Researcher
Chelsea Sierra Voss
Current role not listed
Voss helped develop DALL·E, which turns text descriptions into images,[1] and studied how neural networks represent concepts. She co-authored research showing that CLIP, an image-and-text model, recognizes the same ideas in pictures and written words,[2] and led the execution of GPT-4 evaluations.[3]
OpenAI
MITCitation Trend
Updates 1
Selected contributions
Zero-Shot Text-to-Image Generation
With Aditya Ramesh, Mikhail Pavlov, Gabriel Goh, Scott Gray, Alec Radford, Mark Chen and Ilya SutskeverIntroduced the original DALL·E, which generates images from written descriptions. It used one Transformer to learn from text and images together and could combine concepts in novel ways without retraining for each task. Published at ICML 2021.
References
- Zero-Shot Text-to-Image Generationarxiv.org
Voss co-authored the original DALL·E paper, which describes a model that generates images from text.
- Multimodal Neurons in Artificial Neural Networksdistill.pub
The study identifies CLIP neurons that respond to the same concept in images and text.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed August 28, 2026 · Checked for accuracy and new information.