占俊飞
占俊飞 Junfei Zhan

PhD Student

Imperial College London

About Me

I’m Junfei Zhan, originally from China, and an incoming PhD student in the Department of Computing at Imperial College London, joining in October 2026 under the supervision of Professor Giuliano Casale. I recently earned my master’s degree in Electrical Engineering at the University of Pennsylvania, where my research was supervised by Professor Saswati Sarkar. My previous research in undergrad was supervised by Professor Tengjiao He and Professor Kwan-Wu Chin.

My research interests center on optimization, Large Language Models (LLMs), and data-driven decision-making, particularly within energy-constrained computing environments. I’ve led and co-authored multiple research projects focused on task scheduling, LLMs, and green IoT networks. I’m passionate about combining mathematical modeling, machine learning, and control theory to tackle real-world system challenges.

Outside of academics, grabbing meals with friends and socializing are among my favorite activities. Sports have also been a significant part of my life, with awards earned in sprinting, rock climbing, rowing, and table tennis.

…
> _
Interests
  • Large Language Models
  • Optimization
  • Network Science
Education

  • PhD in Computing

    Imperial College London, UK

  • MS in Electrical Engineering

    University of Pennsylvania, USA

  • Dual BSc in Applied Mathematics and Information Computing Science

    University of Birmingham, UK & Jinan University, China

News
  • Aug 21 2026
    Paper accepted EMNLP 2026

    "Trains but Doesn’t Learn: A Post-Training Delivery Benchmark for LLM Agents as Forward-Deployed Engineers" was accepted to The 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026).

    With Weihang Ding. Sincere thanks to Weihang for his tremendous effort on this work.

  • Jul 10 2026
    Paper accepted ACM MM 2026

    "Seeing is Free, Speaking is Not: Uncovering the True Energy Bottleneck in Edge VLM Inference" was accepted to The 34th ACM International Conference on Multimedia (ACM MM 2026).

    With Haoxun Shen, Mingang Guo, Zixuan Huang, and Prof. Tengjiao He. Sincere thanks to Prof. He for his guidance.

  • May 10 2026
    Journal accepted IEEE Internet of Things Journal

    "Orchestrating Data Collection and Computation in Green IoT Networks" was accepted by the IEEE Internet of Things Journal.

    With Prof. Fei Song, Prof. Tengjiao He, Prof. Kwan-Wu Chin, and Benyu Chen. Sincere thanks to Prof. He and Prof. Chin for their guidance.

  • Apr 27 2026
    Paper accepted MobiSys '26 Posters

    "Graph Learning-based Update Manipulation Attack on Federated Fine-Tuning of LLMs over Wireless Networks" was accepted to The 24th ACM International Conference on Mobile Systems, Applications, and Services (MobiSys '26 Posters).

    With Hanlin Cai (Cambridge), Zheng Lin (HKU), and Xinyi Cai (Cambridge). Many thanks to Hanlin for leading the collaboration.

  • Apr 24 2026
    Award University of Pennsylvania · ESE Department

    Received the 2026 ESE Department Master's Top 10% GPA Award for outstanding academic performance.

  • Feb 13 2026
    Journal accepted IEEE Internet of Things Journal

    "Joint Function Configuration and Multislot Offloading in Solar-Powered Serverless Edge Computing" was accepted by the IEEE Internet of Things Journal.

    With Prof. Tengjiao He. Sincere thanks to Prof. He for his close guidance throughout, and Benyu for all the support.

  • Jan 16 2026
    Oral IEEE ICASSP 2026

    "SpikeBP: Efficient Spike-Driven Transformer for Blood Pressure Waveform Generation with Frequency Knowledge Distillation" was selected for oral presentation at IEEE ICASSP 2026.

    Many thanks to Zixuan for leading this work.

Publications
  • EMNLP 2026

    Trains but Doesn’t Learn: A Post-Training Delivery Benchmark for LLM Agents as Forward-Deployed Engineers

    W. Ding, J. Zhan

    Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2026. To appear

  • ACM MM 2026

    Seeing is Free, Speaking is Not: Uncovering the True Energy Bottleneck in Edge VLM Inference

    J. Zhan, H. Shen, M. Guo, Z. Huang, T. He

    Proceedings of the 34th ACM International Conference on Multimedia (ACM MM), 2026. To appear · arXiv:2607.09520

  • AAAI 2026

    PRISM: Privacy-Aware Routing for Adaptive Cloud–Edge LLM Inference via Semantic Sketch Collaboration

    J. Zhan, H. Shen, Z. Lin, T. He

    Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2026

  • IoT-J 2026

    Orchestrating Data Collection and Computation in Green IoT Networks

    J. Zhan, T. He, K.-W. Chin, B. Chen, F. Song

    IEEE Internet of Things Journal, 2026

  • IoT-J 2026

    Joint Function Configuration and Multi-Slot Offloading in Solar-Powered Serverless Edge Computing

    B. Chen, T. He, M. Zheng, J. Zhan, B. He

    IEEE Internet of Things Journal, 2026

  • ICASSP 2026

    SpikeBP: Efficient Spike-Driven Transformer for Blood Pressure Waveform Generation with Frequency Knowledge Distillation

    Z. Huang, Z. Wang, W. Qiu, J. Zhan, S. Zou, Y. Li, F. Miao

    IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2026 · Oral

  • MobiSys 2026

    Poster: Graph Learning-based Update Manipulation Attack on Federated Fine-Tuning of LLMs over Wireless Networks

    H. Cai, J. Zhan, Z. Lin, X. Cai

    Proceedings of the 24th ACM International Conference on Mobile Systems, Applications, and Services (MobiSys Posters), 2026

  • SMC 2025

    RL-Enhanced Disturbance-Aware MPC for Fast and Robust UAV Trajectory Tracking

    H. Shen, J. Zhan, T. He

    IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2025

  • IEEE LNET 2024

    Task Offloading and Approximate Computing in Solar Powered IoT Networks

    J. Zhan, J. Wu, T. He, K.-W. Chin

    IEEE Networking Letters, 6(1), 26–30, 2024

For the latest publications, please follow my Google Scholar.

Visitor Map