Scholar
Thien Huu Nguyen
Google Scholar ID: Da2FhegAAAAJ
University of Oregon
Citations & Impact
Citations
8,605
H-index
41
i10-index
104
Publications
20
Co-authors
15
list available
Publications
Cited
28
Resume (English only)
Academic Achievements
- In 2024, released Vistral, a Vietnamese large language model based on Mistral 7B, which significantly outperforms ChatGPT on reliable Vietnamese LLM benchmarks
- Led the development of CulturaX, a multilingual dataset with 6.3 trillion tokens in 167 languages, adopted by Stability AI to train Stable LM 2 1.6B
- Created the Okapi framework for evaluating multilingual LLMs across 26 languages, integrated into EleutherAI's Language Model Evaluation Harness
- Published a survey paper on recent advances in NLP via large pre-trained models, accepted by ACM Computing Surveys (Impact Factor: 14.324) in 2023
- Conducted a comprehensive evaluation of ChatGPT across 7 tasks and 37 languages in 2023
- Awarded the NSF CAREER Award in 2023 to support research on multilingual learning and information extraction
