Wei Huang's picture

Wei Huang

AaronHuangWei

·

AI & ML interests

Deep Learning; Efficient AI; Quantization; LLMs

Recent Activity

authored a paper about 6 hours ago

Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusion

authored a paper about 6 hours ago

SpatialEdit: Benchmarking Fine-Grained Image Spatial Editing

authored a paper about 6 hours ago

TriAttention: Efficient Long Reasoning with Trigonometric KV Compression

View all activity

Organizations

Collections 1

Papers 17

arxiv:2604.04921

arxiv:2604.04911

arxiv:2603.13405

arxiv:2512.20557

models 51

AaronHuangWei/Wan2.1-T2V-14B-NVFP4Fake_layer_4o6

Text-to-Video • Updated Mar 4 • 94

AaronHuangWei/Wan2.1-T2V-14B-NVFP4Fake_layer_4o6_search_n100_s0.001_p2.4

Text-to-Video • Updated Mar 3 • 99

AaronHuangWei/Wan2.2-TI2V-5B-NVFP4Fake_layer_4o6_search_n100_s0.001_p2.4

Text-to-Video • Updated Mar 3 • 10

AaronHuangWei/Wan2.2-TI2V-5B-NVFP4Fake_layer_4o6

Text-to-Video • Updated Mar 3 • 8

AaronHuangWei/Wan2.2-TI2V-5B-NVFP4FakeQuant_ignorelayer

Text-to-Video • Updated Feb 15 • 10

AaronHuangWei/Wan2.1-T2V-1.3B-NVFP4RealQuant_weightnoly

Text-to-Video • Updated Jan 15 • 33

AaronHuangWei/Wan2.1-T2V-14B-NVFP4RealQuant_weightnoly

Text-to-Video • Updated Jan 15 • 56

AaronHuangWei/Llama-3.1-Nemotron-Nano-8B-v1-NVFP4A16-GPTQ

5B • Updated Jan 15 • 4

AaronHuangWei/Wan2.1-T2V-14B-NVFP4FakeQuant

Text-to-Video • Updated Jan 13 • 106

AaronHuangWei/Wan2.1-T2V-14B-INT8FakeQuant_pertensor

Text-to-Video • Updated Jan 4 • 100

datasets 1

AaronHuangWei/Inf-OmniStream-Train

Preview • Updated Feb 25 • 30