Introducing Generative Interfaces - a new paradigm beyond chatbots. We generate interfaces on the fly to better facilitate LLM interaction, so no more passive reading of long text blocks. Adaptive and Interactive: creates the form that best adapts to your goals and needs!

Website: Paper: Code:

To achieve this, user queries are first mapped into structured intermediate representations and then decoded into UI code. An iterative refinement process applies adaptive reward functions to optimize the generated interfaces.

We ask human annotators to compare our generative interfaces (GenUI) with conversational interfaces (ConvUI) and interfaces generated through Claude artifacts (Instructed UI). GenUI wins a LOT.

Human preferences vary across different domains, suggesting a balance between generative interfaces and conversational ones to optimize user experience.

Joint work with @CHEN_JIAQI_00 @zhangyt0704 @EchoShao8899 @Diyi_Yang !

i'm trying to understand what's the contribution of the paper. is it just that humans prefer GUIs over text, or do you also have any new way of improving GUI generation

This video is fantastic! I hope I could make such a wonderful short video that clearly demonstrates the contributions of my work. May I know what tools did you use for making this video? Generative UI? When presenting Finite State Machine, how to make the animation effect?

kind of a longwinded way to describe a coding agent. although we appreciate the preference studies