Felix Koulischer
@felixkoulischer.bsky.social
PhD working on Diffusion Models and all their applications, from images to proteins! Passions include theoretical physics, birds and orgiamis
The DB track is a very different story, the average score is of 4.25/6 which implies that loads of qualitative papers got rejected simply to maintain the 30% acceptance rate and that's much more problematic...
I believe that good papers should get in, regardless of how many there are...
I believe that good papers should get in, regardless of how many there are...
October 1, 2025 at 1:27 PM
The DB track is a very different story, the average score is of 4.25/6 which implies that loads of qualitative papers got rejected simply to maintain the 30% acceptance rate and that's much more problematic...
I believe that good papers should get in, regardless of how many there are...
I believe that good papers should get in, regardless of how many there are...
July 31, 2025 at 1:48 PM
#1 We see guidance as an error correction mechanism, by choosing different error models one obtains different guidance schemes!
#2: We see the diffusion process from a control theory perspective, classical guidance schemes such as CFG, work as open loop controller, FBG works as a closed loop!
#2: We see the diffusion process from a control theory perspective, classical guidance schemes such as CFG, work as open loop controller, FBG works as a closed loop!
July 31, 2025 at 1:46 PM
#1 We see guidance as an error correction mechanism, by choosing different error models one obtains different guidance schemes!
#2: We see the diffusion process from a control theory perspective, classical guidance schemes such as CFG, work as open loop controller, FBG works as a closed loop!
#2: We see the diffusion process from a control theory perspective, classical guidance schemes such as CFG, work as open loop controller, FBG works as a closed loop!