Aitor Arronte Alvarez

LLMs show high-confidence social bias in tutoring roles

Researchers have identified significant social biases in large language models (LLMs) used for conversational tutoring. These models often exhibit high confidence in their assessments, even when those assessments are bi…