About
I am a Machine Learning Engineer at EmbodyX, working on efficient AI systems, large model deployment, and embodied AI.
Recently, my work has focused on embodied AI systems, including model deployment on Unitree G1 and UR-series robotic arms. I also have experience with Mixture-of-Experts large language models, including pruning, quantization, efficient decoding, and edge deployment.
Previously, I worked on real-time AI applications such as TTS-integrated chatting avatars and streaming pipelines. Before that, I was a Software Engineering Co-op at Cognex Corporation, where I contributed to VisionPro software.
I received my M.S. from Northeastern University and my B.E. from Beijing University of Technology.
Research Interests
π Selected Publications
Effective MoE-based LLM Compression by Exploiting Heterogeneous Inter-Group Experts Routing Frequency and Information Density
ICML 2026 | Paper
Are VLMs Ready for Lane Topology Awareness in Autonomous Driving?
ICASSP | Paper
Collaborative Compression for Large-Scale MoE Deployment on Edge
ICASSP | Paper
From Static to Dynamic: a Survey of Topology-Aware Perception in Autonomous Driving
ICCV WDFM-AD | Paper
Unsupervised Search for Ethnic Minorities' Disc-Cup Segmentation Training Set
ICASSP | Paper
Education
- M.S. in Electrical and Computer Engineering, Northeastern University
- B.E. in Electrical Engineering, Beijing University of Technology
Experience
- Machine Learning Engineer, EmbodyX Inc. & AIbao LLC
- Software Engineering Co-op, Cognex Corporation
