Xin Jin

Latest

• ExpertPlex: A High-Goodput Disaggregated Serving System for MoE LLMs with Adaptive Persistent Kernels
• ReLibra: Routing-Replay-Guided Load Balancing for MoE Training in Reinforcement Learning
• FastServe: Iteration-Level Preemptive Scheduling for Large Language Model Inference
• Heddle: A Distributed Orchestration System for Agentic RL Rollout
• Epiphron: Resource-Efficient Distributed Key-Value Storage
• TokenLake: A Unified Segment-level Prefix Cache Pool for Fine-grained Elastic Long-Context LLM Serving
• Optimizing RLHF Training for Large Language Models with Stage Fusion
• LoongServe: Efficiently Serving Long-Context Large Language Models with Elastic Sequence Parallelism
• dLoRA: Dynamically Orchestrating Requests and Adapters for LoRA LLM Serving
• A Survey of Resource-efficient LLM and Multimodal Foundation Models
• XRON: A Hybrid Elastic Cloud Overlay Network for Video Conferencing at Planetary Scale
• Transparent GPU Sharing in Container Clouds for Deep Learning Workloads