debiaswe
by tolga-b
Remove problematic gender bias from word embeddings.
AI summary
Bias remover
Removes problematic gender bias from word embeddings to create more neutral representations.
- stars
- 244
- forks
- 90
- watching
- 8
Similar projects
Found by comparing what the projects do, not just their names.
Image captioning bias study
An investigation into bias in image captioning systems using a dataset and a new model design to mitigate this bias
Debugger library
A collection of useful Ember.js debugging functions.
Attribute remover
Enables better element selectors in Ember.js tests by removing unwanted attributes
Model debiasing
Debiasing techniques to minimize hallucinations in large visual language models
Bias test sentences
A dataset of manually crafted sentence templates to evaluate the presence of gender bias in natural language processing systems
Data decontamination tool
Deconvolutes single-cell RNA-seq data from potential doublets to remove contamination
Code cleaner
Removes console statements from JavaScript code to strip debugging information
Bias evaluator
This project provides a data proxy to evaluate bias in facial recognition systems across demographic groups.
Domain adaptation library
An implementation of a federated learning method to reduce domain bias in representation space, enabling improved knowledge transfer between clients and servers.
Word embedding trainer
Improves word embeddings by training with adversarial objectives
Whitespace cleaner
Removes unnecessary whitespace from Handlebars templates to improve template performance and reduce file size.
Stereotype analysis
Quantifies 100 years of gender and ethnic stereotypes through word embeddings.
Whitespace cleaner
A package that removes whitespace from strings and detects blank strings.
Vector updater
A tool to update word vectors to align with linguistic constraints
Word embedding model
A PyTorch implementation of the skip-gram model for learning word embeddings.
