Scholar
Langlin Huang
Google Scholar ID: Mt9xdjYAAAAJ
Washington University in St. Louis
Citations & Impact
Citations
175
H-index
5
i10-index
3
Publications
9
Co-authors
7
list available
Contact
Publications
Resume (English only)
Academic Achievements
- - POSS: Position Specialist Generates Better Draft for Speculative Decoding
- - CrossWordBench: Evaluating the Reasoning Capabilities of LLMs and LVLMs with Controllable Puzzle Generation
- - Efficient Test-Time Scaling via Self-Calibration
- - Divide, Reweight, and Conquer: A Logit Arithmetic Approach for In-Context Learning
- - MoCE: Adaptive Mixture of Contextualization Experts for Byte-based Neural Machine Translation
- - First place in the Chinese-Thai track and second place in the Mongolian-Chinese track at the 18th China Conference on Machine Translation (CCMT 2022)
Research Experience
- - July 2025: CrossWordBench accepted by COLM 2025
- - June 2025: Released PosS, an effective Speculative Decoding method
- - April 2025: Released CrossWordBench, a challenging multimodal Benchmark
- - February 2025: Paper on Adaptive Heterogeneous MoE for flexible tokenization accepted by NAACL 2025 as oral representation
- - February 2025: Released Efficient Test-Time Scaling via Self-Calibration
- - October 2024: Released LARA, a logit aggregation approach for In-Context Learning
- - August 2024: Started a new journey at Washington University in St. Louis
- - June 2024: Paper on Byte-based Language Model accepted to ACL 2024
- - May 21, 2024: Successfully defended Master's dissertation
- - October 2023: Paper on Compositional Phrase Representation accepted to EMNLP 2023
- - June 2023: Released BayLing (百聆), an instruction-following LLM
- - July 2022: Won first place in the Chinese-Thai track and second place in the Mongolian-Chinese track at the 18th China Conference on Machine Translation (CCMT 2022)
- - September 2021: Started M.S. life at University of Chinese Academy of Sciences (UCAS)
- - June 2021: Obtained B.E. degree from University of International Business and Economics (UIBE)
- - June 2020: Published a paper on automatic construction of depression-domain lexicon in JMIR Medical Informatics
Education
- - PhD student at Washington University in St. Louis (WUSTL), advised by Prof. Jiaxin Huang
- - Master's degree from Institute of Computing Technology, Chinese Academy of Sciences (ICT/CAS) in Jul. 2024, studied at ICT Natural Language Processing (ICTNLP) group
- - B.E. degree in Data Science and Big Data Technology from University of International Business and Economics (UIBE) in Jun. 2021
Background
- - Research interests: Natural Language Processing and Large Language Models
- - Particularly interested in: Large Language Model Reasoning, Efficient language model generation, Multilingual large language model, Machine Translation
