From Words to Widgets for Controllable LLM Generation
Chao Zhang, Yiren Liu, Lunyiu Nie, Jeffrey M. Rzeszotarski, Yun Huang + 1 more
TLDR
Malleable Prompting converts natural language preferences into GUI widgets for precise, interactive control over LLM generation.
Key contributions
- Introduces Malleable Prompting, an interactive technique for controllable LLM generation.
- Reifies natural language preference expressions into configurable GUI widgets for steering LLMs.
- Uses an LLM decoding algorithm to modulate token probabilities based on preference widget values.
- User study confirms Malleable Prompting improves precision, control, and transparency over text prompts.
Why it matters
This paper addresses a key challenge in LLM interaction: precisely controlling subjective generation preferences. By introducing interactive widgets, it offers a more intuitive and effective way for users to steer LLMs. This approach enhances user experience and broadens the practical application of LLMs.
Original Abstract
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.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.