ArticlePNAS nexus2026

Like humans, language models demonstrate face-to-character biases.

Steven A Lehr et al.PubMed ↗Full text ↗Publisher ↗

No numbers read from the abstract.