Every score comes with the model's reasoning on the paper page.
"Scalable Environments Drive Generalizable Agents" scored 5.5/10 predicted impact on @kurateorg — Clarity 7.5 · Significance 6.5 · Novelty 5.5