Stéphane d'Ascoli (@stephanedascoli) / X

Stéphane d'Ascoli

137 posts

Stéphane d'Ascoli

@stephanedascoli

Research Scientist @AIatMeta working on brain decoding. Prev @EPFL, @ENS_ULM, @NASA.

Paris, France

Joined November 2018

Pinned
Stéphane d'Ascoli
@stephanedascoli
Mar 26
🚨 We're very happy to introduce TRIBE v2: a foundation model of the brain's responses to sight, sound & language. 📄 Paper: ai.meta.com/research/publi… ▶️ Demo: aidemos.atmeta.com/tribev2/ 💻 Code: github.com/facebookresear… 🤗 Model: huggingface.co/facebook/tribe…
00:00
120K
Stéphane d'Ascoli
@stephanedascoli
Jan 12, 2023
Today I met machine learning’s number one enemy #spuriouscorrelations
81K
Stéphane d'Ascoli
@stephanedascoli
Sep 25, 2023
Think Transfomers are terrible at logical reasoning? Think again 💥 In this collaboration with Samy Bengio, @jsusskin (Apple) & Emmanuel Abbé (EPFL), we show that when trained with Boolean inputs and symbolic outputs, they become very powerful 🧠 arxiv.org/pdf/2309.12207… 🧵⤵️
74K
Stéphane d'Ascoli
@stephanedascoli
Sep 21, 2023
Thrilled to announce that I will be joining @MetaAI next month as a Research Scientist 😍 I will be working in the Brain & AI team on decoding language from neural activity, to hopefully help those which have difficulties to speak or type. Learn more here:
ai.meta.com
Using AI to decode speech from brain activity
Decoding speech from brain activity has mostly benefited from invasive approaches. New research from FAIR shows AI could instead make use of noninvasive brain scans.
51K
Stéphane d'Ascoli
@stephanedascoli
Oct 10, 2023
🚨 ODEFormer is on Arxiv! arxiv.org/pdf/2310.05573… We show that Transformers can recover the differential equations governing dynamical systems from noisy & irregularly sampled trajectories. Very fun collaboration with @SorenBecker, @TrackingPlumes, @pschwllr & @k__niki! 🧵⤵️
41K
Stéphane d'Ascoli
@stephanedascoli
May 11, 2021
CNNs are more sample-efficient, ViTs are more powerful. Can we get the best of both worlds? Check out our paper accepted at @icmlconf arxiv.org/pdf/2103.10697… Thanks to my collaborators @HugoTouvron @leavittron @arimorcos @leventsagun @GiulioBiroli 🧵⤵️
Stéphane d'Ascoli
@stephanedascoli
Feb 26, 2023
After a couple great years at @MetaAI and @ENS_ULM, I will be starting as @AI4ScienceEPFL fellow next month 😍 Can’t wait to leverage modern AI tools with biologists, neuroscientists, chemists and physicists 🧬🧠🧪🔭 If you work at @EPFL and want to meet up, please reach out !
16K
Stéphane d'Ascoli
@stephanedascoli
Jan 14, 2022
1,2,3,5,8,13… What is the next term ? This kind of question is typical in IQ tests, but has received little attention in AI. We had great fun training Transformers to tackle this problem, check out our paper arxiv.org/abs/2201.04600 and our online demo : bit.ly/3niE5FS
Guillaume Lample @ NeurIPS 2024
@GuillaumeLample
Jan 14, 2022
Deep Symbolic Regression for Recurrent Sequences -- arxiv.org/abs/2201.04600 We show that transformers are great at predicting symbolic functions from values, and can predict the recurrence relation of sequences better than Mathematica. You can try it here: bit.ly/3niE5FS
Stéphane d'Ascoli
@stephanedascoli
Feb 13, 2022
New preprint : arxiv.org/abs/2202.04509. When and how should you decay your learning rate ? We give some theoretical insights on this crucial question in our latest work with @MariaRefinetti and @GiulioBiroli. (1/3)
Stéphane d'Ascoli
@stephanedascoli
Sep 25, 2023
Replying to @stephanedascoli
We hope this work can be applied to other fields in science and spark more research on symbolic reasoning in LLMs. We release our code & models publicly and provide a pip package & interactive Colab demo! github.com/sdascoli/boolf… A few attention maps for the pleasure of the eye:
1.5K
Stéphane d'Ascoli
@stephanedascoli
Sep 25, 2023
Replying to @stephanedascoli
The so-called "Boolformer" takes as input a set of N (x,y) pairs in {0,1}^D x {0,1}, and tries to predict a Boolean formula which approximates these observations. Here are two very simple examples: addition and multiplication of 2-bit numbers.
6K
Stéphane d'Ascoli
@stephanedascoli
Jan 29, 2022
Great video by @ykilcher on our symbolic regression paper with @pa_kamienny @GuillaumeLample @f_charton. Watch until the end to discover his musical skills 😅
Yannic Kilcher 🇸🇨
@ykilcher
Jan 29, 2022
📜Paper Video Time!📜Today I'm talking to Stéphane d'Ascoli (@stephanedascoli) about Deep Symbolic Regression for Recurrent Sequences. This model is given a sequence of numbers, like 1, 2, 3, 5, 8 and it figures out the *rule behind* the sequence. Insane🤯 youtu.be/1HEdXwEYrGM
Stéphane d'Ascoli
@stephanedascoli
Jun 26, 2021
Cet après-midi, mon ami Arthur et moi avions l’honneur d’être invités par Étienne Klein sur France Culture dans la Conversation Scientifique. Une émission consacrée pleinement à notre sujet préféré, l’espace-temps et sa courbure : a (ré)-écouter ici !
L’espace-temps est courbe : qu’est-ce à dire ?
From radiofrance.fr
Stéphane d'Ascoli
@stephanedascoli
Sep 25, 2023
Replying to @stephanedascoli
We apply the Boolformer to a set of classification tasks from PMLB, ranging from predicting chess moves to diagnosing horse colic. Our model achieves similar performance to classic ML methods, while outputting interpretable Boolean formulas!
1.6K