Isn't this what Frontiers tried, and it just failed because the incentives for authors to engage is infinitely (literally) larger than the incentive for reviewers to engage? Would need a system where reviewers are incentivised to engage
People are great.
Yes very very impressive Scott Waddell
Oh my god, the alt text is a gift. 😹
This is awesome. Kam and Pip are both great!
My son has his version of Cosyne today :)
This is really cool!
I’m not so familiar with the data here but I’ll believe you. The most convincing evidence I’ve seen comes in owls, but it seems like there were still open questions? But I really haven’t followed closely.
Obviously in each case the understanding has opened up new questions. I’m not saying we should stop working in these systems!! Just that there is something extremely satisfying in each case.
The face patch system is a thing of beauty (thanks Doris,Nancy, Winrich et al.)
but our explanations in neocortex are a very different depth to those above.
Any in neocortex? I don’t think so. Vision is obviously a candidate but our best models for explaining the data are very non-biological.
But some aspects of vision are extremely satisfying. For example the early visual filters have explanations from connectivity to normative theory.
(7) Fly experts might correct me here but it feels like the mushroom body algorithm for associative learning and valence learning is getting there.
(8)somatogastric cpg in crabs (well done Eve)!
(9) integration of fear signals to trigger an escape reflex in rodents. (Well done Tiago:)
(5) Simple cerebellar-like circuits for prediction (eg electric-fish self-field cancellation).
(6) VTA- accumbens loop for TD-like learning ( obviously there is still debate here but the most recent demonstration by Uchida-lab at Cosyne really felt like a very direct demonstration).
(1) ring attractor in central complex of flies
(2) grid cell circuit for path integration in rodents (particularly with the amazing recent unpublished results from Mosers at SfN)
(3)song learning circuit in zebrafinch (we can now even train individual neurons by manipulating feedback).
What are the systems in neuroscience that we really have something that we can call “explanation” at all relevant levels, other than reflexive feed-forward like circuits.
Here are a few that I would argue are getting there. Obviously not complete explanations but genuinely satisfying.
Scientific publishing: Rethinking how research is reviewed and published
Review of how the loss of impact factor affected submissions at eLife - uneven drop across countries, but generally holding up remarkably well and shows a new model is possible
elifesciences.org/articles/110...
As I've said, the biggest obstacles to innovation in sci comm are Impact Factor & PubMed [for many, without one/being listed, you might as well not exist, and the inclusion criteria constrain process, format, etc].
bioRxiv shows you can get by without them. eLife may do it for peer review venues.
Your regular reminder of the extent of the problem
Hiring, promotion and tenure committees, grant review boards, institutional leadership, and every scientist whose knee jerk response to hearing of a new journal is to ask “what’s its impact factor?”, time to step up…
And if @elife.bsky.social were to achieve nothing more than cure scientists of Impact Factor addiction the project could be considered a triumph.
“Authors [&] reviewers tell us that they have had a more constructive experience…choosing eLife not because of our impact factor, but because of our process”
👆 eLife comtinues to provoke debate but this is perhaps the most important point elifesciences.org/articles/110...
It is absolutely astonishingly impressive how much work scientific publishers have managed to shunt onto authors, who aren't paid, while they continue to increase their enormous profits. The submission process itself takes about 5 hours. I miss the 90s 😢
Thanks Mark :)
Cosyne is always the best week of the year but this year it was just spectacular. Congrats @neurokim.bsky.social, @markhisted.org, @glajoie.bsky.social, and Alex Cayco Gajic for one of the best programs ever.
💯agree with the conclusion of the editorial - “Scientific publishing must evolve for the modern era. It must promote science but also provide a framework for honest discourse around science”
1/ Three years ago, we completely changed the way we review and publish research articles.
Today, we’re taking a look at how it’s going and what’s next.
buff.ly/aHwYNIE
The last year has been an extraordinary journey for us at @elife.bsky.social as we have been establishing a new model of publishing and taken on some of the commercial forces in publishing. We have tried to capture some of the takeaways in this editorial.
Really neat work by Fountas and colleagues at UCL:
arxiv.org/abs/2603.04688
They propose that consolidation reflects a form of "predictive forgetting" that aids generalization.
As eLife editor, it has been exciting and we are in a new steady state: quality of papers remains high, variance has dropped; work during the editorial phase has gone up tremendously; review quality is excellent and has improved. Kudos to @behrenstimb.bsky.social and the leadership team.
The last year has been an extraordinary journey for us at @elife.bsky.social as we have been establishing a new model of publishing and taken on some of the commercial forces in publishing. We have tried to capture some of the takeaways in this editorial.