Information Capacity Frame (ICF) and Information-Level Unidirectional Constraint Principle (ILUC) English version

Cross-level information processing has long lacked a unifying constraint grounded in first principles. Beginning from three axioms—finite resources, maintenance of non-equilibrium states, and discrete inputs—this paper demonstrates that physical information systems operating within a rigid Information Capacity Frame (ICF) are subject to irreversible constraints on cross-level encoding transformations. Accordingly, we propose the Information Level Unidirectional Constraint (ILUC). This constraint encompasses five core propositions: irreversible loss in dimensionality reduction; irreversible loss in dimensionality expansion; loss that scales with dimensional difference; intrinsic logical time binding in dimensionality expansion; and an inverse relationship between hierarchical level and processing speed. Substrate-independent, ILUC applies to all hierarchical information systems within finite capacity frames. We extend ILUC's validation domain to software compilation, AI explainability, cognition, and emergence in complex systems, demonstrating that black-box unexplainability, abstraction leakage, and knowledge transfer loss are structurally inevitable consequences of hierarchical transformation, while the explanatory gap constitutes merely a higher-order manifestation within temporal systems. This paper provides independent foundational support for the principal thesis Temporal Integration, Bidirectional Criticality, and Hierarchical Emergence of Consciousness, and delineates information-theoretic boundaries for frontier areas including embodied intelligence, cross-modal fusion, and Sim-to-Real transfer. English version; Chinese version is also available.