About this document

The Troubling Emergence of Hallucination in Large Language Models - An Extensive Definition, Quantification, and Prescriptive Remediations by PRIYANKA is a document available to read on EtoBox.

The document discusses the emergence of hallucinations in Large Language Models (LLMs), proposing a Hallucination Vulnerability Index (HVI) to assess and quantify the susceptibility of different LLMs to generating inaccurate information. It categorizes hallucinations into two main orientations: factual mirage and silver lining, further divided into intrinsic and extrinsic types, and outlines six specific categories of hallucination. The authors also introduce a dataset and mitigation strategies to address t

Author

PRIYANKA

Language

EN