Sheng Wang
PhD Graduate in Computer Science
The University of Hong Kong
Research Interests
About
I am Sheng Wang, a PhD graduate in the School of Computing and Data Science from The University of Hong Kong, where I was advised by Prof. Chuan Wu and Prof. Lingpeng Kong. Before that, I received my Bachelor of Engineering from the School of Artificial Intelligence and Automation at Huazhong University of Science and Technology.
My research interests center on agent systems, LLM super-alignment, and data synthesis. Covering these directions, I have published papers at top-tier venues including NeurIPS, ICLR, ACL, and EMNLP. In research, I believe in the power of impactful features, disruptive innovation, and scalability to keep pushing the frontier of intelligence.
Beyond academia, I am deeply interested in startups and product building, especially in the family education space. To me, academia, engineering, and entrepreneurship are three distinct lifestyles, each revealing a colorful facet of life. As a personal philosophy, I strongly believe in the power of agile iteration, continuously refining mental models, and building a stronger cognitive flywheel. Recently, I have been enjoying using agents to make everyday life a bit more fun, while experimenting with an AI-native way of working and living.
If any of this resonates with you, feel free to reach out.
Selected Publications
View All →ToolSelf: Unifying Task Execution and Self-Reconfiguration via Tool-Driven Intrinsic Adaptation
Jingqi Zhou, Sheng Wang†, DeZhao Deng, Junwen Lu, Junwei Su, Qintong Li, Jiahui Gao, Hao Wu, Jiyue Jiang, Lingpeng Kong, Chuan Wu†
arXiv 2026
TherapyGym: Evaluating and Aligning Clinical Fidelity and Safety in Therapy Chatbots
Fangrui Huang, Souhad Chbeir, Arpandeep Khatua, Sheng Wang, Sijun Tan, Kenan Ye, Lily Bailey, Merryn Daniel, Ryan Louie, Sanmi Koyejo, Ehsan Adeli
arXiv 2026
ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and Wisdom
Jingqi Zhou, Sheng Wang†, Jingwei Dong, Kai Liu, Lei Li, Jiahui Gao, Jiyue Jiang, Lingpeng Kong, Chuan Wu†
Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing
QSpec: Speculative Decoding with Complementary Quantization Schemes
Juntao Zhao*, Wenhao Lu*, Sheng Wang*, Lingpeng Kong, Chuan Wu
Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing
TreeSynth: Synthesizing Diverse Data from Scratch via Tree-Guided Subspace Partitioning
Sheng Wang*, Pengan Chen*, Jingqi Zhou*, Qintong Li, Jingwei Dong, Jiahui Gao, Boyang Xue, Jiyue Jiang, Lingpeng Kong, Chuan Wu
Neural Information Processing Systems (NeurIPS) 2025 - Spotlight
MoS: Unleashing Parameter Efficiency of Low-Rank Adaptation with Mixture of Shards
Sheng Wang*, Liheng Chen*, Pengan Chen, Jingwei Dong, Boyang Xue, Jiyue Jiang, Lingpeng Kong, Chuan Wu
The Thirteenth International Conference on Learning Representations
News
ToolSelf and TherapyGym were prepared as ICML 2026 submissions on agentic systems and therapy chatbots. 🎉
Graduated with a PhD in the School of Computing and Data Science from The University of Hong Kong. 🎉
TreeSynth was selected as a Spotlight paper at NeurIPS 2025. 🎉
QSpec and ProReason appeared at EMNLP 2025, covering efficient inference and multimodal reasoning.
MoS and Forewarned were accepted to ICLR 2025.
I began my entrepreneurial journey in family education. 🎉
PRoLoRA and LoRA Meets Dropout were accepted to ACL 2024.