GERO public research archive: seven numerical reports by Xamit Kadirbekov

Archival collection of seven already public GERO numerical research reports by Xamit Kadirbekov. This upload preserves their original evidence archives byte-for-byte; it is not a new experiment or a claim that every proposed fix was accepted. Original reports and manifests identify the versions tested, limitations, AI assistance, and dated upstream status. Later project status may differ.

Included reports:
1. actuarialmath: the variance selector rejected before its own branch
https://www.gero.uz/research/articles/actuarialmath-uniform-variance-selector.html

2. bond_pricing: perpetuity-rate inversion fixed in release 1.0.1
https://www.gero.uz/research/articles/bond-pricing-perpetuity-rate-release.html

3. Samsung ONE: Gather slices and Mean axes changed by compiler passes
https://www.gero.uz/research/articles/samsung-one-gather-mean-axis.html

4. Empyrical: constant returns and floating-point Sharpe ratios
https://www.gero.uz/research/articles/empyrical-constant-return-dispersion.html

5. Riskfolio BOP: correcting the shrinkage target denominator
https://www.gero.uz/research/articles/riskfolio-bop-target-normalization.html

6. PyPortfolioOpt: a ticker-order mismatch breaks the allocation budget
https://www.gero.uz/research/articles/pypfopt-lp-ticker-alignment.html

7. actuarialmath direct endowment second moments omit the benefit power
https://www.gero.uz/research/articles/actuarialmath-direct-endowment-second-moment.html

Each report remains attributed exactly as in its original publication. The collection README and manifest are licensed CC BY 4.0. Included third-party source code, dependencies and original evidence archives retain their own licenses and notices; the compilation license does not override them. SHA-256 checksums and source URLs are included for reproducibility.

Compiled 26 October 2026. No private client data or unpublished medical findings are included.