Draft:Angela Fan

Angela Fan, Ph.D (born 1994) is a Chinese-American research scientist specializing in machine translation and natural language processing. She is known for her work on large language models, particularly as a contributing developer of Meta AI's Llama 2 and Llama 3 models and as a researcher of many-to-many machine translation models. In 2024 Fan served as a program chair for the Conference on Neural Information Processing Systems (NeurIPS), one of the most influential conferences in machine learning. She has also served as an invited speaker at multiple academic and professional conferences, including Applied Machine Learning Days (AMLD). Fan has lived in Canada, the U.S., and France. As of 2025, she lives in Paris where she works at Facebook AI Research.

Early life and education

Fan was born in Shanghai and grew up in America. As a child, she loved reading stories and frequently visited libraries on her own. Fan's early research focused on applications of generative AI to storywriting, bringing her childhood interest in storytelling together with her interest in AI. Later, her family moved to the U.S., but they continued to speak Shanghainese at home. This early exposure to multilingualism sparked her interest in machine translation and led Fan to eventually research applications of machine learning with low-resource languages, languages that have significantly less data, tools, and infrastructure available online for training NLP models.

After moving to the U.S., Fan attended Stuyvesant High School where she earned a variety of accolades in her senior year. She was awarded the Intel International Science and Engineering Fair Grand Award. Fan was also an Intel Science Talent Search Finalist, a Siemens Competition in Math, Science & Technology Semifinalist, and an attendant of the Stockholm International Youth Science Seminar.

Fan completed her undergraduate education at Harvard University in 2016 where she majored in Statistics and minored in Computer Science. She graduated Magna Cum Laude and was president of the Harvard Premedical Society in her third year.

After graduating, Fan worked as a research engineer at Facebook AI Research for two and a half years. She then returned to academia and earned her Ph.D in Computer Science under her advisors Antoine Bordes, Claire Gardent, and Chloe Braud, specializing in text generation through a joint program between Inria Centre at the University of Lorraine (INRIA Nancy) and Facebook AI Research (FAIR) Paris.

Career and research

Fan’s work focuses on text generation and multilingual AI systems. Fan has made significant contributions to the field of machine translation, particularly in developing systems that support low-resource languages. As a key researcher behind Meta AI's No Language Left Behind (NLLB) project, she helped build a highly accurate multilingual translation model that supports over 200 languages, many of which previously lacked robust AI-driven translation tools. Her work emphasizes improving translation quality through innovations in data efficiency, model scaling, and self-supervised learning techniques. By focusing on linguistic diversity, Fan’s research has helped improve access to information by enabling speakers of underrepresented languages to benefit from AI advancements. Her contributions continue to shape the future of machine translation, making it more inclusive and equitable across global communities.

Fan's work leverages generative AI to address content gaps on platforms like Wikipedia, particularly in representing women in STEM. She led research on using AI-generated text to create high-quality, well-sourced Wikipedia articles about notable women scientists, helping to counteract the systemic underrepresentation of women in STEM fields. In another research project, Fan demonstrates how AI can assist in expanding knowledge equity while maintaining factual accuracy and neutrality by training models to generate biographical entries based on reliable data. These initiatives not only showcase the potential of generative AI in content creation but also highlight its role in promoting diversity and inclusion in digital knowledge spaces.

Notable accomplishments

Fan has achieved a variety of notable accomplishments throughout her research career. She has authored or contributed to over 60 published papers, multiple of which have won various awards. In 2018 Fan, Mike Lewis and Yann Dauphin received Honorable Mention from the Association for Computational Linguistics (ACL) in the category "Best Long Paper" category for their paper Hierarchical Neural Story Generation. In 2023, Fan and her team received the Linguistic Diversity Area Chair Award again by the ACL for their paper Small Data, Big Impact: Leveraging Minimal Data for Effective Machine Translation.

In addition to her research contributions, Fan has been an invited speaker at prestigious academic and professional conferences, including Applied Machine Learning Days (AMLD), where she has shared insights on AI for language technology and inclusion. She has also played an active role in the research community, serving as a program chair for the 2024 Conference on Neural Information Processing Systems (NeurIPS).

Notable papers

• Touvron, Hugo; Martin, Louis; Stone, Kevin; Albert, Peter; Almahairi, Amjad; Babaei, Yasmine; Bashlykov, Nikolay; Batra, Soumya; Bhargava, Prajjwal (July 19, 2023), Llama 2: Open Foundation and Fine-Tuned Chat Models, arXiv:2307.09288
• Ott, Myle; Edunov, Sergey; Baevski, Alexei; Fan, Angela; Gross, Sam; Ng, Nathan; Grangier, David; Auli, Michael (April 1, 2019), fairseq: A Fast, Extensible Toolkit for Sequence Modeling, arXiv:1904.01038
• Grattafiori, Aaron; Dubey, Abhimanyu; Jauhri, Abhinav; Pandey, Abhinav; Kadian, Abhishek; Al-Dahle, Ahmad; Letman, Aiesha; Mathur, Akhil; Schelten, Alan (November 23, 2024), The Llama 3 Herd of Models, arXiv:2407.21783
• Dauphin, Yann N.; Fan, Angela; Auli, Michael; Grangier, David (August 6, 2017). "‪Language modeling with gated convolutional networks‬" (PDF). Proceedings of the 34th International Conference on Machine Learning. 70: 933–941.
• Fan, Angela; Lewis, Mike; Dauphin, Yann (May 13, 2018), Hierarchical Neural Story Generation, arXiv:1805.04833

References

This page will be placed in the following categories if it is moved to the article namespace.

Categories:

• 1994 births
• American artificial intelligence researchers
• Facebook employees
• Harvard University alumni
• Living people
• Stuyvesant High School alumni