Latest updates, announcements, and highlights from CyberPUNK Lab.
🎉 Our PhD student Shuyan Zou's paper "SlaClip: Gradient Norm Slacks can be Indicator for Adaptive Clipping in DP-SGD" was accepted to ICML 2026 as a Spotlight (top 2.2%) paper. Congratulations to Shuyan and all co-authors!
🎉 Our paper "An Adaptive Mini-Batching Strategy for Reliable Streaming Data Delivery in Real-Time" was accepted to IEEE Transactions on Reliability. Congrats to all co-authors!
🎉 Our PhD student Jiawei Wang's paper "Peeling the Layers of Privacy-Utility Onion on Tabular Data" was accepted to the EuroMLSys Workshop, co-located with EuroSys 2026. Congratulations to Jiawei and all co-authors!
Hengyang Yao, from University of Birmingham, joined CyberPUNK lab as a PhD student. Welcome!
🎉 Our paper "FedEvolve: Evolutionary Tabular Data Synthesis in Vertical Federated Learning Systems" was accepted to the RSD Workshop @ AAAI 2026. Congrats to Adethya Srinivasan — such an achievement for a second-year BSc student!
Jiawei Wang, from Sichuan University, joined CyberPUNK lab as a PhD student and is studying machine unlearning. Welcome!
Our lab was invited to Durham University for the RISCS Cybersecurity event, where Han Wu gave a talk on Auditing Privacy-preserving AI.
Our lab supported BSc Year 2 student Adethya Srinivasan in securing the Aniket Singh Scholarship (funded by University of Southampton) to work on a 12-week research project in Vertical Federated Learning.
🔥 Our paper "GTV: Generating Tabular Data via Vertical Federated Learning" was accepted by DSN 2025. Congrats!
Our paper "Impacts of Physical-Layer Information on Epidemic Spreading in Cyber-Physical Networked Systems" was accepted by IEEE Transactions on Circuits and Systems I: Regular Papers.
Han Wu was invited to Heriot-Watt University to give a talk on "Together We Learn, Individually We Protect: A Vision For Differential Privacy".
