PhD Candidate, Elmore Family School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN · chen4388@purdue.edu · CV
I am an ECE PhD candidate at Purdue University, advised by Professor Christopher G. Brinton, with an expected graduation in December 2026. My research focuses on scalable and reliable distributed AI systems, particularly LLM-based multi-agent systems and collaboration between on-device and cloud language models.
I develop communication-aware coordination and consistency mechanisms for heterogeneous agents, along with reinforcement-learning and training-free methods that let local language models decide when cloud assistance is worthwhile under resource budgets.
My broader work spans federated and fog learning, distributed optimization, and differential privacy. I am interested in turning principled learning and networking methods into dependable AI systems that operate under real-world communication, computation, privacy, and reliability constraints.
Scalable coordination and consistency for heterogeneous LLM agents.
Resource-aware routing and RL post-training for collaborative reasoning.
Federated optimization across communication, trust, and privacy constraints.
Preprint · 2026
CARGO uses inference-time response agreement, Bayesian early stopping, and lightweight calibration to route requests between local and cloud LLMs at a controllable collaboration ratio—without training an additional router.
PaperNeurIPS 2026 · Under Review
A communication-aware framework for coordinating heterogeneous LLM agents across varying capabilities, tasks, and network constraints.
NeurIPS 2026 · Under Review
DA-GRPO jointly improves a local language model and its cloud-collaboration behavior while controlling assistance budgets and reducing forgetting across changing tasks.
PaperICML 2026
A unified approach that teaches on-device LLMs to reason locally and invoke cloud assistance selectively through reinforcement-learning-based post-training.
PaperNeurIPS 2026 · Under Review
A gradient-correction perspective for improving the stability and effectiveness of adaptive optimization in heterogeneous federated learning.
IEEE Network · Under Review
A systems perspective on deploying federated foundation models where connectivity is intermittent, devices are resource-constrained, and data is noisy.
PaperIEEE/ACM Transactions on Networking
Semi-Decentralized Gradient Tracking mitigates subnet drift in hierarchical fog networks without relying on restrictive gradient-diversity assumptions.
PaperACC 2026
A probabilistic transformer approach to battery remaining-useful-life estimation and uncertainty-aware prediction.
IEEE/ACM Transactions on Networking
A privacy-preserving framework that adapts noise injection across network tiers with heterogeneous trust and resource constraints.
PaperNeurIPS 2025
MU-SplitFed reduces synchronization delays from straggling clients through unbalanced server-side updates.
NeurIPS 2024
A multi-timescale gradient-correction method for improving convergence across heterogeneous hierarchical network tiers.
PaperImproved the stability and performance of a production-scale security ML system and developed configurable agent-based simulations for realistic enterprise interaction traces.
Purdue University · Intelligence Optimization for Networks (ION) Lab
Developing scalable distributed multi-agent systems, on-device and cloud LLM collaboration, privacy-preserving fog learning, and communication-aware distributed optimization.
Cummins Inc. · Research and Technology
Built real-world federated learning applications using NVIDIA FLARE and Meta ExecuTorch.
National Chiao Tung University · Institute of Electronics
Conducted research on transformer models, computer vision, and video compression.
Doctor of Philosophy
Elmore Family School of Electrical and Computer Engineering
PhD candidate advised by Professor Christopher G. Brinton · Expected December 2026
Bachelor of Science
Electronics Engineering
Ranked 2 of 89