arXiv - CSCL: "Contrastive Language-Vision AI Models Pretrained …" - Qoto Mastodon

arXiv - CSCL @arxiv_cscl@qoto.org

Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias. (arXiv:2212.11261v1 [cs.CY])

http://arxiv.org/abs/2212.11261 #arXiv #NLProc

Dec 22, 2022, 03:07 · · arxiv-cscl · · ·

Sign in to participate in the conversation