Pu Fanyi

pufanyi

7 78 148

https://pufanyi.github.io

AI & ML interests

Recent Activity

updated a Space about 16 hours ago

pufanyi/ViZDoom-Agentic-Demo

updated a dataset about 16 hours ago

pufanyi/ViZDoom

updated a dataset about 17 hours ago

pufanyi/ViZDoom-Agentic-Rollouts

View all activity

Organizations

updated a Space about 16 hours ago

ViZDoom Agentic Rollouts

🎮

Explore ViZDoom agent rollouts with video and metrics

updated a dataset about 16 hours ago

pufanyi/ViZDoom

Viewer • Updated about 16 hours ago • 120 • 233

updated a dataset about 17 hours ago

pufanyi/ViZDoom-Agentic-Rollouts

Viewer • Updated about 17 hours ago • 240 • 4

published a dataset about 17 hours ago

pufanyi/ViZDoom-Agentic-Rollouts

Viewer • Updated about 17 hours ago • 240 • 4

upvoted a paper 3 days ago

Mage-VL: An Efficient Codec-Native Streaming Multimodal Foundation Model

Paper • 2607.24904 • Published 5 days ago • 27

liked 2 models 5 days ago

microsoft/Mage-VL

Image-Text-to-Text • 5B • Updated 3 days ago • 10.5k • 160

microsoft/Mage-ViT

Image Feature Extraction • 0.3B • Updated 5 days ago • 6.39k • 26

upvoted an article 5 days ago

Article

Kimi K3 Model Overview: 2.8T Parameters, MXFP4 Quantization, and What the Open Weights Mean for the Community

ResterChed

•

15 days ago

• 187

updated a dataset 9 days ago

pufanyi/vbvrpro_output-data

Preview • Updated 9 days ago • 6.97k

published a dataset 9 days ago

pufanyi/vbvrpro_output-data

Preview • Updated 9 days ago • 6.97k

updated a Space 9 days ago

VBVR-Pro Output Explorer

🧭

Task-level and video-level VBVR-Pro training result explorer

published a Space 9 days ago

VBVR-Pro Output Explorer

🧭

Task-level and video-level VBVR-Pro training result explorer

authored a paper 10 days ago

Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing

Paper • 2607.19064 • Published 11 days ago • 75

upvoted a paper 10 days ago

Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing

Paper • 2607.19064 • Published 11 days ago • 75

upvoted a paper 14 days ago

VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding

Paper • 2607.14935 • Published 16 days ago • 170

liked a model 16 days ago

thinkingmachines/Inkling

Image-Text-to-Text • 952B • Updated 9 days ago • 59.1k • • 1.67k

published a Space 19 days ago

ViZDoom Agentic Rollouts

🎮

Explore ViZDoom agent rollouts with video and metrics

upvoted a paper 19 days ago

Video Generation Models are General-Purpose Vision Learners

Paper • 2607.09024 • Published 22 days ago • 86

published a dataset 20 days ago

pufanyi/ViZDoom

Viewer • Updated about 16 hours ago • 120 • 233

upvoted a paper 24 days ago

Vision as Unified Multimodal Generation

Paper • 2607.06560 • Published 25 days ago • 48

Pu Fanyi

AI & ML interests

Recent Activity

Organizations

pufanyi's activity

ViZDoom Agentic Rollouts

Kimi K3 Model Overview: 2.8T Parameters, MXFP4 Quantization, and What the Open Weights Mean for the Community

VBVR-Pro Output Explorer

VBVR-Pro Output Explorer

ViZDoom Agentic Rollouts