# Conceptor Debiasing of Word Representations Evaluated on WEAT

## TL;DR

This paper introduces a conceptor-based debiasing method for word embeddings that effectively reduces racial and gender biases, utilizing heterogeneous biased word lists and evaluated using the WEAT metric.

## Contribution

The paper presents a novel conceptor debiasing approach that can be applied post-processing to both traditional and contextualized embeddings, improving bias mitigation.

## Key findings

- Conceptor debiasing reduces racial and gender biases in word embeddings.
- The method effectively uses heterogeneous biased word lists.
- Bias reduction is validated using the WEAT metric.

## Abstract

Bias in word embeddings such as Word2Vec has been widely investigated, and many efforts made to remove such bias. We show how to use conceptors debiasing to post-process both traditional and contextualized word embeddings. Our conceptor debiasing can simultaneously remove racial and gender biases and, unlike standard debiasing methods, can make effect use of heterogeneous lists of biased words. We show that conceptor debiasing diminishes racial and gender bias of word representations as measured using the Word Embedding Association Test (WEAT) of Caliskan et al. (2017).