A multimodal LLM for the non-invasive decoding of spoken text from brain recordings

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.