Program

Date: 11 May 2026, 09:15 - 17:30
Venue: Lamenais 3 (Level 3), Palais des Congrès de Saint Malo

Welcome
9:15 - 9:30
Abbas Rahimi (IBM Zurich), Stephen Xia (Northwestern University), Xiaomin Ouyang (HKUST)

Keynote 1: Is Information Digital? A Defense of Reality
9:30 - 10:30
Speaker: Edward A. Lee (UC Berkeley)
Moderator: Abbas Rahimi (IBM Zurich)

Coffee Break
10:30 - 11:00

Session 1 - Foundation Models and Methods
11:00 - 11:40
Session Chair: Hokeun Kim (ASU)

Exploring the Capabilities of LLMs for Multimodal Group Interaction Prediction
Diana Romero, Xin Gao, Daniel Khalkhali, Salma Elmalaki (University of California, Irvine)

STRIDE: Bidirectional Soft-Token Propagation for Faster Diffusion Language Models
Yejia Liu, Xiaomin Ouyang (HKUST)

Towards Optimizing LLM-based Sensor Data Processing: An Empirical Study of Evolutionary Algorithms
Pengrui Quan (UCLA), Brian Wang (UCLA), Xiaomin Ouyang (HKUST), Kang Yang (UCLA), Mani Srivastava (UCLA)

Can Foundation Models Revolutionize Mobile AR Sparse Sensing?
Yiqin Zhao (Rochester Institute of Technology), Tian Guo (Worcester Polytechnic Institute)

Session 2 - Agents and Human-Centered AI
11:40 - 12:30
Session Chair: Shengzhong Liu (Shanghai Jiaotong University)

AgentExplorer: Parallel Online Exploration for Enhancing On-Device Mobile GUI Agents
Runxi Huang, Xiaomin Ouyang (HKUST)

Agentic Driving Coach: Robustness and Determinism of Agentic AI-Powered Human-in-the-Loop Cyber-Physical Systems
Deeksha Prahlad, Daniel Fan, Hokeun Kim (Arizona State University)

Gradient Angles for Runtime Detection of Body-Position Shift in Human Activity Recognition
Sayantan Kuila (Indian Institute of Technology, Kharagpur), Soumyajit Chatterjee (Brave Software & University of Cambridge), Sandip Chakraborty (Indian Institute of Technology, Kharagpur)

HART: A Pretrained Time-Series Foundation Model and Benchmark for Human Activity Recognition
Suyog Jare, Kajeeth Kumar G, Naman Srivastava, Pandarasamy Arjunan (Indian Institute of Science)

Lunch Break
12:30 - 14:00

Keynote 2: From Foundation Models to Tiny Intelligence: Adaptive AI Systems for the Physical World
14:00 - 15:00
Speaker: Elisabetta Farella (FBK, Italy)
Moderator: Abbas Rahimi (IBM Zurich)

Session 3 - Systems, Infrastructure, and Frameworks
15:00 - 16:00
Session Chair: Xiaomin Ouyang (HKUST)

A KG-LLM Pipeline for Automated Regulatory Alignment in CPS-IoT Systems
Veronica Muriga, Jiening Yang, Jason Li, Swarun Kumar (Carnegie Mellon University)

MindFusion: Communication-Efficient Mobile-Cloud Orchestration for MLLMs
Xuwang Zeng, Qianyi Huang (Sun Yat-sen University)

X-CPS: A Top-Down Hierarchical Framework for Explaining Learning-enabled Cyber-Physical Systems
Guocheng He, Lingwen Deng, Ziyan An, Meiyi Ma (Vanderbilt University)

Characterizing Cellular Network Dynamics for Federated Human Activity Recognition
Muhammad Abdullah Soomro, Muhammad Shayan Nazeer, Taqi Raza, Fatima Anwar (University of Massachusetts Amherst)

UNIFERENCE: A Discrete Event Simulation Framework for Developing Distributed AI Models
Doğaç Eldenk, Stephen Xia (Northwestern University)

Coffee Break
16:00 - 16:30

Panel Discussion: Where Shall We Go Next? Repositioning CPS/IoT Research in the Era of Foundation Models: Autonomy, Models, Safety, and Architectures
16:30 - 17:30
Panel Members: Tarek Abdelzaher (UIUC), Qian Zhang (HKUST), Samarjit Chakraborty (UNC Chapel Hill), Salma Elmalaki (University of California, Irvine), Shengzhong Liu (Shanghai Jiaotong University)
Moderator: Xiaomin Ouyang (HKUST)

Closing Remarks and Awards
17:30 - 17:40
Abbas Rahimi (IBM Zurich), Stephen Xia (Northwestern University), Xiaomin Ouyang (HKUST)