isl-logo


We are the Intelligent System (IntelliSys) Lab at the ECE Department of Stevens Institute of Technology. Our research bridges artificial intelligence (AI) and novel system architectures by designing efficient and robust AI algorithms and developing intelligent and high-performance computing systems. Check our roadmap.


News

08/23/2025 :pencil: Our paper “Taming Latency-Memory Trade-Off in MoE-Based LLM Serving via Fine-Grained Expert Offloading” has been accepted by EuroSys 2026. Find our arXiv paper here.
08/20/2025 :pencil: Our paper “pFedGPT: Hierarchically Optimizing LoRA Aggregation Weights for Personalized Federated GPT Models” has been accepted for presentation at the EMNLP 2025 Main Conference (acceptance rate 17.35%).
08/01/2025 :bookmark: Our two K-12 interns have successfully presented their project ``Donkey car: How to Train your AI’’ at Stevens ACES STARS Launch Day. Check their project here!
06/25/2025 :pencil: Our paper ``Stealthy Backdoor Attack in Federated Learning via Adaptive Layer-wise Gradient Alignment’’ has been accepted by ICCV’25.
06/09/2025 :bookmark: Our work on serverless eco-systems for AI has been reported by News@Indiana University.
More...

Selected Publications

  1. KDD
    CriticalFL: A Critical Learning Periods Augmented Client Selection Framework for Efficient Federated Learning
    In Proceedings of the ACM Special Interest Group on Knowledge Discovery and Data Mining (KDD) 2023
  2. ASPLOS
    RainbowCake: Mitigating Cold-starts in Serverless with Layer-wise Container Caching and Sharing
    In Proceedings of the ACM Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) 2024
  3. ICLR
    Backdoor Federated Learning by Poisoning Backdoor-Critical Layers
    In Proceedings of the International Conference on Learning Representations (ICLR) 2024
  4. SC
    Stellaris: Staleness-Aware Distributed Reinforcement Learning with Serverless Computing (Best Student Paper Finalist)
    In Proceedings of the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC) 2024
  5. EuroSys
    Taming Latency-Memory Trade-Off in MoE-Based LLM Serving via Fine-Grained Expert Offloading
    Hanfei Yu, Xingqi Cui, Hong Zhang, Hao Wang@Rutgers, and Hao Wang
    Accepted by ACM European Conference on Computer Systems (EuroSys) 2026

Sponsors