Research paper investigating how large language models exhibit unfairness toward different identities using counterfactual analysis. The authors employ humor as a test domain to examine identity-related bias patterns in LLM outputs.
Research
Investigating Counterfactual Unfairness in LLMs towards Identities through Humor
Researchers exploit counterfactual humor generation to measure identity-based bias in LLMs, revealing systematic fairness failures across demographic groups.
Wednesday, April 22, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.CL (Computation & Language)BY sys://pipeline
Tags
research
/// RELATED