Twelve LLMs show significant gender bias in contextual embeddings, with female names skewed toward warmth and male names toward competence on stereotype-content dimensions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Profiling Bias in LLMs: Stereotype Dimensions in Contextual Word Embeddings
Twelve LLMs show significant gender bias in contextual embeddings, with female names skewed toward warmth and male names toward competence on stereotype-content dimensions.