MultiQ&A: An Analysis in Measuring Robustness via Automated Crowdsourcing of Question Perturbations and Answers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cho, Nicole, Watson, William
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913681164468224
author Cho, Nicole
Watson, William
author_facet Cho, Nicole
Watson, William
contents One critical challenge in the institutional adoption journey of Large Language Models (LLMs) stems from their propensity to hallucinate in generated responses. To address this, we propose MultiQ&A, a systematic approach for evaluating the robustness and consistency of LLM-generated answers. We demonstrate MultiQ&A's ability to crowdsource question perturbations and their respective answers through independent LLM agents at scale. Our experiments culminated in the examination of 1.9 million question perturbations and 2.3 million answers. Furthermore, MultiQ&A shows that ensembled LLMs, such as gpt-3.5-turbo, remain relatively robust and consistent under perturbations. MultiQ&A provides clarity in the response generation space, offering an effective method for inspecting disagreements and variability. Therefore, our system offers a potential framework for institutional LLM adoption with the ability to measure confidence, consistency, and the quantification of hallucinations.
format Preprint
id arxiv_https___arxiv_org_abs_2502_03711
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MultiQ&A: An Analysis in Measuring Robustness via Automated Crowdsourcing of Question Perturbations and Answers
Cho, Nicole
Watson, William
Computation and Language
Artificial Intelligence
Machine Learning
One critical challenge in the institutional adoption journey of Large Language Models (LLMs) stems from their propensity to hallucinate in generated responses. To address this, we propose MultiQ&A, a systematic approach for evaluating the robustness and consistency of LLM-generated answers. We demonstrate MultiQ&A's ability to crowdsource question perturbations and their respective answers through independent LLM agents at scale. Our experiments culminated in the examination of 1.9 million question perturbations and 2.3 million answers. Furthermore, MultiQ&A shows that ensembled LLMs, such as gpt-3.5-turbo, remain relatively robust and consistent under perturbations. MultiQ&A provides clarity in the response generation space, offering an effective method for inspecting disagreements and variability. Therefore, our system offers a potential framework for institutional LLM adoption with the ability to measure confidence, consistency, and the quantification of hallucinations.
title MultiQ&A: An Analysis in Measuring Robustness via Automated Crowdsourcing of Question Perturbations and Answers
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2502.03711