RESEARCH NOTES

Record what happened.
Keep what did not work.

The notes do not turn a changing research process into a completed story. Observations are separated from hypotheses and later interpretations.

4RESEARCH GENERATIONS
—CURATED RECORDS
30 × 4BALANCED SELECTION
01 / RECORD FORMAT

A result and an interpretation are not the same thing.

OBSERVED

What the run actually produced

Measurements, events, outputs, and provenance are recorded before a conclusion is written.

HYPOTHESIS

What the experiment expected

A hypothesis remains a proposal until the stated test and counterexample conditions hold.

INTERPRETATION

What the result may mean

Later understanding can be added without rewriting the original observation as foreknowledge.

  1. Research question
  2. Hypothesis
  3. Experimental setup
  4. Method
  5. Observed result
  6. Interpretation
  7. Failure and limitations
  8. Next question
02 / CATEGORIES

Different questions, one evidential standard.

01

Architecture

System boundaries, runtime invariants, and core design.

02

Memory & Experience

What is stored, what becomes reusable, and what still belongs to code.

03

World Experiments

Observation, external action, and reconciliation with actual outcomes.

04

Retrieval & Anchors

Search, natural-language surfaces, anchors, and experience reuse.

05

Experimental Results

Runs with measured conditions and reproducible artifacts.

06

Failures

Counterexamples, lost information, false bridges, and discarded structures.

03 / PUBLISHED NOTES

Detailed records grounded in the experiment archive.

Showing 30 records from Current MIRA.

Loading published research notes…

Thirty records are preserved for each generation. Zero, Original, and Canonical records are historical evidence; only Current MIRA describes the active research line.

04 / LINEAGE

Past code is evidence, not present authority.

  1. MIRA Zero

    Meaning accumulation, world residues, and relation experiments

    ARCHIVED
  2. MIRA Original

    ECT interference, temporal residue, recovery, and shadow branches

    ARCHIVED
  3. Canonical MIRA

    Minimal computational spine, invariant contracts, and bounded validation

    FROZEN
  4. Current MIRA

    Memory-first Living Runtime and experiment ledger

    ACTIVE