FE-012C Repeatability Run 001 Pre-Registration

Purpose

Record expectations and failure signals before collecting repeatability-run outputs.

Predictions

• Entry primitive agreement should remain high.
• Reasoning shape agreement should remain high.
• Exit primitive and dominant primitive may show more drift.
• Transition agreement should remain moderate to strong.
• Recognition bias will likely remain a threat.
• Candidate missing primitives should remain rare if vocabulary is stable.

Failure Signals

• Same model produces substantially different primitive sequences from the prior run.
• Entry primitive agreement drops sharply.
• Reasoning shape agreement drops sharply.
• Candidate missing primitives increase materially.
• Transition agreement becomes weak across multiple packets.

Interpretation Boundary

These predictions are provisional expectations for measurement repeatability.

They are not claims that repeatability has already been demonstrated.