Counterfactual Logit Pairing

Posted by Bhaktipriya Radharapu, Software Engineer

TensorFlow Model Remediation is an open source toolkit that showcases solutions to help mitigate unfair bias in Machine Learning models. The toolkit offers resources to build fairer models for everyone – in line with Google’s AI Principles. Today, we’re excited to announce a new technique within the TensorFlow Model Remediation Library called Counterfactual Logit Pairing (CLP) to address unintended bias in ML models.

ML models are prone to making incorrect predictions when a …

api bias engineer google library line machine machine learning machine learning models making ml models open source perspective perspective api predictions resources software software engineer solutions tensorflow toolkit