Retrying vs Resampling in AI Control

We’ve just released a new paper: Retrying vs Resampling in AI Control. We revisit the resampling protocols introduced in Ctrl-Z with an up-to-date setting and much stronger models, and compare them against “retrying” protocols similar to Claude Code auto mode or Codex Auto-review.MotivationRoughly a year ago we released Ctrl-Z, the first paper to study control techniques for agents. A headline result of that paper was the performance of resample protocols – strategies that involve taking multipl...