Safety researcher @OpenAI 🤖 | ML PhD @GeorgiaTe 🐝 | Researching human-AI interaction 🔍 | Prev. intern @GoogleAI @Apple @MSFTResearch 🌱 | @UWMadison '19 🦡
San Francisco, CA
Joined December 2018
- Tweets387
- Following599
- Followers1.2K
- Likes748
Pinned Tweet
Today we're sharing updates that help ChatGPT better recognize context in sensitive conversations and respond safely in rare but important cases. We built these improvements with input from mental health experts, grounding the work in clinical expertise.
Jay Wang retweeted
GPT-5.6 has much better taste in frontend work, visualizations, and artifacts.
In the app, you’ll now see more visualizations and interactive tools inline, all generated in real time by 5.6. I’m a visual thinker, so seeing these instead of a wall of text is incredibly refreshing.
- Visualize data with any type of chart
- Learn a topic with interactive walkthroughs
- Explore 3D models
- Walk through steerable simulations
- Use mini-apps to get things done, like sorting through your inbox
Any of these can become a shareable Site with a single click.
Jay Wang retweeted
sharing out new work that helps ChatGPT better recognize context in sensitive conversations and respond safely in these complex/nuanced scenarios-- both within long conversations and across separate conversations!
see blog post for details: openai.com/index/chatgpt-rec…
Jay Wang retweeted
one of my top 3 most used internal tools
Jay Wang retweeted
I have found Euphony so useful internally! Glad it's now open source!
Try Euphony to visualize your chat data and codex sessions!
Jay Wang retweeted
#CHI2025 began this past weekend where alum Zijie (Jay) Wang received an Outstanding Dissertation Award! It is the latest accolade in a career of achievement for @Jay4w that you can learn more about in the 🔗👇!
Congratulations on the honor, Jay! 🥳🏆
cc.gatech.edu/news/thesis-hu…
ALT Photo of Georgia Tech alumnus Zijie (Jay) Wang (Ph.D. ML-CSE 2024)
Incredible honor to be recognized with ACM SIGCHI Outstanding Dissertation Award. Deepest thanks to all my collaborators and mentors—this wouldn’t have been possible without your support and guidance! ❤️
So honored to be named to @Forbes 30 Under 30 in Science this year!
Extremely grateful to my mentors, especially @PoloChau, and all my colleagues! ❤️
Thank you @jaemariesolyst for breaking the news to me 😆.
forbes.com/profile/zijie-jay…
ALT Zijie Jay Wang Forbes 30 under 30 profile
ALT Zijie Jay Wang Forbes 30 under 30 profile
Jay Wang retweeted
All AI systems make mistakes.
🧐 What if users could leverage AI flaws to understand it & take informed actions?
🚀 Our #CSCW2024 paper on Seamful XAI offers a process to foresee, locate, & leverage AI flaws—boosting user understanding & agency.
But why should you care?⤵️
1/n
Jay Wang retweeted
Such a great use case - When you need recommendation letters for a visa, job, or academic purposes, this tool - (RecRec )makes it easier by finding the most suitable writers who are already familiar with your work through paper citations - by @Jay4w
Need recommendation letters for your visa, job, or academic pursuits?
𝗥𝗲𝗰𝗥𝗲𝗰 makes it easy to find potential letter writers who already know your work—based on citations of your papers!
Save time and boost your chances!
Try it now 👉 xiaohk.github.io/recrec/
#PhDLife
RecRec is open-source and available at github.com/xiaohk/recrec.
Feature requests and feedback will be highly appreciated!
Shout out to @gneubig for the inspiration!! 🙌
nitter.cf/gneubig/status/1772989…
Researchers often have to ask for recommendation letters for visa/job applications, etc.
I wrote a script that allows you to find who cites your papers frequently to create a list of potential letter writers: github.com/neubig/research-c…
Hope it's helpful, improvements are welcome!
Jay Wang retweeted
Transformer Explainer
Interactive Learning of Text-Generative Models
discuss: huggingface.co/papers/2408.0…
Transformers have revolutionized machine learning, yet their inner workings remain opaque to many. We present Transformer Explainer, an interactive visualization tool designed for non-experts to learn about Transformers through the GPT-2 model. Our tool helps users understand complex Transformer concepts by integrating a model overview and enabling smooth transitions across abstraction levels of mathematical operations and model structures. It runs a live GPT-2 instance locally in the user's browser, empowering users to experiment with their own input and observe in real-time how the internal components and parameters of the Transformer work together to predict the next tokens. Our tool requires no installation or special hardware, broadening the public's education access to modern generative AI techniques.