Mathias Sablé-Meyer

Mathias Sablé-Meyer

@mathiassablemeyer.bsky.social

Postdoc working on human cognition about abstract concepts at the SWC, London. Cog/comp-(neuro)scientist wannabe; adequate climber.

295 followers89 following24 posts

Mathias Sablé-Meyer reposted

Mohamady El-Gaby
@melgaby.bsky.social

World models are central to intelligence - but how do we actually use them to behave flexibly? We review emerging findings on how the PFC’s inner dynamics translate structured knowledge into action. osf.io/preprints/ps... 🧵 1/15

Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

It was fun to pilot this! I wonder if eLife already has some data about what fraction of the readers uses this feature? elifesciences.org/articles/106...

elifesciences.orgClient Challenge
eLife@elife.bsky.social

Bring your research to life with an author explainer video. Whether you're talking through your figures or summarising your research, explainer videos can make your work more accessible and engaging. Our tips will help you create yours. elifesciences.org/inside-elife...

elifesciences.orgAuthor explainer videos: Top tips for creating your videosOur author explainer videos allow you to illustrate your research and figures, bringing them to life, but what makes a good explainer video? Read our top tips.

Mathias Sablé-Meyer reposted

Maxime Maheu
@maheump.bsky.social

Happy to be back! Europe *is* attractive to researchers. More stable positions would let more of us go after the difficult questions rather than the next low-hanging fruit. So many great projects didn't make it this year — I hope they all find a home! #ChooseEurope

European Research Council (ERC)@erc.europa.eu

Why do researchers #ChooseEurope? 🇪🇺🔬 In our recent ERC Starting Grants competition, 21 researchers currently based outside Europe were selected for funding; 20 of whom are currently based in the United States. Four new #ERCSTG researchers give their reasons 👇

Mathias Sablé-Meyer reposted

Maxence Pajot
@maxencepajot.bsky.social

New AI models now show incredible math skill, solving problems that had stumped mathematicians for years. But what about basic geometric abilities? In our new paper in PLOS Computational Biology, we investigate this in vision models: doi.org/10.1371/jour... 1/8

doi.orgCan neural networks model the human perception of geometric shapes?Author summary Recent neural networks have achieved remarkable success across a wide range of visual tasks, often rivaling or surpassing human performance. Yet, it remains unclear whether these system...

Mathias Sablé-Meyer reposted

Hope Kean
@hopekean.bsky.social

Is the Language of Thought == Language? A Thread 🧵 New Preprint (link: tinyurl.com/LangLOT) with @alexanderfung.bsky.social, Paris Jaggers, Jason Chen, Josh Rule, Yael Benn, @joshtenenbaum.bsky.social, ‪@spiantado.bsky.social‬, Rosemary Varley, @evfedorenko.bsky.social 1/8

tinyurl.comEvidence from Formal Logical Reasoning Reveals that the Language of Thought is not Natural LanguageHumans are endowed with a powerful capacity for both inductive and deductive logical thought: we easily form generalizations based on a few examples and draw conclusions from known premises. Humans al...

Mathias Sablé-Meyer reposted

Mohamady El-Gaby
@melgaby.bsky.social

It wouldn’t be FENS if there wasn’t some sort of clash - definitely check out awesome work by @aryabhomick.bsky.social in Friday morning’s poster session tracking learning dynamics of abstract task structure in the PFC as well as evidence for complex hierarchical learning in rodents!

Mohamady El-Gaby@melgaby.bsky.social

If you’re still around on Friday morning I’ll be giving a talk about the latest work and ideas from our lab - alongside lots of great theory and experimental talks in the “Latent representations for smart decisions” symposium. 5/5

View full thread
Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

No public funding can match Taylor's pockets when it comes to design. Why reinvent the wheel? We've K-mean clustered your favourite albums, then cherry picked. It’s as simple as scale_color(era.brewer("Lover1", n=5)) in R, or whatever incantation python uses. And it looks better than viridis 🙄

Mathias Sablé-Meyer reposted

Charley Wu
@thecharleywu.bsky.social

🚨 New preprint w/ Valerio Rubino and Peter Dayan: how do people discover and use compositional structure under constraints? osf.io/preprints/ps... A key factor is a simple heuristic that favors reuse of repeated and symmetric fragments across scales, is robust to time pressure, and sped up RTs 🧵👇

Mathias Sablé-Meyer reposted

Lorenzo Ciccione
@lorenzo-ciccione.bsky.social

If you have ever wondered what happens in your brain when you look at those scatterplots from your research data (hoping to find a significant correlation!)... check our paper out !

MIT Press@mitpress.bsky.social

New and #OpenAccess in the January issue of the Journal of Cognitive Neuroscience: The Neural Bases of Graphical Perception: A Novel Instance of Cultural Recycling? by @lorenzo-ciccione.bsky.social and Stanislas Dehaene: direct.mit.edu/jocn/article... @jocn.bsky.social @standehaene.bsky.social

direct.mit.eduThe Neural Bases of Graphical Perception: A Novel Instance of Cultural Recycling?Abstract. Graphical representations of quantitative data abound in our culture, and yet the brain mechanisms of graphicacy, by which viewers quickly extract statistical information from a data…
View full thread
View full thread
Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

We compare many models: various layers of CNNs, various neural network architectures, models based on shape skeletons or on lexical access to shape names, etc. In fMRI, we also zoom in on both ventral areas and areas previously identified as math-responsive.
 And there'll be more—stay 📈 𝒯𝓊𝓃𝑒𝒹 📉!

Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

We find that after visual processing, the mere perception of a geometric shape induces a representation based on exact geometric features. We connect this to work on a Language of Thought where perception triggers a reverse engineering of the shape's "program": www.sciencedirect.com/science/arti...

sciencedirect.comA language of thought for the mental representation of geometric shapesIn various cultures and at all spatial scales, humans produce a rich complexity of geometric shapes such as lines, circles or spirals. Here, we propos…
View full thread
Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

In a simple fMRI visual localiser, with no task, comparing single shapes to other single visual categories (faces, tools, houses) elicits an under-activation of the visual pathways. It also elicits an over-activation of (right) IPS in adults and children, close to typical math-responsive areas.

Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

During passive shape perception in MEG in adults, we find dissociation in space and time between two models of geometric shape perception. Early visual activity resembles bottom-up perception as embodied by a CNN. Subsequent parieto-frontal activity is best explained by geometric features.

Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

First, the behaviour: when faced with finding an intruder within quadrilaterals, the accuracy and response time of humans vary a lot and are well predicted by the exact geometric features of the shapes—more so than by their visual features embodied in a neural network.

Mathias Sablé-Meyer reposted

Q. Pheevr
@q.pheevr.ca

The heartbreaking thing about this is that there’s already a proven way to invest lots of money in a knowledge machine that produces unforeseeable results that include fantastically profitable ideas (and some life-saving ones) and generally benefit society and this machine is called a university

Warren Terra@warrenterra.bsky.social

From the replies (bsky.app/profile/dasb...) here's Sam Altman doing what the quoted post described. He seems serious about it.

Mathias Sablé-Meyer
@mathiassablemeyer.bsky.social

It was really great to meet and work with @maxencepajot.bsky.social as he was trailblazing this—congrats!! 👏 Takeaways: you can predict number frequencies in books by thinking of numbers as programs, testing mental representation ideas from doi.org/10.1016/j.ti... See Maxence's great thread below👇

doi.orgRedirecting
Maxence Pajot@maxencepajot.bsky.social

Some numbers seem to show up everywhere. Think of 10, 12, 24, 36... Others—like 26 or 34—don’t get the same attention.Why? In our new paper with @standehaene.bsky.social and @mathiassablemeyer.bsky.social, we argue it's because of how the mind builds number concepts. 🧵 doi.org/10.1016/j.co...

doi.orgRedirecting

Mathias Sablé-Meyer reposted

Trends in Cognitive Sciences
@cp-trendscognsci.bsky.social

Online Now: Origins of numbers: a shared language-of-thought for arithmetic and geometry?

dlvr.itOrigins of numbers: a shared language-of-thought for arithmetic and geometry?Concepts of exact number are often thought to originate from counting and the successor function, or from a refinement of the approximate number system (ANS). We argue here for a third origin: a shared language-of-thought (LoT) for geometry and arithmetic that involves primitives of repetition, concatenation, and recursive embedding. Applied to sets, those primitives engender concepts of exact integers through recursive applications of additions and multiplications. Links between geometry and arithmetic also explain the emergence of higher-level notions (squares, primes, etc.). Under our hypothesis, understanding a number means having one or several mental expressions for it, and their minimal description length (MDL) determines how easily they can be mentally manipulated. Several historical, developmental, linguistic, and brain imaging phenomena provide preliminary support for our proposal.

Mathias Sablé-Meyer reposted

Thinking about Thinking
@thoughtchannel.bsky.social

We're bringing together leading researchers from top institutions worldwide to Split, Croatia this summer. If you're passionate about brain-inspired AI, computational neuroscience, or fundamental mathematical principles driving intelligence, this conference is the place to be!

Mathias Sablé-Meyer reposted

Altmetric
@altmetric.com

The general trend is crazy. Here is Twitter vs Bluesky volumes for the last week. On some days they are equal in volume.

This post is unavailable.