Pretraining on 50,000 Hours of Unlabeled Brain Data
neuralink.com
Neuralink has collected over 50,000 hours of unlabeled neural data since its clinical trials began to improve its machine-learning decoders. The article says it used more than 50,000 hours of freeform neural data. Decoders are expected to need less calibration and become more accurate and longer-lasting for individual users.
Why it matters
Neuralink has collected over 50,000 hours of unlabeled, freeform neural data since its clinical trials began, using it to improve its machine-learning decoders. The issue notes decoders are user-specific and require calibration before use, but the training aims to make them more accurate and long-lived.