Xiangtai Li

LXT

20 70 27

https://lxtgh.github.io/

AI & ML interests

Computer Vision, Multi-Modal Understanding, Generative AI

Recent Activity

upvoted a paper 16 days ago

On-Policy Self-Distillation in Diffusion Models

upvoted a paper 23 days ago

StateM: Reaching 95.3% Raw Accuracy, or a \$15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling

upvoted a paper about 1 month ago

Scaling Properties of Text Conditioning in Visual Generation

View all activity

Organizations

liked 3 datasets 3 months ago

liked a model 3 months ago

MSALab/LoomVideo

Updated Jun 6 • 8

liked a dataset 4 months ago

marinero4972/VideoZeroBench

Preview • Updated May 6 • 185 • 5

liked a model 7 months ago

inclusionAI/LLaDA2.1-mini

Text Generation • 16B • Updated Apr 13 • 110k • 125

liked a dataset 7 months ago

WeiChow/CrispEdit-2M

Viewer • Updated Dec 29, 2025 • 2.29M • 16.5k • 17

liked a model 9 months ago

WeiChow/EditMGT

Image-to-Image • 1B • Updated Dec 20, 2025 • 11

liked 4 models about 1 year ago

ByteDance/Sa2VA-26B

Image-Text-to-Text • 26B • Updated Sep 8, 2025 • 76 • 32

ByteDance/Sa2VA-1B

Image-Text-to-Text • 1B • Updated Sep 8, 2025 • 381 • 31

ByteDance/Sa2VA-8B

Image-Text-to-Text • 8B • Updated Sep 8, 2025 • 6.26k • 64

black-forest-labs/FLUX.1-Kontext-dev

Image-to-Image • 12B • Updated Jan 1 • 305k • • 2.8k

liked a model over 1 year ago

answerdotai/ModernBERT-base

Fill-Mask • 0.1B • Updated Jan 15, 2025 • 4.73M • 1.09k

liked a Space over 1 year ago

Sa2VA Simple Demo

🐨

Dense Grounded Understanding of Images and Videos

liked a model over 1 year ago

ByteDance/Sa2VA-4B

Image-Text-to-Text • 4B • Updated Sep 8, 2025 • 2.32k • 99

liked a dataset over 1 year ago

zhangtao-whu/OMG-LLaVA

Updated Jul 3, 2024 • 584 • 3

liked a dataset almost 2 years ago

jianzongwu/MangaZero

Viewer • Updated Dec 11, 2024 • 32.7k • 101 • 36

liked a model almost 2 years ago

Collov-Labs/Monetico

Text-to-Image • 1.0B • Updated Oct 28, 2024 • 11 • 66

liked a Space almost 2 years ago

Meissonic Flow

🚀

Generate images from text prompts