Affiliations & Funding
- Integreat —the Norwegian Center for Knowledge-Driven Machine Learning
- Language Technology Group (LTG) in the Department of Informatics
- Marie Skłodowska-Curie Actions (DSTrain)
Academic Interests
I am a Postdoctoral Fellow at the University of Oslo (UiO), specializing in natural language processing and large language models.
My research interests include:
Verifiable & truthful generation in LLMs
Knowledge graph construction & maintenance
Few-shot information extraction (low-resource & multilingual settings)
Postdoctoral Project
Title: " Separating Linguistics from Knowledge in Large Language Models " — funded by the Marie Skłodowska-Curie Actions DSTrain - Data Science Postdoctoral Training Programme .
This project aims to design modular Large Language Models (LLMs) that explicitly separate linguistic competence from factual knowledge. The core innovation is to train a lightweight base model to capture linguistic structure, while using an external, dynamically updated knowledge graph to store and retrieve factual knowledge.
With this project, we address key limitations in current LLMs—hallucinations, inefficiency of updates, and limited interpretability—by creating an architecture in which knowledge and language are disentangled from the outset.
Background
2022 - 2024 Working as a postdoc at Tsinghua University and the China Data Cleaning Center, focusing on financial fraud identification and money laundering detection.
2017 - 2022 Finished a PhD in Computer Science at Tsinghua University in 2022, supervised by Prof. Bin Xu and Prof. Juanzi Li.
My PhD thesis, "Low-resource Event Extraction", proposes multimodal learning, knowledge distillation, and self-labeling methods to handle the data scarcity issue in event extraction.
Collaborators
- Andrey Kutuzov , University of Oslo, Department of Informatics
- Egor Kostylev , University of Oslo, Department of Informatics
Publications
For a complete publication list, see my Google Scholar profile.
.