GP-2026-018 — April 2026 https://doi.org/10.5281/zenodo.19164275 Abstract Existing practitioner frameworks for assessing vulnerability to political violence are diagnostic at the wrong level. VERA-2R, ERG22+, and behavioural risk tools assess the individual who is already showing precursor signs, or the content that is already circulating — they begin, structurally, where the problem has already advanced. The Genealogical Translation Framework operates upstream: it trans

GP-2026-013 | March 2026 Author: Iliyan Kuzmanov Editor: Iliyan Kuzmanov DOI: https://doi.org/10.5281/zenodo.19070816 ABSTRACT Adversarial mobilisation succeeds not by constructing psychology from zero but by activating cognitive-affective configurations that populations carry as historically installed substrates — architectures of identity, threat, grievance, and sacred necessity that persist beneath continuous surface ideological mutation. Focused Genealogical Configur

GP-2026-012 March 2026 https://doi.org/10.5281/zenodo.19007846 Author: Angel Analytical Team Editor: Iliyan Kuzmanov Abstract Anticipatory intelligence names a structural gap that all historically calibrated detection systems share. Their accuracy within familiar distributions is genuine, institutionally earned, and operationally valuable — and it cannot extend to configurations that have not yet produced detectable surface signatures. That interval, between when threateni

GP-2026-011 March 2026 https://doi.org/10.5281/zenodo.19007711 Author: Angel Analytical Team Editor: Iliyan Kuzmanov Abstract Cognitive monoculture in artificial intelligence systems is not primarily a workforce diversity problem. It is a structural design outcome with predictable security consequences. When training data, evaluation benchmarks, and development teams all reflect the same cognitive profile, the resulting system does not merely underrepresent alternative ana

GP-2026-010 March 2026 https://doi.org/10.5281/zenodo.19007646 Author: Angel Analytical Team Editor: Iliyan Kuzmanov Abstract Pattern recognition systems — the algorithmic architectures that underpin modern machine learning — achieve their extraordinary analytical power through a mechanism that simultaneously defines their structural limitation. Trained on historical distributions, these systems identify known categories with high confidence and speed. What they cannot do