my paper co-authors in the @mackelab.bsky.social, as well as Pedro, Jan-Matthis, and Michael paving the way with SBI.

I'm extremely grateful for Gaute for having me on the show, especially considering the roster of previous guests.

All the work from his group advancing our understanding of the LFP has influenced me since day 1 of my PhD. They say never meet your heroes but I think this time it went pretty good.

Once you get past the first 5 minutes of me bumbling around clearly in need of media training, we settle into a nice groove talking about fitting mechanistic models, ML/AI for neuroscience, underconstrainedness vs. degeneracy, and its potential benefits for biology (plus some good chuckles).

a man with a beard is asking what year is . ALT: a man with a beard is asking what year is .

your paper:

dang you're a real fan because that's a solid who-he-play-for

I woke up today and the Detroit Pistons are first in the East. what year is this??

"borrow"?

main goal for this year: find a new job! 🙂

looking for a role with fun & complex technical challenges & within a great community. my main expertise is in signal processing/EEG/MEG, but topic-wise I am quite flexible.

science/industry both great! starting mid-year. nschawor.github.io/cv

please only do bird research from now on

get in before the spots run out, otherwise put your money in the betting pool

we'll take what we can get

I also had a similar interpretation as you though, that as a first approximation, "pattern-y"=dimensionality, such that spikes destructively interfere.

same argument as spatial alignment of neurons and makes sense if one knows LFP, but I like this paper because it lays it out quantitatively

I think you have a much more nuanced view of spikes and LFPs than the strawhumans in my head Ben

just my perception, but I think people claiming spikes >> LFP rarely cite empirical evidence favoring one way or the other, but make that statement reflexively.

tbc I don't think LFP >> spikes universally, the much more interesting question is when one or the other, like this paper is showing.

but good drinking

you're a butt

I have lived many lives

I will die on this hill over and over until it becomes a mountain

🔥

this reminds me of when Scott showed PAC is waveform hahah

a man in a suit and tie talks to another man and says " some men just want to watch the world burn " ALT: a man in a suit and tie talks to another man and says " some men just want to watch the world burn "

noooo whyyyy

time for the loglog fight, v2

I was working with an undergrad RA 6 years ago on a power law project, and I shit you not the meme name we came up with for the project was galaxybrain.

I'm starting to think the argument between low vs. high dimensional neural activity literally just comes from plotting the PCs in linear vs. log y-scale

free energy tho

I think this by definition makes it the best workshop at neurips

Thrilled to start 2026 as faculty in Psych & CS
@ualberta.bsky.social + Amii.ca Fellow! 🥳 Recruiting students to develop theories of cognition in natural & artificial systems 🤖💭🧠. Find me at #NeurIPS2025 workshops (speaking coginterp.github.io/neurips2025 & organising @dataonbrainmind.bsky.social)