A comprehensive taxonomy of hallucinations in Large Language Models
Manuel Cossio
TL;DR
This work presents a comprehensive taxonomy of hallucinations in large language models, arguing that hallucinations are an inevitable byproduct of computable LLMs. It systematically classifies manifestations into intrinsic/extrinsic and factuality/faithfulness categories, then details a wide spectrum of factual, contextual, temporal, ethical, and task-specific hallucinations across domains such as code and multimodal tasks. The paper analyzes root causes across data, model, and prompt levels, integrates cognitive and human factors, and reviews evaluation benchmarks and mitigation strategies, including retrieval grounding, tool use, and guardrails, framed within a call for robust, human-in-the-loop safeguards. It also surveys web-based monitoring resources and advocates a hybrid, context-aware approach to mitigation and deployment to enable safer, more reliable LLM use in high-stakes settings.
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.