diffusion forcing: next-token prediction meets full-sequence diffusion
Published 4 months ago • 289 plays • Length 11:02Download video MP4
Download video MP3
Similar videos
-
![[qa] diffusion forcing: next-token prediction meets full-sequence diffusion](https://i.ytimg.com/vi/9G8c5i-Ahis/mqdefault.jpg)
8:35
[qa] diffusion forcing: next-token prediction meets full-sequence diffusion
-
1:05:12
diffusion forcing: next-token prediction meets full-sequence diffusion | boyuan chen
-
51:50
skimming hella ai paper abstracts - nov 5, 2024
-
5:01
prdp: proximal reward difference predictionfor large-scale reward finetuning of diffusion models
-
![[cvpr 2024] exploiting diffusion prior for generalizable dense prediction](https://i.ytimg.com/vi/z4W0LmI-4C8/mqdefault.jpg)
5:21
[cvpr 2024] exploiting diffusion prior for generalizable dense prediction
-
0:10
efficient diffusion models with deep compression autoencoder
-
20:18
why does diffusion work better than auto-regression?
-
2:06:38
this is why deep learning is really weird.
-
30:54
diffusion models from scratch in pytorch
-
![[short] deconstructing denoising diffusion models for self-supervised learning](https://i.ytimg.com/vi/iaxlQ1siRvE/mqdefault.jpg)
2:44
[short] deconstructing denoising diffusion models for self-supervised learning
-
38:11
diffusion models from scratch | score-based generative models explained | math explained
-
![[qa] active-dormant attention heads: mechanistically demystifying extreme-token phenomena in llms](https://i.ytimg.com/vi/IyCbGZc6oD0/mqdefault.jpg)
8:01
[qa] active-dormant attention heads: mechanistically demystifying extreme-token phenomena in llms
-
1:08
reorientdiff: diffusion model based reorientation for object manipulation
-
![[qa] adapting language models via token translation](https://i.ytimg.com/vi/7DM2PyocYXQ/mqdefault.jpg)
8:29
[qa] adapting language models via token translation
-
22:01
transfusion: predict the next token and diffuse images with one multi-modal model - arxi
-
15:23
multi-token prediction (forget next token llm?)
-
9:41
adapting language models via token translation
-
0:59
paper walkthrough: why use attention as explanation (https://arxiv.org/abs/2010.05607) #llm #ai
-
2:47
can we use diffusion probabilistic models for 3d motion prediction?
-
14:32
evaluating diffusion models with pickscore
-
![[short] pass: parallel speculative sampling](https://i.ytimg.com/vi/UsqvA8-DEpw/mqdefault.jpg)
2:29
[short] pass: parallel speculative sampling
-
![sigir 2024 t2.2 [fp] denoising diffusion recommender model](https://i.ytimg.com/vi/j46Tuu9W2tg/mqdefault.jpg)
14:58
sigir 2024 t2.2 [fp] denoising diffusion recommender model