Sheng-Lun (Kevin) Wei
Postdoctoral Researcher @ Academia Sinica
Adjunct Instructor @ National Taiwan University
Sheng-Lun (Kevin) Wei is a Postdoctoral Researcher at Academia Sinica and an Adjunct Instructor at National Taiwan University. Kevin received a PhD in Computer Science from 
National Taiwan University in 2026, with research on the reliability of large language models, including the biases they carry, how they behave when judging other models, and how multimodal models reason. Kevin also brings seven years of industry experience as a machine learning and data engineer at 
KKStream, 
Junyi Academy, and 
ShopBack, bridging research and production systems. From 2022 to 2025, Kevin co-founded and served as CTO of 
OrbitNext, a creator economy AI startup. Kevin teaches programming at NTU's Center of General Education and Department of Economics, and in 2016 founded 
ccClub (社團法人攜曦程式推廣學會), a non-profit for programming education that has served more than 4,500 learners.
Research Interests
Selected Publications
Recent research in Natural Language Processing and Large Language Models
Unveiling Selection Biases: Exploring Order and Token Sensitivity in Large Language Models
Sheng-Lun Wei, Cheng-Kuang Wu, Hen-Hsen Huang, Hsin-Hsi Chen
Bias in the Ear of the Listener: Assessing Sensitivity in Audio Language Models Across Linguistic, Demographic, and Positional Variations
Sheng-Lun Wei, Yu-Ling Liao, Yen-Hua Chang, Hen-Hsen Huang, Hsin-Hsi Chen
No One Fits All: From Fixed Prompting to Learned Routing in Multilingual LLMs
Wei-Chi Wu, Sheng-Lun Wei, Hen-Hsen Huang, Hsin-Hsi Chen
Induct-Learn: Short Phrase Prompting with Instruction Induction
Po-Chun Chen, Sheng-Lun Wei, Hen-Hsen Huang, Hsin-Hsi Chen