Often the process of learning how to do the science depends on having implemented it yourself, though. This is one example of many where LLMs benefit more experienced people in a different way to those with less experience - eg you gain intuition from doing the modeling yourself first
I like neuroscience and history, and over a decade ago I started a list of neurohistory events that influence my scientific views. I've been meaning to format as a timeline for a while, and doomscroll prevention + GenAI left me no excuse: neurojak.github.io/neurohistory/
Congratulations Roxana!
Congratulations!
How does an animal choose between exploring for a better food source and taking advantage of a known one? Our recent work in Current Biology demonstrates how recent feeding and metabolic state dynamically influence fly local search. bit.ly/3PfrIv3 #Science #Neurosky #Foraging #Drosophila
New paper hot off the (pre-)press! We dig into the evolutionary origins of neural computations for behavioral control across mice, monkeys, and humans: www.biorxiv.org/content/10.6....
As our lab's first foray into comparative analysis of neural dynamics, I’m super excited about this work! 1/18
Speakers include @clopathlab.bsky.social, Luke Coddington, Celine Drieu, @katenuss.bsky.social, @jpillowtime.bsky.social, Cristina Savin, and @saxelab.bsky.social.
Excited to be co-organising a #cosyne2026 workshop with Alison Comrie on 'algorithms for learning from scratch'! With a great line-up of speakers, we'll be tackling the question of what processes enable naive biological & artificial agents to adapt to new situations. Info here: tinyurl.com/4u8enf7k
A new preprint, co-authored with @johnwkrakauer.bsky.social:
The Deliberation Taboo
Cognitive science is, nominally, the science of thinking. We argue that the field has no theory of what thinking is and, even worse, that the topic has largely dropped out of focus. 1/
osf.io/preprints/ps...
I used to have a collection of fun rodent illustrations from papers, you’ve inspired me to hunt it down again…
The biggest problem holding neuroscience back right now isn’t data or tools, thanks in large part to the BRAIN Initiative.
It’s fragmentation across species. I wrote this to hopefully spark discussion around an issue that can only be solved as a community👇
www.thetransmitter.org/animal-model...
Excited to share “Orofacial behaviors, not eye movements, govern neural activity in mouse visual cortex”
www.biorxiv.org/content/10.6...
Summary below...
This is not limited to #zebrafish but includes all organismal research. If you can spare 15 minutes, you may help safeguard the future of responsible animal research in the UK. 🙏
Statistical Tests as Thought Experiments
“The only populations that can be referred to in a test of significance have no objective reality, being exclusively the product of the statistician’s imagination through the hypotheses which he has decided to test” (Fisher, 1956, p. 77).
BG GRC Oath: I acknowledge that the go/no-go model of the BG was useful but it is outdated and incorrect, or at least incomplete. I pledge not to use the go/no-go model as a strawman to motivate my work.
Taking the #GRCBasalGanglia Oath ✋🏼
I acknowledge that the go/no-go model of the BG was useful but it is outdated and incorrect, or at least incomplete. I pledge not to use the go/no-go model as a strawman to motivate my work.
Also love the idea of a more holistic view of what a manuscript even is (a collection of resources rather than just the main text)
It’s just kind of refreshing not to have to do so much hunting around for (what feels to me) some crucial details that are often not made explicit in the hybrid methods/results format style papers.
Again, I do understand where you’re coming from though, especially when it’s these big chunky sys neuro papers and it feels like a dense methods section might break the narrative.
These papers still have additional, more detailed, supplementary methods. But having a dedicated methods section in the main text gives a nice baseline understanding of some of the more critical aspects.
I’ve actually found reading papers with this other format surprisingly clarifying! You go into the results with a better idea of why the authors chose their approach, there’s less “oh wait did they think about this?”
Yeah, I see your points. Perhaps I’m confounding ordering of methods/results with honesty/completeness. Moving methods out of the main paper narrative makes it easier to shape or skew said narrative, I think
I was also thinking about elife! Cool that they were open to this format (even if “just” in the context of a methods paper)
Ah, yeah I’m not blaming authors per se…but as a field this somehow became an accepted norm. Agreed journals drive a lot of these decisions. Maybe we can use preprinting to take back some control over this. Was just highlighting something that is easy to take for granted
Sure, but why? I do think it’s field-specific to an extent too - most psychology or even human neuroscience journals format with methods after intro.
Or maybe it’s more a function of journal? Either way I’m fully convinced that the rigor of a lot of science would increase if we were all forced to put our methods front and centre
I’ve recently read a couple of papers (mostly from psychopharm) where the methods are placed after the intro and not buried at the end of the paper. It’s so refreshing! N animals excluded, plans for statistical tests, etc. all up front and easy to access. Why did we stop doing this in sys neuro?
Junior scientists 👉 applications are open for our workshop on the mechanistic basis of #cognition. 🧠
🤝 Joint sessions with our #TheoreticalNeuroscience workshop
✈️ Hotel, meals + reasonable travel expenses covered
Apply by May 7 ➡️ janelia.news/CNW26
@michaelreiser.bsky.social @jvoigts.bsky.social
Beautiful work as always Drew, congrats!
My final year group project as an undergraduate was in visual psychophysics and the effect sizes were so large we only needed an n of 4 participants - which ended up just being ourselves!
Right, agreed - depends a lot on whether there’s a model that’s a sort of foregone conclusion or whether you actually want to evaluate the model. Going back to your original question, as an experimentalist first I understand the desire to look at the “ground truth” first to build intuitions
