Researchers have developed three generative models to understand how people differ in their judgments of various stimuli like faces, artworks, and jokes. By analyzing data from approximately 20,000 raters across ten public datasets, the models were able to reconstruct an individual's judgment space from a small number of ratings. The study found that while reweighting shared dimensions was sufficient for judging faces, judging artworks revealed a more complex spread of individual differences, with some raters holding opinions opposite to the population consensus. AI
RANK_REASON Academic paper published on arXiv detailing a new modeling approach for understanding human judgment. [lever_c_demoted from research: ic=1 ai=0.4]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →