AI & ML interests
None defined yet.
Recent Activity
Papers
MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities
(1D) Ordered Tokens Enable Efficient Test-Time Search
spaces 8
Running on Zero
Agents
1
Video-4M
🎬
Generate videos from captions, frames, or other modalities
Running on Zero
Agents
5
MODUS
🎨
MODUS any-to-any multimodal demo (16 aligned modalities)
Running on Zero
Agents
4
Search over Tokens
🔎
Test-time search over ordered visual tokens.
Running on Zero
Agents
3
VideoFlexTok
🎞
VideoFlexTok: flexible-length coarse-to-fine video tokenizer
Running on Zero
Agents
Featured
203
4M Demo
⚡
4M: Massively Multimodal Masked Modeling
Running on Zero
Agents
17
FlexTok
🖼
FlexTok flexible sequence length autoencoding demo
models 55
EPFL-VILAB/MODUS-Hunyuan
Image-to-Image • 83B • Updated • 39
EPFL-VILAB/Video-4M-tokenizers
Updated • 1
EPFL-VILAB/Video-4M-models
Image-to-Video • Updated • 1
EPFL-VILAB/FlexAR-3B-T2I
Text-to-Image • 3B • Updated • 53
EPFL-VILAB/GridAR-3B-T2I
Text-to-Image • 3B • Updated • 33
EPFL-VILAB/FlexAR-1B-T2I
Text-to-Image • 1B • Updated • 38
EPFL-VILAB/FlexAR-382M-T2I
Text-to-Image • 0.5B • Updated • 21
EPFL-VILAB/FlexAR-113M-T2I
Text-to-Image • 0.2B • Updated • 24 • 1
EPFL-VILAB/MODUS
Any-to-Any • 15B • Updated • 3
EPFL-VILAB/videoflextok_d18_d18_k600
Other • 1B • Updated • 1.44k