ORNATE INTELLIGENCE LLC · INA-001 · PRELIMINARY EXPERIMENT · SEPTEMBER 26, 2026

INA-001: Selective Information Access

A first causal engineering test of one proposed INA function.

Question

Does a selective retrieval mechanism deliver predeclared relevant information substantially more often than a permission- and packet-size-matched random selector?

Recorded result

In the preserved synthetic #26001 benchmark, the recorded selector covered required evidence in 46 of 50 answerable cases (92%). Under the exact matched-random null model used in the analysis, the expected success count is approximately 2.24 of 50.

What this supports

The result supports the narrow engineering claim that selective information access is functioning as a measurable, non-random component of the prototype. It provides a concrete way to manipulate and test one element of the INA framework.

What this does not establish

• It does not validate the full cross-substrate INA hypothesis.
• It does not show that human memory retrieval and electronic retrieval use the same mechanism.
• It does not demonstrate consciousness, subjective experience, or personhood.
• It is not independent validation; the benchmark and prototype were produced within the Ornate development process.

Next test

The planned INA-001B stage is a prospective ablation study comparing no persistent state, selectively retrieved state, full permitted state, and deliberately mismatched permitted state under frozen questions and scoring criteria.