Gabriel Goh
Also known as Gabe Goh
Gabriel Goh is a researcher whose work includes image generation and multimodal models. In 2017 he published 'Why Momentum Really Works' in Distill. He co-authored two papers in February 2021: 'Zero-Shot Text-to-Image Generation', the transformer paper behind DALL·E, and 'Learning Transferable Visual Models From Natural Language Supervision'. In March 2021 he was research lead on a Distill article about multimodal neurons. As of October 2026, he is a machine learning researcher at OpenAI.
Timeline
-
2017
Why momentum really works
He publishes 'Why Momentum Really Works' in Distill in 2017.
-
February 2021
Zero-shot text-to-image generation
In February 2021 he co-authors 'Zero-Shot Text-to-Image Generation', the transformer paper behind DALL·E.
-
February 2021
Learning from natural language supervision
A February 2021 paper, 'Learning Transferable Visual Models From Natural Language Supervision', lists him as a co-author.
-
March 2021
Research lead on multimodal neurons
He serves as research lead on 'Multimodal Neurons in Artificial Neural Networks', a Distill article published in March 2021.
-
As of October 2026
Machine learning research at OpenAI
He is a machine learning researcher at OpenAI, with interests in interpretability and convex optimization.
Sources and credits
Photo: We have no published photo of Gabriel Goh yet. Send us a free-licensed photo.