Open to Work

18 27 3

Taki WU

taki555

https://wutaiqiang.github.io/

AI & ML interests

None yet

Recent Activity

commentedon a paper 22 days ago

OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond

upvoted a paper 22 days ago

OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond

authored a paper about 2 months ago

Attention Sink in Transformers: A Survey on Utilization, Interpretation, and Mitigation

View all activity

Organizations

Collections 3

View 3 collections

Papers 20

models 15

datasets 1

taki555/DeepScaleR-Easy

Viewer • Updated Feb 25 • 19.6k • 19 • 1

Taki WU

AI & ML interests

Recent Activity

Organizations

Collections 3

The Art of Efficient Reasoning: Data, Reward, and Optimization

taki555/DeepScaleR-Easy

taki555/Qwen3-0.6B-Art

taki555/Qwen3-1.7B-Art

taki555/Qwen3-0.6B-Timber

Timber: Training-free Instruct Model Refining with Base via Effective Rank

The Art of Efficient Reasoning: Data, Reward, and Optimization

taki555/DeepScaleR-Easy

taki555/Qwen3-0.6B-Art

taki555/Qwen3-1.7B-Art

taki555/Qwen3-0.6B-Timber

Timber: Training-free Instruct Model Refining with Base via Effective Rank

Papers 20

models 15

taki555/Qwen3-30B-A3B-Thinking-2507-Art

taki555/Qwen3-30B-A3B-Instruct-2507-Art

taki555/Qwen3-4B-Instruct-2507-Art

taki555/Qwen3-1.7B-Art

taki555/Qwen3-0.6B-Art

taki555/Qwen3-4B-Thinking-2507-Art

taki555/Qwen3-0.6B-Timber

taki555/Qwen3-14B-Shadow-FT-BAAI-2k

taki555/Qwen3-8B-Shadow-FT-BAAI-2k

taki555/Qwen3-4B-Shadow-FT-BAAI-2k

datasets 1

taki555/DeepScaleR-Easy

Taki WU

AI & ML interests

Recent Activity

Organizations

Collections 3

Papers 20

models 15 Sort: Recently updated

datasets 1

models 15