A Unified Definition of Hallucination: It's The World Model, Stupid!

TL;DR

This paper proposes a unified definition of hallucination in language models as inaccurate world modeling, aiming to clarify evaluation standards and facilitate the development of benchmarks to improve model reliability.

Contribution

It introduces a comprehensive framework unifying prior hallucination definitions based on world model inaccuracies, enabling clearer evaluation and comparison across benchmarks.

Findings

Unified hallucination definition based on world model inaccuracies

Framework distinguishes true hallucinations from other errors

Plans for benchmarks using synthetic reference models

Abstract

Despite numerous attempts at mitigation since the inception of language models, hallucinations remain a persistent problem even in today's frontier LLMs. Why is this? We review existing definitions of hallucination and fold them into a single, unified definition wherein prior definitions are subsumed. We argue that hallucination can be unified by defining it as simply inaccurate (internal) world modeling, in a form where it is observable to the user. For example, stating a fact which contradicts a knowledge base OR producing a summary which contradicts the source. By varying the reference world model and conflict policy, our framework unifies prior definitions. We argue that this unified view is useful because it forces evaluations to clarify their assumed reference "world", distinguishes true hallucinations from planning or reward errors, and provides a common language for comparison across benchmarks and discussion of mitigation strategies. Building on this definition, we outline plans for a family of benchmarks using synthetic, fully specified reference world models to stress-test and improve world modeling components.