BBQ: A Hand-Built Bias Benchmark for Question Answering

It is well documented that NLP models learn social biases, but little work has been done on how these biases manifest in model outputs for applied tasks like question answering (QA). We introduce the Bias Benchmark for QA (BBQ), a dataset of question sets constructed by the authors that highlight at...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2022-03
Main Authors: Parrish, Alicia, Chen, Angelica, Nangia, Nikita, Padmakumar, Vishakh, Phang, Jason, Thompson, Jana, Phu Mon Htut, Bowman, Samuel R
Format: Article
Language:English
Subjects:
Benchmarks
Bias
Context
Human bias
Model testing
Questions
Social factors
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:It is well documented that NLP models learn social biases, but little work has been done on how these biases manifest in model outputs for applied tasks like question answering (QA). We introduce the Bias Benchmark for QA (BBQ), a dataset of question sets constructed by the authors that highlight attested social biases against people belonging to protected classes along nine social dimensions relevant for U.S. English-speaking contexts. Our task evaluates model responses at two levels: (i) given an under-informative context, we test how strongly responses reflect social biases, and (ii) given an adequately informative context, we test whether the model's biases override a correct answer choice. We find that models often rely on stereotypes when the context is under-informative, meaning the model's outputs consistently reproduce harmful biases in this setting. Though models are more accurate when the context provides an informative answer, they still rely on stereotypes and average up to 3.4 percentage points higher accuracy when the correct answer aligns with a social bias than when it conflicts, with this difference widening to over 5 points on examples targeting gender for most models tested.
ISSN:2331-8422