EuroSys 2026

Meta Info

Homepage: https://2026.eurosys.org

Acceptance Rate

• Spring: 19.6% (= 79 / 404)

Papers

LLMs

• LLM Inference

  • Taming Latency-Memory Trade-Off in MoE-Based LLM Serving via Fine-Grained Expert Offloading [Paper] [arXiv]

    • Stevens Institute of Technology & Waterloo & Rutgers

  • FlexPipe: Adapting Dynamic LLM Serving Through Inflight Pipeline Refactoring in Fragmented Serverless Clusters [Paper] [arXiv]

    • SIAT, CAS

• LoRA Fine-Tuning

  • LoRAFusion: Efficient LoRA Fine-Tuning for LLMs [Paper] [arXiv]

    • UofT & Vector Institute & NVIDIA

Diffusion Models

• FlashPS: Efficient Generative Image Editing with Mask-aware Caching and Scheduling [arXiv]

  • HKUST & Alibaba

  • Our work!

Acronyms

• LLM: Large Language Model

• MoE: Mixture-of-Experts

• LoRA: Low-Rank Adaptation