Sentient Agent as a Judge: Evaluating Higher-Order Social Cognition in Large Language Models

The paper introduces SAGE, an evaluation framework for assessing LLMs' social cognition through simulated emotional responses, revealing significant performance gaps among models in empathetic dialogue. https://arxiv.org/abs//2505.02847 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016 Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers