A Novel Metric for Measuring the Robustness of Large Language Models in Non-adversarial Scenarios

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ackerman, Samuel, Rabinovich, Ella, Farchi, Eitan, Anaby-Tavor, Ateret
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917827957489664
author Ackerman, Samuel
Rabinovich, Ella
Farchi, Eitan
Anaby-Tavor, Ateret
author_facet Ackerman, Samuel
Rabinovich, Ella
Farchi, Eitan
Anaby-Tavor, Ateret
contents We evaluate the robustness of several large language models on multiple datasets. Robustness here refers to the relative insensitivity of the model's answers to meaning-preserving variants of their input. Benchmark datasets are constructed by introducing naturally-occurring, non-malicious perturbations, or by generating semantically equivalent paraphrases of input questions or statements. We further propose a novel metric for assessing a model robustness, and demonstrate its benefits in the non-adversarial scenario by empirical evaluation of several models on the created datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2408_01963
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Novel Metric for Measuring the Robustness of Large Language Models in Non-adversarial Scenarios
Ackerman, Samuel
Rabinovich, Ella
Farchi, Eitan
Anaby-Tavor, Ateret
Computation and Language
Applications
We evaluate the robustness of several large language models on multiple datasets. Robustness here refers to the relative insensitivity of the model's answers to meaning-preserving variants of their input. Benchmark datasets are constructed by introducing naturally-occurring, non-malicious perturbations, or by generating semantically equivalent paraphrases of input questions or statements. We further propose a novel metric for assessing a model robustness, and demonstrate its benefits in the non-adversarial scenario by empirical evaluation of several models on the created datasets.
title A Novel Metric for Measuring the Robustness of Large Language Models in Non-adversarial Scenarios
topic Computation and Language
Applications
url https://arxiv.org/abs/2408.01963