Arthur Prat-Carrabin

neuroscience + dance --- Harvard

[Itai Yanai @ItaiYanai]

Doing good science is 90% finding a science buddy to constantly talk to about the project.

[Arthur Prat-Carrabin @arthurpr4t]

If you need to combine old and new information, you need to remember how much to weigh one vs the other. If you're imprecise about it, but with a (Bayesian) central tendency, you'll overweight past evidence at the beginning, and new information at the end. ⇒ primacy + recency.

[Arthur Prat-Carrabin @arthurpr4t]

It turns out that primacy and recency effects are not that boring after all
https://t.co/TiMUVh5zCr

[Arthur Prat-Carrabin @arthurpr4t]

I have posted on OSF the data for the numerosity estimation and discrimination tasks discussed in my paper with Mike Woodford "Endogenous Precision of the Number Sense". https://t.co/lVcRfnTnQA Btw it's now on eLife as a reviewed, not-revised preprint https://t.co/AiMm5TEMd3

[Arthur Prat-Carrabin @arthurpr4t]

@edmundlalor @UofR @URNeuroscience Hi! Is there a deadline for this?

[Arthur Prat-Carrabin @arthurpr4t]

"This raises the crucial question to what extent variability in risky choices really reflects subjective valuation (as opposed to noisy perception), as usually assumed in standard economic models." Very cool!

[Arthur Prat-Carrabin @arthurpr4t]

Important work by @GillesdeH and @ChristianCRuff's group! Perceptual processes account for risk preferences, and their reversal. "inter- and intra-individual variability in risky choice behaviour can be explained by fluctuations in basic perceptual neurocognitive representations"

[GillesdeH @GillesdeH]

🎉 Thrilled to announce that the first project I worked on from start to finish in Zurich is now on bioRxiv! We show how risk preferences are fundamentally shaped by how we process choice variables, and how these can be measured in the parietal cortex.

https://t.co/qrTYXkv7Nh

[Arthur Prat-Carrabin @arthurpr4t]

@MBarrettoGarcia @BBRFoundation Awesome!! Congrats!

[Sam Gershman @gershbrain]

Arthur Prat-Carrabin has some new work on numerosity perception showing that Weber's law can be broken by manipulations of the stimulus distribution and reward function: https://t.co/Om6lB4CQSO
The pattern of results is broadly consistent with a Bayesian observer model.

[Sebastian Olschewski @Seb_Olschewski]

Great defining slide for what is Cognitive Economics presented by @webbr14 at #ESAworld24 @EcScienceAssoc

[Del Monte Institute for Neuroscience @URNeuroscience]

"I think that, actually, the brain is pretty clever."

Cognitive neuroscientist @KiaNobre of @WuTsaiYale explains why she doesn't believe limited resources drive the brain's behavior.

Watch Neuroscience Perspectives🎙️ w/@JohnnyFoxe ➡️https://t.co/ZVkD38mdVc

[Arthur Prat-Carrabin @arthurpr4t]

@SuryaGanguli That's efficient winning

[Arthur Prat-Carrabin @arthurpr4t]

Susan Sontag, 1964: "Most American novelists and playwrights are really either journalists or gentlemen sociologists and psychologists."

[Arthur Prat-Carrabin @arthurpr4t]

The question of how we represent stuff is fundamental and comes with many implications. This sheds some principled light on how it happens for simple stuff (1D variables), while exhibiting interesting experimental results.

[Arthur Prat-Carrabin @arthurpr4t]

Last month I posted a preprint that presents experimental and theoretical results about the brain's representation of magnitudes. We looked at the variability of subjects in both an estimation task and a discrimination task, and with different priors. https://t.co/lfeoEruL43

[Arthur Prat-Carrabin @arthurpr4t]

The cost we deduce is not found in other studies but it is the only one that explains all our results simultaneously. And it turns out that it makes a lot of sense: you "pay" for getting more signals about what you want to represent.

[Rava Azeredo da Silveira @Rava_daSilveira]

Applications are open for the CNeuro2024 summer school in computational neuroscience to be held in Beijing, 8-15 July 2024. Rolling admissions starting March 1. See https://t.co/5aaVG2zQdZ for details on program and application process.