2 papers
cs.CL2026
Auditing Proprietary Alignment in Large Language Models: A Comparative Framework Without a Ground-Truth Standard
Alireza Arbabi, Florian Kerschbaum
Large language models (LLMs) are increasingly released and deployed through opaque development and deployment pipelines, enabling model providers to inject intentional, provider-sp…
cs.CL2025
Relative Bias: A Comparative Framework for Quantifying Bias in LLMs
Alireza Arbabi, Florian Kerschbaum
The growing deployment of large language models (LLMs) has amplified concerns regarding their inherent biases, raising critical questions about their fairness, safety, and societal…