What Prompts Don't Say: Understanding and Managing Underspecification in LLM Prompts

TL;DR

This paper analyzes prompt underspecification in LLMs, revealing its fragility and proposing requirements-aware optimization and systematic management to improve reliability.

Contribution

It introduces requirements-aware prompt optimization methods and advocates for proactive discovery and monitoring of requirements to address underspecification issues.

Findings

LLMs infer unspecified requirements 41.1% of the time

Underspecified prompts are twice as likely to regress with model or prompt changes

Requirements-aware optimization improves performance by 4.8% on average

Abstract

Prompt underspecification is a common challenge when interacting with LLMs. In this paper, we present an in-depth analysis of this problem, showing that while LLMs can often infer unspecified requirements by default (41.1%), such behavior is fragile: Under-specified prompts are 2x as likely to regress across model or prompt changes, sometimes with accuracy drops exceeding 20%. This instability makes it difficult to reliably build LLM applications. Moreover, simply specifying all requirements does not consistently help, as models have limited instruction-following ability and requirements can conflict. Standard prompt optimizers likewise provide little benefit. To address these issues, we propose requirements-aware prompt optimization mechanisms that improve performance by 4.8% on average over baselines. We further advocate for a systematic process of proactive requirements discovery, evaluation, and monitoring to better manage prompt underspecification in practice.