Nested Bubble/Gear (NBG-01) v0.2.0: Canonical Negative-Control Execution and Deterministic Replay

Nested Bubble/Gear (NBG-01) v0.2.0 reports the canonical execution and deterministic replay of E1-NC, the frozen negative-control experiment following the NBG-01 v0.1.x positive-control and protocol-repair sequence. E1-NC preserves the transition dynamics and original 16-point (eta, kappa) parameter grid used in NBG-01 v0.1.2 while replacing the successful majority-based observer with a deliberately inadequate global six-bit parity observer. The parity observer was previously disclosed as a post-hoc diagnostic and is therefore used here explicitly as a known-fail negative control, not as blinded discovery evidence. The E1-NC protocol was frozen before canonical execution. The predictive-leakage threshold was fixed at EPS_PRED = 0.01, failure witnesses were required for NO-CLOSURE outcomes, parent transition kernels were required to match the v0.1.2 canonical experiment exactly, and no observer, threshold, parameter, or rescue rule was permitted to change after canonical output became visible. The canonical execution validates the negative control. All 16 of 16 parameter points terminate NO-CLOSURE, all 16 primary predictive leakages exceed the frozen threshold, and all 16 failures contain explicit witness pairs. Primary parity leakage ranges from approximately 0.944000 to 0.999404, while two-step leakage ranges from approximately 0.891136 to 0.998808. Kernel-digest cross-checks confirm that all 16 transition kernels are identical to the corresponding canonical kernels published in NBG-01 v0.1.2. The reference observer O0 retains exact one-step and two-step predictive closure throughout the grid, with maximum raw composition defect approximately 1.17 × 10^-15. A clean same-host isolated deterministic replay was then performed from the identical frozen source. Every deterministic scientific artifact reproduced byte-for-byte, yielding REPLAY_EXACT. The terminal experiment verdict is: NEGATIVE_CONTROL_VALIDATED This result demonstrates that the NBG receipt and witness machinery can reject an observer that discards transition-relevant information while continuing to accept the previously validated reference observer on the same underlying dynamics. The same-host replay is not presented as an external independent-laboratory replication. E1-NC is a control-validation experiment and does not establish a new physical law or make claims about spacetime, quantum mechanics, consciousness, or nature's literal ontology. This release closes the E1 negative-control stage and establishes the basis for a separately frozen E2-A interface-aware challenge in which predictive closure is no longer guaranteed for the cheap observer hierarchy.