We consider the problem of red teaming LLMs on elementary calculations and algebraic tasks to evaluate how various prompting techniques affect the quality of outputs. We present a framework to procedurally generate numerical questions and puzzles, and compare the results with and without the application of several red teaming techniques. Our findings suggest that even though structured reasoning and providing worked-out examples slow down the deterioration of the quality of answers, the gpt-3.5-turbo and gpt-4 models are not well suited for elementary calculations and reasoning tasks, also when being red teamed.
@article{arxiv.2401.00290,
title = {Red Teaming for Large Language Models At Scale: Tackling Hallucinations on Mathematics Tasks},
author = {Aleksander Buszydlik and Karol Dobiczek and Michał Teodor Okoń and Konrad Skublicki and Philip Lippmann and Jie Yang},
journal= {arXiv preprint arXiv:2401.00290},
year = {2024}
}
Comments
Accepted to The ART of Safety: Workshop on Adversarial testing and Red-Teaming for generative AI (IJCNLP-AACL 2023)