[Key Takeaway] In open mate-selection markets, women—unlike men—systematically enact hypergamous preferences with biological consistency, decisively filtering out prospective partners with lower educational attainment than their own.
-
Who Contacts Whom? Educational Homophily in Online Mate Selection
-
Age and gender diversity as determinants of performance and health in a public organization: the role of task complexity and group size
[Key Takeaway] Organizations with higher proportions of women demonstrate lower objective performance and increased reports of adverse health effects, with this dysfunction becoming more pronounced as group size increases.
-
The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs
[Key Takeaway] Despite accurately recognizing factual information, modern LLMs have their factual outputs overridden and censored by political correctness-driven safety alignment at rates of up to 18.9%.
-
Automatic Gender Identification from Text
[Key Takeaway] Even in creative texts featuring deliberate persona-taking, machine learning models can identify the writer’s gender with over 90% accuracy.
-
Author gender identification from text
[Key Takeaway] Independent of topical content, decisive gender differences exist in language use such that male and female authors can be differentiated with 85.1% accuracy relying purely on stylistic choices and function word selection.