# Learning to Clarify Underspecified Intents Under Limited Interaction

arXiv: https://arxiv.org/abs/2610.04719 · HTML: https://mlsanity.com/paper/2610.04719/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 71 references that are in the map.
- First posted on arXiv 3 Oct 2026.

## Builds on (most cited first, 27 of 27)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [Asking Clarifying Questions in Open-Domain Information-Seeking Conversations](https://mlsanity.com/paper/1907.06554/) (2019)
- [DOCCI: Descriptions of Connected and Contrasting Images](https://mlsanity.com/paper/2404.19753/) (2024)
- [CollabLLM: From Passive Responders to Active Collaborators](https://mlsanity.com/paper/2502.00640/) (2025)
- [Learning to Ask Good Questions: Ranking Clarification Questions using Neural Expected Value of Perfect Information](https://mlsanity.com/paper/1805.04655/) (2018)
- [STaR-GATE: Teaching Language Models to Ask Clarifying Questions](https://mlsanity.com/paper/2403.19154/) (2024)
- [SOTOPIA-π: Interactive Learning of Socially Intelligent Language Agents](https://mlsanity.com/paper/2403.08715/) (2024)
- [Eliciting Human Preferences with Language Models](https://mlsanity.com/paper/2310.11589/) (2023)
- [Modeling Future Conversation Turns to Teach LLMs to Ask Clarifying Questions](https://mlsanity.com/paper/2410.13788/) (2024)
- [Tell Me More! Towards Implicit User Intention Understanding of Language Model Driven Agents](https://mlsanity.com/paper/2402.09205/) (2024)
- [PromptCharm: Text-to-Image Generation through Multi-modal Prompting and Refinement](https://mlsanity.com/paper/2403.04014/) (2024)
- [How to Build User Simulators to Train RL-based Dialog Systems](https://mlsanity.com/paper/1909.01388/) (2019)
- [Answer-based Adversarial Training for Generating Clarification Questions](https://mlsanity.com/paper/1904.02281/) (2019)
- [Prompting AI Art: An Investigation into the Creative Skill of Prompt Engineering](https://mlsanity.com/paper/2303.13534/) (2023)
- [Evaluating Mixed-initiative Conversational Search Systems via User Simulation](https://mlsanity.com/paper/2204.08046/) (2022)
- [Proactive Agents for Multi-Turn Text-to-Image Generation Under Uncertainty](https://mlsanity.com/paper/2412.06771/) (2024)
- [Adapting User Interfaces with Model-based Reinforcement Learning](https://mlsanity.com/paper/2103.06807/) (2021)
- [A Framework for Studying AI Agent Behavior: Evidence from Consumer Choice Experiments](https://mlsanity.com/paper/2509.25609/) (2025)
- [The Dearth of the Author in AI-Supported Writing](https://mlsanity.com/paper/2404.10289/) (2024)
- [Gemini Embedding 2: A Native Multimodal Embedding Model from Gemini](https://mlsanity.com/paper/2605.27295/) (2026)
- [What's Producible May Not Be Reachable: Measuring the Steerability of Generative Models](https://mlsanity.com/paper/2503.17482/) (2025)
- [MARLUI: Multi-Agent Reinforcement Learning for Adaptive UIs](https://mlsanity.com/paper/2209.12660/) (2022)
- [Probing Perceptual Priors of MLLMs via Gibbs Sampling with Interpretable Generative Controls](https://mlsanity.com/paper/2608.27727/) (2026)
- [Improving Information Extraction with Learned Queries](https://mlsanity.com/paper/2608.31058/) (2026)

## Abstract (arXiv)

AI assistants receive requests that leave out information needed for a good outcome, for example about users' preferences or goals. They must then either speculate or ask for more information before proceeding. We reconceptualize this as a value-of-information problem: the assistant should acquire information whose absence causes the greatest avoidable loss in user utility. This is rarely known ex ante; rather, assistants must predict it in order to optimally allocate limited user interactions. We instantiate this problem in image generation and derive a reinforcement learning framework using multi-turn simulated users to maximize utility recovery under uncertainty. In a preregistered study with 456 interactive sessions across 76 human participants, this helped users significantly better match reference images with significantly fewer questions, less total interaction time, and lower cost. This points toward a simple and scalable framework for training language model assistants to better disambiguate user intent by asking more informative questions.