Guardado en:
Detalles Bibliográficos
Autores principales: Murray, Malcolm, Papadatos, Henry, Quarks, Otter, Gimenez, Pierre-François, Campos, Simeon
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2503.04299
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929750513025024
author Murray, Malcolm
Papadatos, Henry
Quarks, Otter
Gimenez, Pierre-François
Campos, Simeon
author_facet Murray, Malcolm
Papadatos, Henry
Quarks, Otter
Gimenez, Pierre-François
Campos, Simeon
contents The literature and multiple experts point to many potential risks from large language models (LLMs), but there are still very few direct measurements of the actual harms posed. AI risk assessment has so far focused on measuring the models' capabilities, but the capabilities of models are only indicators of risk, not measures of risk. Better modeling and quantification of AI risk scenarios can help bridge this disconnect and link the capabilities of LLMs to tangible real-world harm. This paper makes an early contribution to this field by demonstrating how existing AI benchmarks can be used to facilitate the creation of risk estimates. We describe the results of a pilot study in which experts use information from Cybench, an AI benchmark, to generate probability estimates. We show that the methodology seems promising for this purpose, while noting improvements that can be made to further strengthen its application in quantitative AI risk assessment.
format Preprint
id arxiv_https___arxiv_org_abs_2503_04299
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mapping AI Benchmark Data to Quantitative Risk Estimates Through Expert Elicitation
Murray, Malcolm
Papadatos, Henry
Quarks, Otter
Gimenez, Pierre-François
Campos, Simeon
Artificial Intelligence
The literature and multiple experts point to many potential risks from large language models (LLMs), but there are still very few direct measurements of the actual harms posed. AI risk assessment has so far focused on measuring the models' capabilities, but the capabilities of models are only indicators of risk, not measures of risk. Better modeling and quantification of AI risk scenarios can help bridge this disconnect and link the capabilities of LLMs to tangible real-world harm. This paper makes an early contribution to this field by demonstrating how existing AI benchmarks can be used to facilitate the creation of risk estimates. We describe the results of a pilot study in which experts use information from Cybench, an AI benchmark, to generate probability estimates. We show that the methodology seems promising for this purpose, while noting improvements that can be made to further strengthen its application in quantitative AI risk assessment.
title Mapping AI Benchmark Data to Quantitative Risk Estimates Through Expert Elicitation
topic Artificial Intelligence
url https://arxiv.org/abs/2503.04299