Twitter | Pretraživanje | |
Tony Zador
Neuroscientist at CSHL. Interested in neuroAI, connectomics, and auditory cortex.
542
Tweetovi
380
Pratim
3.365
Osobe koje vas prate
Tweetovi
Tony Zador proslijedio/la je tweet
Andrew Hardaway 2. velj
Where do people currently stand with the best opsin to use for pathway-specific inhibition?
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 1. velj
Odgovor korisniku/ci @generuso @neuroecology i 2 ostali
I'm not sure i understand. The way i think about it, the functional goals (eg vision) determine which circuits (eg ones with orientation columns) evolution selects for. no?
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 1. velj
Odgovor korisniku/ci @neuroecology
True. Although I was thinking that "journal papers" might also be considered to fulfill the same role.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 1. velj
also, I don’t think developmental plasticity is necessarily more efficient in general. I think it is complementary, more precise in that it has the potential to be more specific (sensitive to pair wise neural interactions)
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 1. velj
My guess is it was the other way: Hebbian learning hijacked activity dependent developmental plasticity. Not sure though
Reply Retweet Označi sa "sviđa mi se"
Tony Zador proslijedio/la je tweet
Adam J Calhoun 1. velj
In dark, tunicates (a small animal with only 177 neurons!) will swim upward. When it's light, they don't distinguish between up and down. How do they switch between these contextual behaviors? 👇how antennal cells and photoreceptors interact
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 31. sij
Exactly. Although people often distinguish between "activity-dependent" and "activity-independent" developmental rules, from a computational point of view the more relevant issue is whether the rule is "data-dependent" or "data-independent" (of data from the outside world).
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 31. sij
Some rules can be specified more compactly using activity. Suppose you need a L3 neuron to get input from L4 neurons driven by L or R eye (but not both). Could be done with molecular tags, but easier to do with a Hebbian winner-take-all circuit.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador proslijedio/la je tweet
NRDlab 29. sij
We are looking for postdocs or grad students to act as conference mentors for (aka ) undergraduate travel grant winners. Free registration for the main meeting and a chance to help students at the start of their journey. Apply here:
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 28. sij
Odgovor korisniku/ci @AdamHantman
completely agree. It's not either/or. The overall project can be curiosity based. And there can be room for creativity within the project. But it is easy to default to a top-down model that may not be satisfying for worker bees.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 28. sij
Team big science is great, as long as it's not at the expense of curiosity driven science A foundation of modern mobio, PCR, can be traced to fortuitous discoveries by people wondering how life exists in boiling water One example of many
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 28. sij
Odgovor korisniku/ci @AdamHantman
Bell labs used to give people a lot of freedom to pursue their ideas.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 28. sij
Odgovor korisniku/ci @AdamHantman
I think a lot of physicists fled physics (for neuroscience and other fields) precisely to get away from the "cog in the wheel" model of research. Some people like the freedom to pursue their own crazy ideas.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 27. sij
Odgovor korisniku/ci @tahoppe05404
I think this explains each army alone but not how the two armies maintain borders and keep the peace. (But I didn’t read beyond abstract)
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 27. sij
How does this work?
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 25. sij
Odgovor korisniku/ci @MatteoCarandini
I am genuinely puzzled by the idea of a virtual talk. Why not just make a youtube video, and perhaps answer questions at the end? Maybe even by email, or in the comments section? Or maybe transcribe the talk (nicely formatted, along with slides)?
Reply Retweet Označi sa "sviđa mi se"
Tony Zador proslijedio/la je tweet
Justus 25. sij
Odgovor korisniku/ci @TonyZador
So, what is this paper about? SYNPLA is a method to label synapses in a specific pathway that recently underwent LTP. The principle is simple. We exploit the fact that after LTP GluA1 temporarily moves into the synapse from extra-synaptic pools.
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 24. sij
Odgovor korisniku/ci @mchevillet @KordingLab
in what way is it interactive? I had the impression from that there wasn't much of a feedback channel...
Reply Retweet Označi sa "sviđa mi se"
Tony Zador proslijedio/la je tweet
Justus 24. sij
New paper out in PNAS today. Long-running collaboration between Malinow and labs brought to fruition by Kim Dore! Check it out. SYNPLA, a method to identify synapses displaying plasticity after learning
Reply Retweet Označi sa "sviđa mi se"
Tony Zador 23. sij
Odgovor korisniku/ci @KordingLab
Then these could be gathered into a compilation, and distribute to scientists in the field (maybe along with a monthly table of contents)?
Reply Retweet Označi sa "sviđa mi se"