CL-25 Phase 1 — classical teacher rollouts

← CL-25 home  ·  benchmark previews
Every card below is a scripted (classical) teacher rollout, not a random agent — this is whether the task is solved, not just whether the scene renders.
Competence bar: SR ≥ 0.90, measured at n = 128. A card with no SR is not a result — every number here carries its n and the HEAD it was measured on.
Loading task results…