lebellig
@lebellig.bsky.social
Ph.D. student on generative models and domain adaptation for Earth observation 🛰
Previously intern @SonyCSL, @Ircam, @Inria
🌎 Personal website: https://lebellig.github.io/
Previously intern @SonyCSL, @Ircam, @Inria
🌎 Personal website: https://lebellig.github.io/
Great article! But can the preference score go up to 2? You know, because 1 just isn’t aesthetic enough.
October 31, 2025 at 9:37 AM
Great article! But can the preference score go up to 2? You know, because 1 just isn’t aesthetic enough.
Only because of the font size and the large figures!
Tbh, I think this is a super valuable resource for anyone with even a small amount of background in diffusion models or generative models who’s interested in diffusion model research (and wants to avoid the early papers' notations)
Tbh, I think this is a super valuable resource for anyone with even a small amount of background in diffusion models or generative models who’s interested in diffusion model research (and wants to avoid the early papers' notations)
October 28, 2025 at 2:13 PM
Only because of the font size and the large figures!
Tbh, I think this is a super valuable resource for anyone with even a small amount of background in diffusion models or generative models who’s interested in diffusion model research (and wants to avoid the early papers' notations)
Tbh, I think this is a super valuable resource for anyone with even a small amount of background in diffusion models or generative models who’s interested in diffusion model research (and wants to avoid the early papers' notations)
Reposted by lebellig
Strong afternoon session: Ségolène Martin on how to go from flow matching to denoisers (and hopefully come back?) and Claire Boyer on how learning rate and working in latent spaces affect diffusion models
October 24, 2025 at 3:04 PM
Strong afternoon session: Ségolène Martin on how to go from flow matching to denoisers (and hopefully come back?) and Claire Boyer on how learning rate and working in latent spaces affect diffusion models
Very nice challenge! As a student, I would have loved it. Has anybody tried to cheat and involve the classifier in the training to target its modes (far from the decision boundaries)?
October 23, 2025 at 9:49 PM
Very nice challenge! As a student, I would have loved it. Has anybody tried to cheat and involve the classifier in the training to target its modes (far from the decision boundaries)?
two days in, and I’m still waiting for an H100 :)
October 17, 2025 at 11:58 AM
two days in, and I’m still waiting for an H100 :)