Back to paper
Full reasoning
Oct 19, 2026
cs.LG · cs.CL
Share this scorecard
Every score comes with the model's reasoning on the paper page.
Suggested post
"Strength Controls Change How Interventions Compare Across Sparse Autoencoder Codes" scored 4.0/10 predicted impact on @kurateorg — Reproducible 7.5 · Rigor 6.5 · Difficulty 6.0