I'm Jay Wang, a safety resaercher at OpenAI. I design, code, research, and write.

My work focuses on making AI more accessible, interpretable, and accountable, by designing and developing novel Interactive Interfaces for people to easily and enjoyably interact with Machine Learning systems at scale.

I love open source. I lead and maintain the following open-source research projects on GitHub. ⬇️

Please reach out if you have any questions or want to chat!

MeMemo: Vector database and vector search in your browser! (SIGIR'24) 62	Farsight: In situ intelligent widgets for responsible AI! (CHI'24) 18	SuperNOVA: Design strategies for notebook visualization tools (CHI'24) 63
Wordflow: Social and customizable AI writing assistant! (ACL'24) 186	WizMap: Explore and interpret millions of embeddings in your browser! (ACL'23) 419	DiffusionDB: Large-scale prompt gallery dataset for image generative models (ACL'23) 1214
WebSHAP: Explain any ML models anywhere! (TheWebConf'23) 51	GAM Coach: Personal coach to help you obtain desired AI decisions! (CHI'23) 17	Angler: Prioritize ML model improvements with interactive visualizations (CHI'23) 66
TimberTrek: Curate decision trees that align with human values! (VIS'22) 152	NOVA: Easy way to create notebook-ready visual analytics tools (VIS Poster'22) 28	GAM Changer: Edit ML models to reflect human knowledge and values (KDD'22) 123
StickyLand: Break the linear presentation of computational notebooks! (CHI'22) 531	Dodrio: Explore transformer models with interactive visualization (ACL'21) 344	CNN Explainer: Learn about CNNs with interactive visualization (TVCG'20) 8090