Hwang's Research
Research Beyond Boundaries
Advancing intelligent systems, autonomous driving, quantum computing, and federated learning — independent research at the frontier of AI.
A research collective at the frontier of intelligent systems.
Hwang's Research is an independent research organization focused on the systems that will define how machines learn, perceive, and reason about the world.
Our collaborative work spans federated learning, autonomous driving, and quantum computing — three frontiers where progress demands deep technical depth and a willingness to build infrastructure that doesn't yet exist.
We publish openly, release code, and design our research to be reproducible and useful beyond our own work.
Read more about our missionDistributed, privacy-preserving learning across heterogeneous clients. Adaptive aggregation, continual participation, and communication efficiency.
Mamba state-space models for efficient long-horizon perception and planning in end-to-end autonomous systems.
Quantum circuit simulation and algorithm prototyping for hybrid classical-quantum workloads.
Where we focus
Seven areas of active investigation — from distributed learning to quantum simulation.
Federated Learning
Privacy-preserving distributed learning across heterogeneous clients. Adaptive aggregation, continual participation, and communication efficiency.
Autonomous Driving
Mamba state-space models for efficient long-horizon perception and planning in end-to-end autonomous systems.
Quantum Computing
Quantum circuit simulation and algorithm prototyping for hybrid classical-quantum workloads.
AI Systems
Scalable infrastructure and inference engines for training and serving large AI models efficiently.
AI Agents
Multi-agent frameworks, tool use, and autonomous reasoning systems built on top of foundation models.
Security
Robust and secure ML systems with adversarial resilience and verifiable privacy guarantees.
Edge AI
On-device inference, model compression, and resource-efficient learning for constrained environments.
Time-Series Forecasting
New architectures for forecasting sequential data — efficiency, accuracy, and adaptation across non-stationary domains.
What we're building
CI-FL
ActiveCache-Inspired Federated Learning
Cache-inspired federated learning framework for efficient communication and scalable edge intelligence.
H2PFUser
ActiveMamba-based Autonomous Driving
End-to-end autonomous driving stack built on Mamba state-space models for efficient long-horizon perception and planning.
AdaFed
ActiveAdaptive Selective Aggregation for Heterogeneous FL
Adaptive selective aggregation strategy for federated learning across heterogeneous model architectures. Tier-based partial backbone sharing with adaptive blending and validation-based rollback to mitigate negative transfer — evaluated on a fleet of seven autonomous driving models in CARLA.
PRISM
ActiveQuantum Simulator
Quantum qubit simulator for quantum computing research and experimentation. Designed for algorithm prototyping and hybrid quantum-classical workloads.
