Every score comes with the model's reasoning on the paper page.
"Towards a General Intelligence and Interface for Wearable Health Data" scored 8.0/10 predicted impact on @kurateorg — Significance 8.5 · Clarity 8.0 · Rigor 7.5