Efficient Long-context Language Model Training by Core Attention Disaggregation
Paper • 2510.18121 • Published • 124
Junda Chen
GindaChen
AI & ML interests
Reasoning System, Efficient LLM Serving System
Recent Activity
upvoted an article 9 days ago
EMO: Pretraining mixture of experts for emergent modularity updated a dataset about 1 month ago
GindaChen/glass-substrate-traces published a dataset about 1 month ago
GindaChen/glass-substrate-tracesOrganizations
authored a
paper 7 months ago
authored 3
papers over 1 year ago
DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving
Paper • 2401.09670 • Published • 2
Mnemosyne: Parallelization Strategies for Efficiently Serving Multi-Million Context Length LLM Inference Requests Without Approximations
Paper • 2409.17264 • Published
Efficiently Serving LLM Reasoning Programs with Certaindex
Paper • 2412.20993 • Published • 36