FairMonitor: A Dual-framework for Detecting Stereotypes and Biases in Large Language Models

arXiv:2405.03098v1 Announce Type: new
Abstract: Detecting stereotypes and biases in Large Language Models (LLMs) is crucial for enhancing fairness and reducing adverse impacts on individuals or groups when these models are applied. Traditional methods, which rely on embedding spaces or are based on probability metrics, fall short in revealing the nuanced and implicit biases present in various contexts. To address this challenge, we propose the FairMonitor framework and adopt a static-dynamic detection method for a comprehensive evaluation of stereotypes and …

abstract arxiv biases cs.cl embedding fairness framework impacts language language models large language large language models llms metrics probability spaces stereotypes type