Angela Fan
Angela Fan is an AI researcher whose papers span story generation, long-form question answering, structured dropout and multilingual translation. In 2018 she co-wrote Hierarchical Neural Story Generation, and in 2019 she co-wrote ELI5: Long Form Question Answering. Reducing Transformer Depth on Demand with Structured Dropout was posted to arXiv in September 2019 with two other authors. Beyond English-Centric Multilingual Machine Translation followed in 2021, and Llama 2 (2023) and The Llama 3 Herd of Models (2024) also list her among their authors. She is on OpenAI's scaling team, which she joined in July 2025 after working as an AI researcher at Meta.
Timeline
-
2018
Hierarchical story generation
In 2018 she co-writes Hierarchical Neural Story Generation with Mike Lewis and Yann Dauphin, a paper built on 300K human-written stories paired with writing prompts.
-
2019
ELI5 long-form question answering
She co-writes ELI5: Long Form Question Answering in 2019 with Yacine Jernite, Ethan Perez, David Grangier, Jason Weston and Michael Auli.
-
September 2019
Structured dropout for transformers
Reducing Transformer Depth on Demand with Structured Dropout is posted to arXiv in September 2019 with two other authors.
-
2021
Beyond English-centric translation
She co-authors Beyond English-Centric Multilingual Machine Translation in 2021 with Shruti Bhosale, Holger Schwenk and others.
-
July 2023
Co-authoring Llama 2
Llama 2: Open Foundation and Fine-Tuned Chat Models is posted to arXiv in July 2023 with her name on the author list.
-
2024
Co-authoring the Llama 3 paper
The Llama 3 Herd of Models, posted to arXiv in July 2024, lists her among its authors.
-
Since 2025
Joining OpenAI's scaling team
OpenAI adds Angela Fan, an AI researcher from Meta, to its scaling team in July 2025.
Sources and credits
Photo: We have no published photo of Angela Fan yet. Send us a free-licensed photo.
- [1805.04833] Hierarchical Neural Story Generation
- Hierarchical Neural Story Generation - ACL Anthology
- ELI5: Long Form Question Answering - ACL Anthology
- [1909.11556] Reducing Transformer Depth on Demand with Structured Dropout
- Beyond English-Centric Multilingual Machine Translation
- [2307.09288] Llama 2: Open Foundation and Fine-Tuned Chat Models
- [2407.21783] The Llama 3 Herd of Models
- OpenAI Poaches Top Engineers from Tesla, xAI, and Meta in Escalating AI Talent War
- What Is OpenAI's Scaling Team? Tesla, xAI, Meta Engineers Join Their Ranks as AI Hiring Competition Heats Up