A comprehensive taxonomy of hallucinations in Large Language Models

TL;DR

This paper presents a detailed taxonomy of hallucinations in large language models, analyzing their types, causes, and implications, and discusses evaluation and mitigation strategies for responsible deployment.

Contribution

It offers the first comprehensive taxonomy of LLM hallucinations, including definitions, causes, manifestations, and mitigation approaches, highlighting their theoretical inevitability.

Findings

Hallucinations are categorized into intrinsic and extrinsic types.

Evaluation benchmarks and metrics for hallucination detection are surveyed.

Mitigation strategies include architectural and systemic approaches.

Abstract

Large language models (LLMs) have revolutionized natural language processing, yet their propensity for hallucination, generating plausible but factually incorrect or fabricated content, remains a critical challenge. This report provides a comprehensive taxonomy of LLM hallucinations, beginning with a formal definition and a theoretical framework that posits its inherent inevitability in computable LLMs, irrespective of architecture or training. It explores core distinctions, differentiating between intrinsic (contradicting input context) and extrinsic (inconsistent with training data or reality), as well as factuality (absolute correctness) and faithfulness (adherence to input). The report then details specific manifestations, including factual errors, contextual and logical inconsistencies, temporal disorientation, ethical violations, and task-specific hallucinations across domains like code generation and multimodal applications. It analyzes the underlying causes, categorizing them into data-related issues, model-related factors, and prompt-related influences. Furthermore, the report examines cognitive and human factors influencing hallucination perception, surveys evaluation benchmarks and metrics for detection, and outlines architectural and systemic mitigation strategies. Finally, it introduces web-based resources for monitoring LLM releases and performance. This report underscores the complex, multifaceted nature of LLM hallucinations and emphasizes that, given their theoretical inevitability, future efforts must focus on robust detection, mitigation, and continuous human oversight for responsible and reliable deployment in critical applications.