I am pursuing a Ph.D. in Computer Science at the University of Texas at Arlington, advised by Dr. Miao Yin. I am also a research intern at Argonne National Laboratory, advised by Dr. Sheng Di.
My current research centers on systems and infrastructure for agentic AI, with a focus on accelerating reinforcement learning (RL) training for large-scale agentic models, including parallelism, scheduling, and communication that make long-horizon, tool-using agents tractable to train at scale. Earlier, I worked on scalable distributed training for 3D Gaussian Splatting, and on HPC data reduction and lossy compression as well as efficient training and adaptation of foundation models. Across these directions, I emphasize algorithm–system co-design, treating research as a cycle of observation, hypothesis, and validation.
Outside of academics, I enjoy freestyle swimming and previously represented Tongji University in swimming competitions in Shanghai.
🔥 News
- 2026.05: 🎉 Started my research internship at Argonne National Laboratory, working on agentic RL research.
- 2026.05: 🎉 Honored that our HPC research line, to which I contributed three first-authored papers (GWLZ, NeurLZ, FLARE), provided the foundation for Dr. Miao Yin’s 2026 NSF CAREER Award on algorithm–hardware co-design for neural-enhanced scientific data compression.
📖 Educations
- Ph.D. in Computer Science, University of Texas at Arlington, Arlington, TX, USA
- Advisor: Dr. Miao Yin
- Focus: Systems and infrastructure for agentic AI and RL training
- M.S. in Mechanical Engineering, Tongji University, Shanghai, China
- B.S. in Mechanical Engineering, Tongji University, Shanghai, China
- National Scholarship of China, awarded twice (Top 0.2%)
📝 Publications
🎯 3D Reconstruction
Wenqi Jia*, Zhewen Hu*, Ying Huang*, Yu Gong, Stavros Kalafatis, Yuke Wang, Wei Niu, Chengming Zhang, Ang Li, Sheng Di, Yuede Ji, Bo Fang, Miao Yin
- Pixel-level communication framework that reduces spatial and saturation redundancy for scalable distributed 3D Gaussian Splatting training.
Yangming Zhang*, Wenqi Jia*, Wei Niu, Miao Yin
- A novel optimization-based framework that simplifies 3D Gaussian Splatting for compact storage while preserving high-quality rendering.
💻 High Performance Computing
Wenqi Jia, Zhewen Hu, Baixi Sun, Yafan Huang, Jiannan Tian, Boyuan Zhang, Daoce Wang, Sian Jin, Luanzheng Guo, Sheng Di, Yuede Ji and Miao Yin
- A scalable hardware architecture that combines dataflow-aware design with neural-hybrid approaches for efficient scientific lossy compression.
NeurLZ: An Online Neural Learning-based Method to Enhance Scientific Lossy Compression
Wenqi Jia, Zhewen Hu, Youyuan Liu, Boyuan Zhang, Jinzhen Wang, Jinyang Liu, Wei Niu, Stavros Kalafatis, Junzhou Huang, Sian Jin, Daoce Wang, Jiannan Tian, Miao Yin
- An online neural learning-based method to enhance scientific lossy compression for large-scale scientific simulations and data analysis.
Advancing Scientific Data Compression via Cross-field Prediction
Youyuan Liu, Wenqi Jia, Taolue Yang, Jiang Bo, Miao Yin, Sian Jin
- Advanced scientific data compression technique utilizing cross-field prediction to enhance compression efficiency for multi-field scientific datasets.
GWLZ: A Group-wise Learning-based Lossy Compression Framework for Scientific Data
Wenqi Jia, Sian Jin, Jinzhen Wang, Wei Niu, Dingwen Tao, Miao Yin
- Group-wise residual learning framework that improves scientific lossy compression with compact learned model weights.
- Enhancing Lossy Compression Through Cross-field Information for Scientific Applications Youyuan Liu, Wenqi Jia, Taolue Yang, Miao Yin, Sian Jin. SC 2024 Workshop
🚀 Efficient Foundation Models
AdaRing: Towards Ultra-light Vision-language Adaptation via Cross-layer Tensor Ring Decomposition
Ying Huang, Yuanbin Man, Wenqi Jia, Zhengzhong Tu, Junzhou Huang, Miao Yin
- Ultra-light parameter-efficient fine-tuning framework for vision-language models using cross-layer tensor ring decomposition with diverse adapters.
ECP-ViT: Real-time Core-periphery Guided ViT with Smart Data Layout Selection on Mobile Devices
Zhihao Shu*, Xiaowei Yu*, Zihao Wu, Wenqi Jia, Yinchen Shi, Miao Yin, Tianming Liu, Dajiang Zhu, Wei Niu
- Real-time Vision Transformer framework employing brain-inspired core-periphery principles with hardware-friendly optimizations for efficient deployment on mobile devices.
Cheng Yang, Yang Sui, Jinqi Xiao, Lingyi Huang, Yu Gong, Yuanlin Duan, Wenqi Jia, Miao Yin, Yu Cheng, Bo Yuan
- Two-stage compression method for Mixture of Experts models combining inter-expert pruning and intra-expert low-rank decomposition to reduce model size while maintaining performance.
🎖 Honors and Awards
- Jeff and Lisa Smith Outstanding Graduate Researcher Award (Ph.D. period, three awardees)
- John S. Schuchman Outstanding Doctoral Student Award (Ph.D. period, three awardees)
- International Conference on Supercomputing (ICS) 2025 Travel Grant (Ph.D. period, four awardees)
🤝 Academic Services
Conference Reviewer
- Conference on Neural Information Processing Systems (NeurIPS), 2026
- IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
- Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL), 2025
- ACM/IEEE Design Automation Conference (DAC), 2025
🛠 Technical Skills
• Programming: Python, C++, CUDA (custom kernels)
• ML & LLM Systems: PyTorch, SGLang, Megatron, slime, Hugging Face Transformers
• Distributed & HPC: MPI, NCCL, OpenMP, SLURM, GPU profiling (NVIDIA Nsight)
• Tools: Git, Docker, Linux/Bash, LaTeX, cuda-gdb
• Mathematical Foundations: Probability & Statistics, Linear Algebra, Convex Optimization, ODE/PDE
