Back to paper
Full reasoning
May 21, 2026
cs.AI
Share this scorecard
Every score comes with the model's reasoning on the paper page.
Suggested post
"TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks" scored 7.0/10 predicted impact on @kurateorg — Clarity 7.5 · Significance 7.0 · Novelty 7.0