DATASET_DIG_IN_GIT_WITNESS: Reproducible Hardware and Mirror-Test Runs for the DIG-in-GIT Framework (Raw Counts, Audit-Fixed Evaluation)

This dataset provides the complete, audit-fixed experimental material for the DIG-in-GIT witness analysis. It contains: All hardware run packs (ibm_fez, ibm_torino) Raw count files (no mitigation, no post-selection) Reference packs used during evaluation Twin-sign reports Sector fallback reports (where applicable) Runner scripts used to generate the hardware jobs Mirror-test run material README documentation describing the evaluation protocol Evaluation discipline: Raw-count basis only No error mitigation No post-selection Fixed reference pack per run Fixed evaluator logic No parameter tuning after data acquisition The dataset allows full third-party reproduction of the reported DIG-in-GIT witness results, including twin-sign classification and dimensional consistency checks. All materials are provided exactly as used in the reported analysis.A SHA256 checksum file accompanies the archive to ensure integrity verification. Scope: This dataset documents the operational layer of the DIG-in-GIT framework.It makes no ontological claims and introduces no new dynamical laws.It provides reproducible evidence regarding directional consistency structure under audit-fixed evaluation conditions. Correspondence regarding this work may be directed to:kaya@cab-film.com