Birdwatch Note Rating
2023-03-03 19:34:47 UTC - HELPFUL
Rated by Participant: 45FEE49BCFCE4A1A17CA18A1FA73059C6C807EF15AF136F0EBD9FC7AFC9A6700
Participant Details
Original Note:
The model used text prompts, but the prompts were decoded from fMRI signals. The researchers didn't generate them. Also, the tweet clarifies that the paper is not peer-reviewed. "We decoded latent text representations c from fMRI signals within higher (ventral) visual cortex." https://www.biorxiv.org/content/10.1101/2022.11.18.517004v2.full
All Note Details