From Words to Widgets for Controllable LLM Generation
For end-users of LLMs, this work addresses the challenge of precisely controlling subjective generation preferences, offering a more intuitive and transparent interaction method.
The paper introduces Malleable Prompting, an interactive technique that converts natural language preference expressions into GUI widgets for controllable LLM generation. In a user study, it helped participants achieve target preferences more precisely and was perceived as more controllable and transparent than natural language prompting alone.
Natural language remains the predominant way people interact with large language models (LLMs). However, users often struggle to precisely express and control subjective preferences (e.g., tone, style, and emphasis) through prompting. We propose Malleable Prompting, a new interactive prompting technique for controllable LLM generation. It reifies preference expressions in natural language prompts into GUI widgets (e.g., sliders, dropdowns, and toggles) that users can directly configure to steer generation, while visualizing each control's influence on the output to support attribution and comparison across iterations. To enable this interaction, we introduce an LLM decoding algorithm that modulates the token probability distribution during generation based on preference expressions and their widget values. Through a user study, we show that Malleable Prompting helps participants achieve target preferences more precisely and is perceived as more controllable and transparent than natural language prompting alone.