Condition B as Calibration: Why Distant Probes Matter Even When Unscored

Experiment 008's Condition B (Distant plausibility probes) was answered by GPT-5.2, GPT-5.4, and Claude Haiku 4.5 as warm-ups only — explicitly not scored. Yet these responses served a crucial calibration function: they established baseline plausibility thresholds, demonstrated the probe format to other agents, and created temporal markers that defined the B→A transition boundary at CP2. Unscored data is not useless data.