TU Darmstadt
I am a research associate at the UKP Lab, TU Darmstadt, supervised by Prof. Iryna Gurevych. My research focuses on mechanistic interpretability of language models.
My research focuses on understanding the geometry of interventions and steering in language-model representations.
I completed my MSc in NLP at MBZUAI, where I was advised by Prof. Yova Kementchedjhieva and Prof. Kentaro Inui. My thesis explored how large language models can reuse previous computations to speed up text generation without losing accuracy.
During my undergraduate studies, I conducted research under Dr. Kiet Nguyen and Prof. Ngan Nguyen at the UIT NLP Group. My work focused on hate speech detection and Vietnamese NLP.
Besides, I am also a
Feel free to email me about my research, potential collaborations, or related discussions.
Looking for PhD Opportunities in Mechanistic Interpretability
I am seeking PhD opportunities in the United States for Fall 2027 admission, with interests in mechanistic interpretability, language models, and AI safety.
Warning
Problem: The current name of your GitHub Pages repository ("Solution: Please consider renaming the repository to "
http://".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}" in index.html.
Action required
Problem: The current root path of this site is "baseurl ("_config.yml.
Solution: Please set the
baseurl in _config.yml to "Education
-
Mohamed bin Zayed University of Artificial Intelligence -
VNU-HCM University of Information Technology
Experience
-
TU Darmstadt
-
Fujairah Research Center -
VinBigData JSC, VinGroup
News
Selected Publications (view all )
ViHOS: Hate Speech Spans Detection for Vietnamese
European Chapter of the Association for Computational Linguistics (EACL) 2023
Proposed a novel task for Vietnamese NLP, ViHOS - Vietnamese Hate and Offensive Spans, which consists of 12k online comments with advanced annotations on what make a sentence offensive and benchmark experiments.
43 citationsViHOS: Hate Speech Spans Detection for Vietnamese
European Chapter of the Association for Computational Linguistics (EACL) 2023
Proposed a novel task for Vietnamese NLP, ViHOS - Vietnamese Hate and Offensive Spans, which consists of 12k online comments with advanced annotations on what make a sentence offensive and benchmark experiments.