Generative MUI
Summary
Generative MUI lets you use an LLM to build UIs that follow your own Material UI theme.
Similar Articles
@alphabatcher: 25 render tools can cost ~10,000 tokens before the user asks anything Generative UI has 3 build patterns Controlled > p…
Discusses three build patterns for generative UI (controlled, declarative, open-ended) and provides a decision tree for choosing the right approach based on use case.
From Words to Widgets for Controllable LLM Generation
Malleable Prompting is a novel interactive technique that reifies natural language preferences into GUI widgets (sliders, toggles, dropdowns) for direct manipulation, with a decoding algorithm that modulates token probabilities based on widget values to enable precise control over LLM generation. A user study shows it outperforms natural language prompting in precision, controllability, and transparency.
@Saboo_Shubham_: https://x.com/Saboo_Shubham_/status/2062220865643982875
The article explains three patterns of Generative UI (Controlled, Declarative, Open-ended) and how CopilotKit implements them via the AG-UI protocol for building dynamic agent-driven interfaces.
@tom_doerr: Generates UI with 67% fewer tokens than JSON https://github.com/thesysdev/openui…
OpenUI is an open-source framework for generative UI that uses a streaming-first language to produce UI with up to 67% fewer tokens than JSON, including a React runtime and chat interfaces.
Macaron-A2UI: A Model for Generative UI in Personal Agents
Presents Macaron-A2UI, a model for generative UI in personal agents that synthesizes dynamic interfaces with lightweight executable actions, moving beyond text-only chat. The paper introduces a large-scale corpus, the A2UI-Bench benchmark, and trains models up to 754B parameters using LoRA fine-tuning and reinforcement learning, achieving strong results.