Joséphine Raugel

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

[Joséphine Raugel @JRaugel]

@JeanRemiKing @valentinwyart @maxseitzer @MarcSzafraniec @huyvvo @oriane_simeoni @monsieurlabatut @p_bojanowski @martin_hebart @Emsquem @AIatMeta @ENS_ULM 15/ Finally thanks to @ylecun and @pascalefung for the insightful discussions on this topic

[Joséphine Raugel @JRaugel]

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
- 🔗https://t.co/yrb4otBEYk
Thread below 🧵

[Joséphine Raugel @JRaugel]

@JeanRemiKing @valentinwyart @maxseitzer @MarcSzafraniec @huyvvo @oriane_simeoni @monsieurlabatut @p_bojanowski @martin_hebart @Emsquem @AIatMeta @ENS_ULM 14/ and 🙏to all the open source and open data #NeuroAI communities for making this possible!

[Joséphine Raugel @JRaugel]

@JohnYue122333 Indeed, testing whether predictive-coding signals align better with brain activity and brain organization would be an interesting next step

[Joséphine Raugel @JRaugel]

@ZebinWilson Good question. We do find that low-level brain areas are best predicted by gradients of deepest model layers. This doesn’t imply stronger error-correction signal there, rather that we don’t observe the expected hierarchy of bio-plausible backprop (early regions↔early gradients).

[Jarod Levy @JarodLevy]

💃 DANCE 🕺: Detect and Classify Events in EEG

Today, we’re releasing DANCE, an end-to-end model that detects and classifies events in EEG in one pass.

📄 Paper: https://t.co/0TLhzxXHK0
💻 Code: https://t.co/NEfQ7wxXLJ

🧵🧵🧵 More details below 🧵🧵🧵