Accurate Decoding of Natural Sentences from<br>Non-Invasive Brain Recordings | Research - AI at Meta
Products<br>AI Research<br>Resources<br>About<br>Get Llama<br>Try Meta AI
RESPONSIBLE AI
Accurate Decoding of Natural Sentences from<br>Non-Invasive Brain Recordings
June 29, 2026
Abstract<br>Restoring communication for people who have lost the ability to speak or move after a brain injury is a major challenge. While intracranial implants now enable high-performing brain-computer-interfaces, non-invasive alternatives are still lagging behind. Here, we present Brain2Qwerty v2, a model that can decode the production of natural sentences solely from real-time magnetoencephalography (MEG) recordings. By collecting 22,000 sentences typed by nine subjects, each recorded for 10 hours, our<br>model leverages character, word and sentence-level representations to achieve an average word error rate (WER) of 39%. For our best participant, the model accurately decodes half of the sentences<br>with one word error or less. Critically, decoding accuracy log-linearly improves with data volume, suggesting that the performance gap with intracranial approaches could be partially bridged through<br>data scaling. We show that AI enables this performance in three main ways: the substitution of hand-crafted pipelines for event detection with deep learning, the finetuning of large language models to extract semantic representations, and the deployment of AI agents to iteratively refine our decoding pipeline via automated code development. Together, these results show that non-invasive brain-to-text decoding starts to operate at a level of accuracy previously thought exclusive to surgical implants, opening a path toward safe and efficient brain-computer-interfaces.
Download the Paper
AUTHORS
Written by<br>Mingfang (Lucy) Zhang *<br>Jarod Levy *<br>Cédric Rommel<br>Jérémy Rapin<br>Corentin Bel<br>Julie Bonnaire<br>Daniel Nieto<br>Pierre Bourdillon<br>Svetlana Pinet<br>Stéphane d'Ascoli<br>Thomas Moreau<br>Jean Remi King
Publisher<br>arXiv
Research Topics<br>Speech & Audio
Related Publications<br>February 13, 2026
RESPONSIBLE AI
FERRET: Framework for Expansion Reliant Red Teaming
Ninareh Mehrabi, Vítor Albiero, Maya Pavlova, Joanna Bitton
February 13, 2026
Read the Paper
December 26, 2025
REINFORCEMENT LEARNING<br>NLP
Safety Alignment of LMs via Non-cooperative Games
Anselm Paulus, Ilia Kulikov, Brandon Amos, Remi Munos, Ivan Evtimov, Kamalika Chaudhuri, Arman Zharmagambetov
December 26, 2025
Read the Paper
September 24, 2025
RESEARCH<br>NLP
Code World Model Preparedness Report
Daniel Song, Peter Ney, Cristina Menghini, Faizan Ahmad, Aidan Boyd, Nathaniel Li, Ziwen Han, Jean-Christophe Testud, Saisuke Okabayashi, Maeve Ryan, Jinpeng Miao, Hamza Kwisaba, Felix Binder, Spencer Whitman, Jim Gust, Esteban Arcaute, Dhaval Kapil, Jacob Kahn, Ayaz Minhas, Tristan Goodman, Lauren Deason, Alexander Vaughan, Shengjia Zhao, Summer Yue
September 24, 2025
Read the Paper
June 13, 2025
FAIRNESS<br>INTEGRITY
Measuring multi-calibration
Ido Guy, Daniel Haimovich, Fridolin Linder, Nastaran Okati, Lorenzo Perini, Niek Tax, Mark Tygert
June 13, 2025
Read the Paper
See All Papers
Help Us Pioneer The Future of AI
We share our open source frameworks, tools, libraries, and models for everything from research exploration to large-scale production deployment.
Join our Team
Our approach<br>About AI at Meta<br>People<br>Careers
Research<br>Infrastructure<br>Resources<br>Demos
Meta AI<br>Explore Meta AI<br>Get Meta AI<br>AI Studio
Latest news<br>Blog<br>Newsletter
Foundational models
Llama
Our approach
Our approachAbout AI at MetaPeopleCareers
Research
ResearchInfrastructureResourcesDemos
Meta AI
Meta AIExplore Meta AIGet Meta AIAI Studio
Latest news
Latest newsBlogNewsletter
Foundational models
Llama
Privacy Policy<br>Terms<br>Cookies
Meta © 2026