|
Yun-Ta Hsieh I am currently pursuing my M.A.S. in Computer Science at the University of Pennsylvania, after receiving my M.S. in Digital and Material Technologies from the University of Michigan, and B.Arch. from Huazhong University of Science and Technology. I am also a research assistant at The Ohio State University, working with Prof. Mi Zhang. My current work focuses on multimodal agents, vision-language-action systems, and efficient ML systems. CV / Google Scholar / GitHub |
Hello!Welcome to my personal academic website. My research interests include multimodal agents, efficient vision-language-action systems, and efficient ML systems. I am always open to conversations and collaborations with people who share these interests. |
Beyond Computer ScienceBeyond computer science, I was trained in architecture and digital fabrication. My earlier work focused on spatial design, material systems, and computational workflows for making physical things. That background still shapes how I think about AI systems: I care about how models interact with interfaces, environments, and real-world constraints, not only how they perform on static tasks. |
Selected Publications
QuantVLA: Scale-Calibrated Post-Training Quantization for Vision-Language-Action Models
Jingxuan Zhang*, Yunta Hsieh*, Zhongwei Wan, Haokun Lin, Xin Wang, Ziqi Wang, Yingtie Lei, Mi Zhang CVPR 2026 Project / Code
OS-Omni: A Cross-Platform Benchmark for Generalist Computer-Using Agents
Hui Shen*, Yunta Hsieh*, Jianing Ma*, Ziyuan Liu*, Qi Han*, Xiuqi Xu*, Yanheng Shang*, ..., Ben Athiwaratkun, Qiushi Sun, Mi Zhang, Ping Luo, Wenhu Chen, Ngai Wong Submitted to NeurIPS 2026 Preprints
MathGen: Revealing the Illusion of Mathematical Competence through Text-to-Image Generation
R Liu, Hui Shen, Peng Zhang, Yunta Hsieh, Yuyue Zhang, J Xu, S Chen, et al. arXiv preprint, 2026
MMSpec: Benchmarking Speculative Decoding for Vision-Language Models
Hui Shen, Xin Wang, Peng Zhang, Yunta Hsieh, Qi Han, Zhongwei Wan, Zhe Zhang, Jingxuan Zhang, et al. arXiv preprint, 2026
SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills
Yingtie Lei, Zhongwei Wan, Jiankun Zhang, Samiul Alam, Zixuan Zhong, Peizhou Huang, Xin Wang, Jingxuan Zhang, Donghao Zhou, Yunta Hsieh, Zhihao Dou, Hui Shen, et al. Submitted to NeurIPS 2026
Speculative Decoding for Multimodal Models: A Survey
Yuyue Zhang, Y Wang, Yunta Hsieh, Xin Wang, Peng Zhang, Z Yang, J Ma, Z Zhao, et al. TMLR 2026
MMFormalizer: Multimodal Autoformalization in the Wild
Jing Xiong, Qi Han, Yunta Hsieh, Hui Shen, Hongsheng Xin, Chaofan Tao, Chenyu Zhao, et al. arXiv preprint, 2026
PhyX: Does Your Model Have the "Wits" for Physical Reasoning?
Hui Shen, Taiqiang Wu, Qi Han, Yunta Hsieh, Jizhou Wang, Yuyue Zhang, Yuxin Cheng, et al. arXiv preprint, 2025 OpenReview |
|
University of Pennsylvania, USA
|
|
University of Michigan, USA
|
|
Huazhong University of Science and Technology, China
|
|
Email: yunta@seas.upenn.edu GitHub: Kiky-88 CV and profile photo will be added later. |