Reproduce “EmplifAI: a Fine-grained Dataset for Japanese Empathetic Medical Dialogues in 28 Emotion Labels” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Wan Jou She, Lis Kanashiro Pereira, Fei Cheng, Sakiko Yahata, Panote Siriaraya, Eiji Aramaki
EmplifAI creates Japanese patient–supporter dialogues for chronic-care situations and 27 emotions plus neutral, using crowdsourced situations and two-turn exchanges. The authors use 4,125 dialogues across 280 situations to fine-tune Japanese models. For LLM-jp-3.1-13b-instruct4, mean empathy ratings rose from 1.00 to 2.47 and 2.40, but Gemini-2.5-Flash judged them, not patients, and clinical benefit remains untested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.