Reproduce “EmpathicStories++: A Multimodal Dataset for Empathy towards Personal Experiences” 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

Jocelyn Shen, Yubin Kim, Mohit Hulse, Wazeer Zulfikar, Sharifa Alghowinem, Cynthia Breazeal, Hae Won Park

EmpathicStories++ introduces a longitudinal, multimodal dataset and benchmark for predicting a participant’s empathy toward a story. The authors recorded 41 participants across 269 in-home sessions over one month, collecting words, audio, video, and ratings. In Reflection, LF-LSTM with video and text achieved r = 0.560 ± 0.049 and ρ = 0.559 ± 0.049; these are benchmark correlations, not evidence of personalized empathy.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.