Joséphine Raugel

@JRaugel

PhD student @ Meta & ENS, on Large models and Brains processing language and vision together. jraugel.github.io

Paris, France Joined October 2021

201 Following 667 Followers 50 Posts

Latest posts 5

The EEG/EMG Foundation Challenge 2026 starts in 4 days 💫
🔗 neural-interfaces26.github.io/index.html

Four tracks 🧠 :
🖼️ EEG-to-IMG
➗ BCI decoding
😴 Predicting sleep onset
✋ EMG-to-Pose

Let's push the frontiers of neural decoding further 🔥

neural-interfaces26.github.io EEG/EMG Foundation Challenge 2026 | Brain and Body Workshop at NeurIPS EEG/EMG Foundation Challenge 2026. Sealed evaluation November 5-21 for Tracks 01, 02 and 04. Track 03 opens later. $20,000 in prizes.

Joséphine Raugel @JRaugel Jun 29

Brain2Qwerty V2 is out 💫

Jarod Levy @JarodLevy Jun 29

🧠⌨️ Decode language from brain activity without surgery. 🧠⌨️

Brain2Qwerty V1 is officially published in Nature Neuroscience. Today, we're releasing Brain2Qwerty V2.
We achieve unprecedented performance for a non-invasive MEG setup.

Details below 🧵👇

View quoted post on X

Joséphine Raugel @JRaugel May 28

1/ We’re so glad to share this new study 💫
Does the brain learn like a Deep Net? 🧠⚙️
- 📄Misalignment Between Backpropagation and the Hierarchy of Brain Responses to Images
- 🔗arxiv.org/abs/2605.28693
Thread below 🧵

Joséphine Raugel @JRaugel May 7

It's time to flex your brain foundation model: NeuralBench is out.
36 EEG tasks, 94 datasets, 9,500 subjects.
Congratulations @hubertjbanville for leading this tremendous work

Hubert Banville @hubertjbanville May 6

🧠 Introducing NeuralBench: a unified, open-source framework to benchmark NeuroAI models.

v1.0: 36 EEG tasks, 94 datasets, task-specific + foundation models. MEG/fMRI ready.

MIT-licensed, FAIR's Brain & AI @AIatMeta.

Code: github.com/facebookresear…
Paper: ai.meta.com/research/publi…

View quoted post on X

Joséphine Raugel @JRaugel Apr 24

We’re thrilled to share DinoBrain at @iclr_conf:
Can AI help us understand how the brain learns to see? 🧠
📍Poster P3 #1610 | 3:15pm
- DINOv3’s training mirrors some striking aspects of brain development, especially when trained on human-centric data.
- Reproduced on 8 models.

Jean-Rémi King @JeanRemiKing Sep 1, 2025

Can AI help understand how the brain learns to see the world?

Our latest study, led by @JRaugel from FAIR at @AIatMeta and @ENS_ULM, is now out!
📄 arxiv.org/pdf/2508.18226

🧵 A thread:

View quoted post on X