Uncertainty Propagation in LLM-Based Systems

2026-04-25

AI Summary

This study addresses the propagation and accumulation of uncertainty within large language model (LLM) systems across model internals, system components, and human–AI interactions. It proposes, for the first time, a three-tiered classification framework encompassing intra-model (P1), system-level (P2), and sociotechnical (P3) dimensions to systematically analyze uncertainty propagation mechanisms. Through conceptual modeling, systems analysis, and taxonomic construction, the work distills cross-domain engineering principles and identifies five critical open challenges. By elucidating how uncertainty permeates LLM-based systems at multiple levels, this research establishes a theoretical foundation and outlines key directions for the design of trustworthy AI systems.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Reasoning under Uncertainty: Uncertainty Representations
• Multiagent Systems: Multiagent Systems under Uncertainty

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Uncertainty in large language model (LLM)-based systems is often studied at the level of a single model output, yet deployed LLM applications are compound systems in which uncertainty is transformed and reused across model internals, workflow stages, component boundaries, persistent state, and human or organisational processes. Without principled treatment of how uncertainty is carried and reused across these boundaries, early errors can propagate and compound in ways that are difficult to detect and govern. This paper develops a systems-level account of uncertainty propagation. It introduces a conceptual framing for characterising propagated uncertainty signals, presents a structured taxonomy spanning intra-model (P1), system-level (P2), and socio-technical (P3) propagation mechanisms, synthesises cross-cutting engineering insights, and identifies five open research challenges.

Problem

• Uncertainty Propagation
• Large Language Models
• Compound Systems
• Socio-technical Systems
• Error Compounding

Innovation

• uncertainty propagation
• large language models
• systems-level analysis
• socio-technical systems
• structured taxonomy