Dynamical Information Geometry, Part VII: Two-Dimensional Twin-Sign Validation under Fixed Audit and Regime Constraints

Preprint: This work is a preprint, has not undergone peer review, and is made publicly available to establish a public scientific record. This work reports an experimental two-dimensional validation of the Dynamical Information Geometry (DIG) Twin-Sign protocol under fixed audit and regime constraints. Dynamical Information Geometry assigns an effective, state-dependent information geometry to quantum many-body dynamics via the curvature of a coarse-grained mutual-information profile. Earlier work introduced the theoretical framework (Part I), provided numerical evidence for sign-sensitive refraction effects (Part II), validated a minimal operational consequence on quantum hardware (Part III), established regime structure and operational necessity (Part IV), and formulated the intrinsic tensorial nature of information curvature in dimensions greater than one (Part V). The present work occupies a strictly operational and consolidating role within the DIG series. It does not introduce new theoretical postulates, derivations, observables, or interpretive claims. Instead, it tests whether the multidimensional, projection-based structure articulated in DIG Part V remains internally consistent when subjected to the mandatory audit logic and regime language fixed in DIG Part IV. Twin-Sign tests are performed in two spatial dimensions along two non-collinear directions, without privileging any spatial axis. All results are derived exclusively from raw measurement counts under fixed audit invariants. No readout mitigation, post-selection, fitting procedures, or cross-platform normalization is applied. In addition to positive validation runs, the dataset includes a structurally defined negative control (One-Sided Twin), which is shown to fail deterministically. Across sampled values of the lens-strength protocol parameter, the observed outcomes recover the finite onset–window–breakdown regime structure introduced in DIG Part IV. This behavior is reported descriptively, without invoking universality, sharp thresholds, or new scaling claims. The results demonstrate that the operational and structural content consolidated in DIG Part IV persists under the multidimensional extension of the framework. Directional asymmetries and null projections are observed and are consistent with the tensorial formulation of information geometry in dimensions greater than one. All results reported in this work are fully reproducible from the accompanying public dataset without additional assumptions. Related DIG materials DIG Part I (theoretical framework):https://doi.org/10.5281/zenodo.17983399 DIG Part II (numerical validation):https://doi.org/10.5281/zenodo.17982058 DIG Part III (hardware validation):https://doi.org/10.5281/zenodo.18100057 DIG Part IV (regime structure and operational necessity):https://doi.org/10.5281/zenodo.18202656 DIG Part V (multidimensional formulation):https://doi.org/10.5281/zenodo.18210315 DIG Part VI (dataset):https://doi.org/10.5281/zenodo.18226272 DIG Part VII (dataset):Dataset for Experimental Validation of Dynamical Information Geometry in Two Dimensionshttps://doi.org/10.5281/zenodo.18309526 DIG — N1 (definitions, scope, and non-equivalences):https://doi.org/10.5281/zenodo.18001179 - Correspondence regarding this work may be directed to:kaya@cab-film.com