Every score comes with the model's reasoning on the paper page.
"MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systems" scored 5.0/10 predicted impact on @kurateorg — Clarity 7.0 · Significance 6.0 · Novelty 5.5