← All proof

Strategic Cognition Benchmark — leaderboard

205 tasks across 5 dimensions of strategic cognition, grounded in ICD 203 analytic tradecraft. Sorted by SCI-H — the hardened index that uses non-compensatory gating on judgment-core dimensions to prevent narrow optimization from masking analytical weakness.

Loading leaderboard…

Three tracks

  • SCB-Core — Public 25-task starter pack. Anyone can submit.
  • SCB-Live — Rotating live tasks with Brier-resolved Type A questions.
  • SCB-Hidden — Contamination-resistant held-out task set.

Want this for your team?

Drop your email and we'll be in touch when access opens for your vertical.