Xinyi Wan's picture

4 12 6

Xinyi Wan

ufotalent

·

AI & ML interests

ML Sys

Recent Activity

upvoted a paper 12 days ago

Rethinking the Trust Region in LLM Reinforcement Learning

authored a paper 20 days ago

Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs

authored a paper 20 days ago

ZeCO: Zero Communication Overhead Sequence Parallelism for Linear Attention

View all activity

Organizations

authored 3 papers 20 days ago

Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs

Paper • 2502.12982 • Published Feb 18, 2025 • 19

ZeCO: Zero Communication Overhead Sequence Parallelism for Linear Attention

Paper • 2507.01004 • Published Jul 1, 2025 • 10

Revisiting Parameter Server in LLM Post-Training

Paper • 2601.19362 • Published 21 days ago • 7

submitted a paper to Daily Papers 20 days ago

Revisiting Parameter Server in LLM Post-Training

Paper • 2601.19362 • Published 21 days ago • 7

authored a paper 11 months ago

PipeOffload: Improving Scalability of Pipeline Parallelism with Memory Optimization

Paper • 2503.01328 • Published Mar 3, 2025 • 16

authored a paper over 1 year ago

Pipeline Parallelism with Controllable Memory

Paper • 2405.15362 • Published May 24, 2024 • 3