作者位次 #4/6ACL 2026Findings· Findings of ACL
Audit Me If You Can: Query-Efficient Active Fairness Auditing of Black-Box LLM s
David Hartmann · Lena Pohlmann · Lelia Hanslik · Noah Gießing · Bettina Berendt · Pieter Delobelle
摘要
Large Language Models (LLMs) exhibit systematic biases across demographic groups. Auditing is proposed as an accountability tool for black-box LLM applications, but suffers from resource-intensive query access. We conceptualise auditing as uncertainty estimation over a target fairness metric and...