| # v0.2.2 — ScenarioView report and financial validation | |
| Documentation/report release dated 2026-10-07. Model weights, tokenizer and loader are unchanged from v0.2.0; this is NOT an improved-accuracy model release. | |
| - Publishes a mobile-friendly ScenarioView report using the previous local report layout styles and section order. | |
| - Includes public macro observations with observation dates and source attribution, RT definitions, aggregate action metrics, and section-level provenance. | |
| - Frozen MiDM action accuracy: 54.37%; B3+MiDM stacking: 50.93%; development-selected horizon/position residual: 47.49%. No candidate promoted. | |
| - Evaluation: KOSPI, 2024-01-02 to 2026-08-20, 756 hypothetical decisions / 129 sampled dates, previously seen retrospective data; not independent trades or portfolio returns. | |
| - No validated economic leading/coincident/lagging status. New prompt context packs are prepared locally but have not been inference-tested. | |
| - No private source prices, per-date predictions, generated private narratives, checkpoint files or credentials published. | |
| - The report is issued today but does not contain a freshly inferred today-market signal. No automatic daily schedule is configured. | |
| Report: https://yeohoonyun.github.io/midm-decision-models/ | |
| The prior Zenodo DOI continues to refer to v0.2.1; no new Zenodo deposit is claimed. | |