VLMs encode female associations internally for ambiguous images of female-stereotyped occupations but output male due to asymmetric layer-wise suppression, revealed by the new LALS metric across 15 occupations and four models.
Hila Gonen and Yoav Goldberg
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Vision-Language Models Suppress Female Representations Under Ambiguous Input
VLMs encode female associations internally for ambiguous images of female-stereotyped occupations but output male due to asymmetric layer-wise suppression, revealed by the new LALS metric across 15 occupations and four models.