Alignment reduces expressed gender bias in LLM outputs but does not remove the underlying encoded gender associations in internal representations.
In: ICML (2024)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Alignment Reduces Expressed but Not Encoded Gender Bias: A Unified Framework and Study
Alignment reduces expressed gender bias in LLM outputs but does not remove the underlying encoded gender associations in internal representations.