Value of Information: A Framework for Human-Agent Communication

TL;DR

This paper introduces a decision-theoretic framework based on Value of Information for human-agent communication, allowing dynamic, context-aware decision-making that balances utility gains and user effort without hyperparameter tuning.

Contribution

It presents a novel, parameter-free VoI-based approach for adaptive agent communication that outperforms traditional confidence threshold methods across diverse real-world tasks.

Findings

VoI-based method matches or exceeds manually-tuned baselines

Achieves up to 1.36 utility points improvement in high-cost scenarios

Works seamlessly across varied domains from games to medical diagnosis

Abstract

Large Language Model (LLM) agents deployed for real-world tasks face a fundamental dilemma: user requests are underspecified, yet agents must decide whether to act on incomplete information or interrupt users for clarification. Existing approaches either rely on brittle confidence thresholds that require task-specific tuning, or fail to account for the varying stakes of different decisions. We introduce a decision-theoretic framework that resolves this trade-off through the Value of Information (VoI), enabling agents to dynamically weigh the expected utility gain from asking questions against the cognitive cost imposed on users. Our inference-time method requires no hyperparameter tuning and adapts seamlessly across contexts-from casual games to medical diagnosis. Experiments across four diverse domains (20 Questions, medical diagnosis, flight booking, and e-commerce) show that VoI consistently matches or exceeds the best manually-tuned baselines, achieving up to 1.36 utility points higher in high-cost settings. This work provides a parameter-free framework for adaptive agent communication that explicitly balances task risk, query ambiguity, and user effort.