REVIEW 3 cited by
Black is to Criminal as Caucasian is to Police: Detecting and Removing Multiclass Bias in Word Embeddings
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Online texts -- across genres, registers, domains, and styles -- are riddled with human stereotypes, expressed in overt or subtle ways. Word embeddings, trained on these texts, perpetuate and amplify these stereotypes, and propagate biases to machine learning models that use word embeddings as features. In this work, we propose a method to debias word embeddings in multiclass settings such as race and religion, extending the work of (Bolukbasi et al., 2016) from the binary setting, such as binary gender. Next, we propose a novel methodology for the evaluation of multiclass debiasing. We demonstrate that our multiclass debiasing is robust and maintains the efficacy in standard NLP tasks.
Forward citations
Cited by 3 Pith papers
-
VLBiasBench: A Comprehensive Benchmark for Evaluating Bias in Large Vision-Language Model
VLBiasBench is a new large-scale benchmark with 128,342 samples covering nine social bias categories plus two intersectional ones to evaluate biases in LVLMs.
-
CaTE Data Curation for Trustworthy AI
A synthesis of data curation practices for trustworthy AI, framed around an actionable definition of trustworthiness and a decision tree.
-
Multimodal Political Bias Identification and Neutralization
A proposed multimodal pipeline to identify and reduce political bias in news text and images remains unvalidated: the report presents architecture and qualitative examples, without quantitative results for most components.
Discussion (0). Continue with ORCID to comment.