KVAE 2.0
Collection
KVAE 2.0 is a family of video tokenizers with a time compression ratio of 4 and spacial compression ratio of 8 and 16 β’ 2 items β’ Updated β’ 2
Artem
kabachuha
AI & ML interests
deep learning, natural language processing, text2image, text2video, computer vision, fusion power
Recent Activity
new activity about 15 hours ago
Qwen/Qwen3.6-27B:Wan when? liked a model 3 days ago
google/gemma-4-26B-A4B-itOrganizations
upvoted a collection 7 days ago
upvoted a paper 21 days ago
upvoted a collection 21 days ago
upvoted a paper about 2 months ago
upvoted 3
papers 2 months ago
Effective Reasoning Chains Reduce Intrinsic Dimensionality
Paper β’ 2602.09276 β’ Published β’ 11
LLaDA2.1: Speeding Up Text Diffusion via Token Editing
Paper β’ 2602.08676 β’ Published β’ 71
AudioSAE: Towards Understanding of Audio-Processing Models with Sparse AutoEncoders
Paper β’ 2602.05027 β’ Published β’ 63
upvoted 3
papers 3 months ago
GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization
Paper β’ 2601.05242 β’ Published β’ 230
Cross-Frame Representation Alignment for Fine-Tuning Video Diffusion Models
Paper β’ 2506.09229 β’ Published β’ 7
LTX-2: Efficient Joint Audio-Visual Foundation Model
Paper β’ 2601.03233 β’ Published β’ 177
upvoted 2
papers 4 months ago
upvoted 2
papers 5 months ago
upvoted 3
papers about 1 year ago
One-Step Residual Shifting Diffusion for Image Super-Resolution via Distillation
Paper β’ 2503.13358 β’ Published β’ 95
Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders
Paper β’ 2503.03601 β’ Published β’ 233
LLM-Microscope: Uncovering the Hidden Role of Punctuation in Context Memory of Transformers
Paper β’ 2502.15007 β’ Published β’ 175
upvoted 3
papers almost 2 years ago
Grass: Compute Efficient Low-Memory LLM Training with Structured Sparse Gradients
Paper β’ 2406.17660 β’ Published β’ 5
An Image is Worth 32 Tokens for Reconstruction and Generation
Paper β’ 2406.07550 β’ Published β’ 60
Make It Count: Text-to-Image Generation with an Accurate Number of Objects
Paper β’ 2406.10210 β’ Published β’ 78