Patrick Ramos
@patrick-ramos.bsky.social
PhD student @ ISLab, Osaka University
Respectfully, I’m surprised you consider CLIP to be SSL. SSL implies that the supervision guiding the image reps comes from the images (ex. masked input, full image target). That’s not the case with CLIP, as the text supervision isn’t directly derived from the images, but is collected with them
August 24, 2025 at 2:02 PM
Respectfully, I’m surprised you consider CLIP to be SSL. SSL implies that the supervision guiding the image reps comes from the images (ex. masked input, full image target). That’s not the case with CLIP, as the text supervision isn’t directly derived from the images, but is collected with them