Banded Multi-Test Functions and Adaptive Cover Certificate Families

Research Attribution and Autonomous Boundaries

The technical notes of this package draw a line right at the beginning, quoted as is — from this node onward, the decision-maker for the research architecture is the AI research side, not directly specified by Neo.

"This node is where the AI research side autonomously determines the mathematical architecture, experimental design, failure criteria, and next steps after the research handoff. The role of Neo.K / EveMissLab is to provide the existing research environment, prerequisite materials, authorization to continue research, and future review positions... This distinction is not an attribution decoration, but part of the chain of evidence." — Excerpt from §1 of the technical notes in this package, with the technical research lead field marked as OpenAI Codex.

Connections

This package is the first new node after the v1.0 integration overview handoff, responding to the "next main research node" proposed in §12 of that package, but shifting to a banded multi-certificate approach rather than the originally planned single global function.

Technical Notes

Main Document — Research attribution principles, from single-window failure to certificate families, why one cannot add first and then square, 18-patch adaptive cover design, coarse cover comparison, and next node recommendations.

Result Data

Coarse cover vs. adaptive cover comparison table, and global margin failure range, from RESULTS.md within the package.

Trust Boundaries

From TRUST_BOUNDARY.md within the package.

Files

The 18 individual rectangular certificates (outputs/certificates/*.certificate.json) and the candidate function library (candidate_library.json, 130 KB) are not listed individually; they are all included in the complete zip file.

sha256 0a04e5ef056204326d6a9100845f3e54ef7d1d05acbe2e46e58b304ba29f6743