Back to paper
Full reasoning
May 21, 2026
cs.AI · cs.CV
Share this scorecard
Every score comes with the model's reasoning on the paper page.
Suggested post
"ST-SimDiff: Balancing Spatiotemporal Similarity and Difference for Efficient Video Understanding with MLLMs" scored 7.0/10 predicted impact on @kurateorg — Clarity 7.5 · Novelty 7.0 · Significance 6.5