The Fifth Industrial Revolution (Industry 5.0) foregrounds human–machine collaboration, sustainability, and resilience as organizing principles for next-generation cyber-physical systems. Yet the identity and access management (IAM) architectures inherited from Industry 4.0 remain perimeter-centric, policy-static, and blind to the behavioral dynamics of human–AI teaming. This paper introduces the Human-Centric Zero Trust Identity Architecture (HC-ZTIA), a novel framework that repositions identity as the adaptive control plane for Industry 5.0 environments. HC-ZTIA integrates three mutually reinforcing innovations: (1) a Joint Embedding Predictive Architecture (JEPA)-driven Behavioral Identity Assurance Engine (BIAE) that learns abstract world models of operator and machine-agent behavior to perform continuous, context-aware identity verification without relying on raw biometric surveillance; (2) a Privacy-Preserving Adaptive Authorization Protocol (PP-AAP) employing zero-knowledge proofs and federated policy evaluation to enforce least-privilege access across human, non-human, and hybrid identity classes while satisfying data-minimization mandates; and (3) a Resilience-Oriented Trust Degradation Model (RO-TDM) that guarantees fail-safe identity governance under adversarial, degraded, or disconnected operating conditions characteristic of operational technology (OT) and critical infrastructure. The framework is grounded in the Agile-Infused Design Science Research Methodology (A-DSRM) and formally extends NIST SP 800-207 and the CISA Zero Trust Maturity Model by addressing five identified gaps in human-centric identity governance. We present the formal system model, threat model, architectural specification, and a multi-scenario evaluation spanning energy-sector OT, smart manufacturing, and vehicle-to-everything (V2X) environments. Simulation results, validated through Monte Carlo trials with 95% confidence intervals, demonstrate that HC-ZTIA reduces identity-related breach exposure by 73.2% (±4.1%) while maintaining sub-200 ms authorization latency, offering a principled bridge between Zero Trust rigor and Industry 5.0 human-centricity.
Privacy-preserving aggregation is a cornerstone for AI systems that learn from distributed data without exposing individual records, especially in federated learning and telemetry. Existing two-server protocols (e.g., Prio and successors) set a practical baseline by validating inputs while preventing any single party from learning users' values, but they impose symmetric costs on both servers and communication that scales with the per-client input dimension $L$. Modern learning tasks routinely involve dimensionalities $L$ in the tens to hundreds of millions of model parameters. We present TAPAS, a two-server asymmetric private aggregation scheme that addresses these limitations along four dimensions: (i) no trusted setup or preprocessing, (ii) server-side communication that is independent of $L$ (iii) post-quantum security based solely on standard lattice assumptions (LWE, SIS), and (iv) stronger robustness with identifiable abort and full malicious security for the servers. A key design choice is intentional asymmetry: one server bears the $O(L)$ aggregation and verification work, while the other operates as a lightweight facilitator with computation independent of $L$. This reduces total cost, enables the secondary server to run on commodity hardware, and strengthens the non-collusion assumption of the servers. One of our main contributions is a suite of new and efficient lattice-based zero-knowledge proofs; to our knowledge, we are the first to establish privacy and correctness with identifiable abort in the two-server setting.
Facing two challenges in distributed power trade, that is, blockchain throughput bottleneck problem and transaction privacy leakage problem, this paper designs an efficient consensus mechanism based on dynamic clustering and zero knowledge proof. Specifically, the network is firstly divided into multiple consensus groups in parallel by multi-dimensional feature dynamic clustering algorithm, which improves the communication topology; then, zero knowledge proofs are applied into consensus process, and groups can generate globally verified proofs for all transactions in certain period without leaking any information; finally, a hierarchical hybrid consensus architecture is designed to achieve fast local sorting and efficient global confirmation. The experimental results show that the mechanism can achieve 2150 TPS with 300 nodes, and the success rate of privacy attack is lower than 10 %. Meanwhile, the mechanism still maintains high stability under dynamic perturbation. The research proves that the mechanism can effectively solve the core requirements of efficiency and privacy in distributed energy trade.
Enterprise identity systems can authenticate workloads, credentials, and attested platforms, but they do not close the composition layer where runtime model identity enters authorization. A token can verify that a service is running in a trusted environment, that its credentials are valid, and that its actions are authorized — without ever establishing which neural network is actually computing. When model identity evidence is inserted into standard authorization flows, new security properties emerge that are not inherited from the underlying protocols and must be formally established rather than presumed. This paper presents a live integration architecture for model-identity attestations in JWT and SPIFFE-style token flows, grounded in real measurements from six neural networks executed inside an NVIDIA H100 Confidential Computing enclave. It formally verifies four composition properties — non-separability, temporal binding, issuer authenticity, and reference integrity — across three Coq proof files with zero unfinished proof obligations. Every remaining trust dependency is explicitly named, traced to an integration control, and paired with a concrete falsification witness. The result is a formally hardened composition layer where no security property is left implicit and no assumption is left silent. Supplementary Material This paper is accompanied by three Coq proof files — ComposableIdentity.v, IssuerAuthenticity.v, and ReferenceIntegrity.v — that formally verify the four composition properties described in §§4–6: non-separability, temporal binding necessity, issuer authenticity, and reference integrity. Together the files prove thirteen theorems from eleven named axioms, each paired with a concrete falsification witness and an integration control. No file contains unresolved obligations (Admitted), and all three compile cleanly under the Rocq Prover 9.1.1 (the current release of the Coq proof assistant, compiled with OCaml 5.4.0). They are available for download as supplementary files attached to this record. The Neural Network Identity Series — Mathematical foundations, empirical validation, and governance frameworks for verifying which model is running Newest addition: Technical Note: The Disappearing Window — AI Logprob Access Withdrawal and the Structural Verifiability of Frontier Model Contracts (DOI: 10.5281/zenodo.20362098) Paper 1: The δ-Gene: Inference-Time Physical Unclonable Functions from Architecture-Invariant Output Geometry (DOI: 10.5281/zenodo.18704275) Paper 2: Template-Based Endpoint Verification via Logprob Order-Statistic Geometry (DOI: 10.5281/zenodo.18776711) Paper 3: The Geometry of Model Theft: Distillation Forensics, Adversarial Erasure, and the Illusion of Spoofing (DOI: 10.5281/zenodo.18818608) Paper 4: Provenance Generalization and Verification Scaling for Neural Network Forensics (DOI: 10.5281/zenodo.18872071) Paper 5: Beneath the Character: The Structural Identity of Neural Networks — Mathematical Evidence for a Non-Narrative Layer of AI Identity (DOI: 10.5281/zenodo.18907292) Paper 6: Which Model Is Running?: Structural Identity as a Prerequisite for Trustworthy Zero-Knowledge Machine Learning (DOI: 10.5281/zenodo.19008116) Paper 7: The Deformation Laws of Neural Identity (DOI: 10.5281/zenodo.19055966) Paper 8: What Counts as Proof? — Admissible Evidence for Neural Network Identity Claims (DOI: 10.5281/zenodo.19058540) Paper 9: Composable Model Identity — Formal Hardening of Structural Attestations in the Enterprise Identity Stack (DOI: 10.5281/zenodo.19099911) Paper 10:Where Identity Comes From: Path Sensitivity and Endpoint Underdetermination in Neural Network Training (DOI: 10.5281/zenodo.19118807) Paper 11: Post-Hoc Disclosure Is Not Runtime Proof: Model Identity at Frontier Scale (DOI: 10.5281/zenodo.19216634) Paper 12: Family-Dependent Response to Reasoning Distillation Across Structural and Functional Identity Layers (DOI: 10.5281/zenodo.19298857) Paper 13: Safety-Alignment Removal as a Model-Identity Failure — Structural Evidence from Published Weight-Level Mutation Checkpoints (DOI: 10.5281/zenodo.19383019) Technical Note: Agent Identity Is Not Model Identity (DOI: 10.5281/zenodo.19240883) Technical Note: Gap Invariance: Why PPP Measurements Are Domain-Independent by Construction (DOI: 10.5281/zenodo.19275524) Technical Note: Measured Model Substitution Under Valid Agent Credentials (DOI: 10.5281/zenodo.19342848) Technical Note: Artifact Identity Is Not Runtime Identity — Trustfall Lite and the Boundary of File-Level Model Verification (DOI: 10.5281/zenodo.20019127) Formal Verification Stack for Neural Network Structural Identity (IT-PUF Coq Proofs) (DOI: 10.5281/zenodo.18930621) Copyright (c) 2026 Anthony Ray Coslett / Fall Risk AI, LLC. All Rights Reserved. Confidential and Proprietary. Patent Pending (Applications 63/982,893, 63/990,487, 63/996,680, 64/003,244).
Annotatsiya.Ushbu maqolada blockchain texnologiyasining moliyaviy tranzaksiyalarni nazorat qilishdagi arxitekturaviy imkoniyatlari IMRAD metodologiyasi doirasida tahlil qilinadi. Distributed Ledger Technology (DLT), smart-kontraktlar va kriptografik audit izlarining moliyaviy nazorat tizimiga integratsiyasi nazariy va empirik jihatdan asoslantirilgan. Yettita konsensus algoritmi (PoW, PoS, PBFT, DPoS, PoA, Raft) moliyaviy sektor uchun qiyosiy baholangan. An’anaviy va blockchain asosidagi nazorat tizimlarining olti o‘lchamda qiyosiy tahlili o‘tkazildi. ISO 31000:2018 asosida yettita asosiy risk (R₁–R₇) baholandi. Estoniya, Dubay, Singapur, JPMorgan, Braziliya va O‘zbekiston misolida xalqaro tajriba sintezi amalga oshirildi. Samaradorlik tahlili shuni ko‘rsatadiki, blockchain asosidagi nazorat tizimi tranzaksiya narxini 40–60%, operatsiya vaqtini 60–70%, xatoliklar sonini 80–90% va audit jarayonini 50–60% qisqartirishi mumkin. O‘zbekiston uchun Hyperledger Fabric + PBFT/PoA konsensus asosidagi ruxsatli blockchain modeli tavsiya etiladi. Kalit so‘zlar: blockchain, distributed ledger technology (DLT), smart-kontrakt, konsensus algoritmi, kriptografik audit izi, moliyaviy nazorat, zero-knowledge proof, CBDC, Hyperledger Fabric, O‘zbekiston raqamli moliyasi. Аннотация. В данной статье в рамках методологии IMRAD анализируются архитектурные возможности технологии блокчейн в сфере контроля финансовых транзакций. Теоретически и эмпирически обоснована интеграция технологии распределенного реестра (DLT), смарт-контрактов и криптографических аудиторских следов в систему финансового контроля. Проведена сравнительная оценка семи алгоритмов консенсуса (PoW, PoS, PBFT, DPoS, PoA, Raft) для финансового сектора. Осуществлен сравнительный анализ традиционных и основанных на блокчейне систем контроля по шести параметрам. На основе стандарта ISO 31000:2018 оценены семь ключевых рисков (R1-R7). Выполнен синтез международного опыта на примере Эстонии, Дубая, Сингапура, JPMorgan, Бразилии и Узбекистана. Анализ эффективности показывает, что система контроля на основе блокчейна может сократить стоимость транзакций на 40-60%, время операций - на 60-70%, количество ошибок - на 80-90% и длительность аудита - на 50-60%. Для Узбекистана рекомендуется модель разрешенного блокчейна на базе Hyperledger Fabric с консенсусом PBFT/PoA. Ключевые слова: блокчейн, технология распределенного реестра (DLT), смарт-контракт, алгоритм консенсуса, криптографический аудиторский след, финансовый контроль, доказательство с нулевым разглашением, CBDC, Hyperledger Fabric, цифровые финансы Узбекистана. Abstract. This article analyzes the architectural capabilities of blockchain technology in monitoring financial transactions, utilizing the IMRAD methodology. It provides theoretical and empirical justification for integrating Distributed Ledger Technology (DLT), smart contracts, and cryptographic audit trails into financial monitoring systems. Seven consensus algorithms (PoW, PoS, PBFT, DPoS, PoA, Raft) are comparatively evaluated for their suitability in the financial sector. A comparative analysis of traditional and blockchain-based monitoring systems is conducted across six dimensions. Seven key risks (R1-R7) are assessed in accordance with ISO 31000:2018. A synthesis of international experience is presented, drawing on case studies from Estonia, Dubai, Singapore, JPMorgan, Brazil, and Uzbekistan. The efficiency analysis indicates that a blockchain-based monitoring system can reduce transaction costs by 40-60%, operational time by 60-70%, error rates by 80-90%, and audit process duration by 50-60%. For Uzbekistan, a permissioned blockchain model based on Hyperledger Fabric, utilizing a PBFT or PoA consensus algorithm, is recommended. Keywords: blockchain, distributed ledger technology (DLT), smart contract, consensus algorithm, cryptographic audit trail, financial monitoring, zero-knowledge proof, CBDC, Hyperledger Fabric, Uzbekistan digital finance.
Open access
2 source records
Advanced Computational Techniques in Science and Engineering
Climate change, driven by global warming and associated greenhouse gas (GHG) emissions, poses a significant global challenge. International organizations and governments are actively pursuing emission reduction strategies, yet these efforts are often constrained by the direct relationship between emissions and national economic activity. This paper proposes a blockchain-based carbon footprint (CF) management system named P u r e C a r b o P r i n t , that leverages data from IoT devices, and security is ensured by zero-knowledge proofs (ZKP) to track and reduce CF at the individual level. Individual data is collected and converted into carbon coin, a hybrid (online-offline) crypto coin operating on the Pure Chain network. A smart contract, deployed on the Pure Chain network using the Pure Chain coin, governs the proposed system. Additionally, zk-SNARK is applied to implement ZKP for the validity and integrity of the information verification without revealing private information. Based on theoretical models and previous studies on behavior-based carbon reductions, it is expected that the system can achieve up to a 30% reduction in CF per user within the first year.
Language isn't just a rigid system of symbols. Instead, it's a living, embodied phenomenon, deeply intertwined with our physical experience and shaped by our interaction with the environment (Wang, 2019;Zhou & Luo, 2024). However, the dynamic nature of language brings a significant challenge to cognitive science: the well-known "stability-plasticity dilemma" (Grossberg, 1980). On one hand, for clear communication, meanings of words need to be stable and widely recognized, so everyone can understand them, no matter when or who speaks them. On the other hand, these meanings must also be flexible and adaptable in varying contexts. While traditional computational models, from early generative grammar to standard Bayesian approaches, have excelled at modeling these stable meanings, they often treat semantic ambiguity as "noise" that needs to be eliminated, rather than a valuable resource (Gärdenfors, 2014).Even with the significant "probabilistic turn" in cognitive science, which brought Bayesian models to handle uncertainty, most of these models still rely on classical probability theory. They assume the meaning of a concept is a pre-defined distribution over a set of fixed features. As Bruza and Cole (2005) pointed out, this dependence on classical set theory creates a major epistemological barrier because it treats semantic ambiguity as "noise" rather than a fundamental part of meaning construction. Though scholars have recently developed more complex tools, like Gradient Symbolic Representations (GSR), to model meanings as weighted mixtures (Smolensky et al., 2014;Mondal, 2024), these approaches are still limited by Kolmogorovian probability. They still follow the Law of Total Probability, which forces conflicting meanings to be simply added together and mixed. We argue that this basic "mixture" method isn't enough to describe or handle complex situations where meanings are incompatible or interfere with each other in context. Therefore, much empirical evidence suggests that capturing these dynamic features requires a non-classical, quantum probability framework (Surov et al., 2021).The importance of this paradigm shift becomes most clear when we analyze how everyday language works and how we interpret deep meanings in complex literary works. A classic example of such "semantic superposition" is the iconic "big fish" in Ernest Hemingway's The Old Man and the Sea.Within the novel's narrative structure, this phrase isn't a static label. Instead, it operates simultaneously on multiple, even mutually exclusive, semantic levels. Here, it serves as a biological marlin, a worthy adversary, and a transcendent symbol of life's ultimate tragedy. A classical probabilistic model fails to capture the dynamic tension that a reader feels, because it forces these meanings to compete for probability mass, implying only one can be dominant. In contrast, human reading suggests that meaning exists in a "superposition" state. It stays that way until a specific context makes it "collapse" into a concrete interpretation. Crucially, these overlapping meanings aren't simple probabilistic blends. They are coherent "quantum states" within a complex adaptive system.This study proposes that Quantum Cognition offers the necessary mathematical formalism to resolve the "stability-plasticity dilemma". This is supported by its proven success in solving decision-making paradoxes in psychology (Busemeyer & Bruza, 2012;Widdows et al., 2023;Huang et al., 2025). We introduce an integrated quantum theoretical model. In this model, the interaction between embodied experience and linguistic context is characterized as a genuine quantum interference phenomenon.This framework reinterprets the tension between stability and plasticity through the lens of Wave-Particle Duality. In our model, the "particle" corresponds to the stable, discrete symbols used for communication. The "wave" captures the fluid, context-sensitive potential that allows for creative interpretation.Next, by employing the mathematical formalism of Hilbert space, we will mathematically demonstrate how semantic ambiguity can be maintained as a useful resource, rather than mere noise.This approach effectively overcomes the limitations inherent in traditional methods like static vectors and gradient symbolic mixtures. To ground these abstract formalizations, we focus on the "Big Fish" motif in Hemingway's The Old Man and the Sea. Through this case analysis, we will reveal how meaning dynamically evolves, similar to "state vector collapse". Our study also extends to address the fundamental limitations of current Artificial Intelligence, particularly Large Language Models (LLMs). We argue that current LLMs, relying heavily on static statistical correlations, lack the "grounding" for true understanding. Therefore, we propose a pathway toward Quantum-Embodied AI and photonic intelligent systems by incorporating quantum-semantic principles. These systems could mimic the non-algorithmic fluidity of the human mind. Quantum probability is not an exotic addition to linguistics but a fundamental requirement for describing dynamic meaning. The research will first analyze the evolution from gradient representations to quantum interference, then formally express the wave-particle duality of meaning using mathematical methods. We will then validate this theoretical framework through the "big fish" case study and neurophysiological evidence, concluding with an exploration of its practical implications for Generative AI and Photonic we need the "quantum we the evolution of semantic theory. We will focus on mathematical models often to capture the nature of meaning. We will how meaning with static symbolic then to gradient we will the limitations of these these limitations a paradigm shift a quantum probability this "quantum turn" in deep that the basic of quantum such as and can be as of human meaning and interpretation. quantum theory can be as a framework for describing a to by like is a or is a In classical computational this is through vector Models like and a meaning by its fixed in a space, from While this method effectively captures stable such as the vector between and it becomes rigid when et that a static vector to a like creates a This vector becomes a weighted of its meanings, the in a semantic that is to models or to resolve this with dynamic they are still on They treat as a that simply the However, they lack an to model the the context is & these models assume the of not the This is by research on which that the of and et et these computational offers into how meaning from like have that meaning is not a fixed but a dynamic However, mathematical to describe the probabilistic of mutually & our of It can how but its is limited how these "collapse" in et al., et al., address the of static a major is the Symbolic Representations by et We can of as a that to the fluid, of with the of symbolic traditional vector models that a symbol or not at allows a symbol to have This approach creates a computational where or semantic meanings can simultaneously as a weighted This a more of how our language on recently this used gradient symbols to the of with the fluidity of cognitive research that linguistic meaning is as a not a discrete This allows for in semantic with evidence that meaning is and & the nature of human this model It captures the that a concept can to at a theoretical still these the framework and model follow the basic of classical probability. In these gradient models, meanings are as an not a true This in cognitive to quantum probability & & This is and we can it with an a classical gradient model an like a that on or but is by a Here, our is simply a lack of In contrast, quantum probability suggests its is more like a which is simultaneously and a dynamic until context or this in to the Law of Total that the probability of a is just the simple of its argue that this "mixture" model is for complex or linguistic are characterized not just by but by theoretical is supported by et "quantum in They that human of semantic often classical probability the & research that the interaction between a and its context creates interference similar to the and in a These interference can a to model classical theory. Therefore, this suggests that the human ambiguity not as a lack of but as a coherent cognitive resource representations have in and This to a "quantum turn" in This shift to like context-sensitive concept and meaning or even and using quantum to This a concept isn't it like a quantum on context and They the and similar so with quantum like and interference, of simple set as & This a for quantum in language as quantum not as a simple of fixed features with et This approach how we understand and rigid traditional specific concept the framework of quantum cognitive offers theoretical for in language meaning. between "quantum and "quantum that can formally features like quantum such as and However, this that quantum physical in our These are to more cognitive In this the of our the and meaning with like a varying Crucially, how these quantum to semantic and concept in quantum representations of the meaning of a isn't simply from its computational semantic models vector to be This can quantum such as of meaning and or rather than just to Bruza and Cole (2005) that semantic can be as true They deeply how quantum over can model context and our practical In this they the of meaning to quantum on the and that that are by quantum probability than by classical using and have of in semantic the that is a of meaning et approaches to quantum computational interpret meanings as quantum and and quantum et with quantum how can be and in quantum with and quantum as for modeling and semantic et "quantum turn" in computational language and the of intelligent systems of meaning on can intelligent systems more et language treat words as over semantic and to in This on like and "semantic et al., Language by to "quantum using genuine quantum to meanings et al., This method allows AI to potential or semantic in which then dynamically "collapse" on quantum This paradigm shift will semantic from methods to more context-sensitive models, which is for AI to understand dynamic dynamic interaction between and when or can also be using quantum and the concept of "quantum theory that "superposition" and into can like the into et quantum of the how quantum and and decision-making in we the limitations of traditional semantic models, inherent in capturing the and the nature of linguistic meaning. This the "quantum turn" in and an for cognitive in While this fundamental shift evidence and models for in meaning a more is still to the dynamic and of meaning within a on these basic we propose a theoretical framework on an the of our are formally similar to quantum The of this framework is the of used to understand "semantic a for the and of framework proposes an the of our are formally similar to quantum It the dilemma" in cognitive Language needs to be stable and to flexible enough to meanings on context. We argue that this is to the of quantum have the of the as a a like a discrete in a evidence a dynamic where and are theory on that are more like than words with this dynamic that meaning is like into a stable within a semantic where meanings are simultaneously and shaped by the of quantum on the theoretical by and Bruza and and we propose that a concept not be as a fixed set of a but as a "state of a This is by probability for this model, a a on its In the to a in the it as a probability an of the can be as a superposition" of semantic such as each to a & In this the a of meanings, effectively a superposition" with interference potential semantic & when we a in a specific or this as a This interaction a where the becomes a "particle" of into the specific This effectively traditional models, which treat words as to capture the dynamic in meaning our of the of we the mathematical framework of Quantum Hilbert this complex traditional classical set theory or vector everyday have Kolmogorovian probability when with and interference in meaning our model, a meaning is as a vector in a complex Hilbert The is a of meaning vectors This "superposition" can be by the the in the is not a classical but a probability It of a and a is the The of the is the theoretical of our model. In are not in probability but they how a cognitive will with in contexts. they describe the between on the from a cognitive this can be as a or in classical probabilistic but a in and of this formalism becomes when the probability of a meaning. In the probability of the of is simply the of However, in Quantum Probability, we the to the of which an a and the probability is the are the meaning is modeling the in a when are the meaning is mathematically human semantic can classical (Surov et al., is and a exists in a semantic To model the interaction between a and its context we the framework by et We describe the semantic using the have used in and to capture that are not to features et al., et al., this space, the semantic can a simple an that the semantic of the and context are no the of the is on the of the context. This the of meaning the meaning of a complex is an of the linguistic system and be to a simple of its et is not it as the narrative or We propose that the cognitive of a a by a of the meaning is by the evolution have to and dynamic in cognitive in (Busemeyer & Bruza, the the or the tension of the This how the in the Hilbert over modeling the of a reader a validate the theoretical framework we from abstract mathematical to a practical the approach to analyze the of in Ernest Hemingway's The Old Man and the specific is to how words in a and how meaning into a understanding. This "quantum isn't just for it's a It will demonstrate how "quantum probability can mathematically model the of deep meaning from semantic a that classical models have with when with et & Hemingway's the as a simultaneously On one hand, the the biological with a and the of and This its biological the the it to a the the a and an in for This in and will this In a traditional semantic or standard the of and are They in of a a cognitive such conflicting simultaneously to and Therefore, a classical model that the of to semantic or However, the experience isn't one of or but a This suggests that the human simple it in physical with Our is to mathematically this of mathematically the "semantic we a complex Hilbert This is by each a in the In this the vector for the meaning the is an and a of The other vector the the as a adversary, a and a symbol of the or when the is in its meaning is it's in a of To capture the between its biological and we assume a This we to the biological and the in a the "state vector is as a coherent a fundamental a specific narrative context forces "semantic the probability of the biological or meaning is Crucially, the between the is set to a coherent where This mathematical a cognitive where the system potential in until a specific context to a meaning of our is how the specific narrative context of the forces the meaning of to traditional models that treat context as a static our "quantum narrative context as a of the This the cognitive of Hemingway's in we where is through physical rather than by this we introduce a that the cognitive This is It the "semantic We that to capture the for a of with the the an abstract a of with the a biological a of is also implying or like a between matter and the meaning we a to the This specific is because it a context that is a the biological This captures the literary that the is a but its and the of that our can be as we traditional and our quantum method in how a reader a into a the between becomes We that corresponds to the probability of the the context we to this model using traditional we will a significant theoretical In the traditional classical the and of the are as not like that can The Law of Total requires to the of each the context. for the biological and we the probability of the biological with the context and it to the probability of the with its context This way of by as and simply the classical model is to implying the from its in a semantic However, this our experience of reading the when we to the the quantum we we first We the between the and the context method the to The A is by the of the the biological and the This in a of the this to the we quantum method a that approaches than the classical This significant is mathematically to the the and the context in the Hilbert by they theoretical It offers mathematical for a that literary have but to it that Hemingway's on physical in the not from the it as an This biological as a that rather than the This with on modeling of and interference, how context and can probability in and et modeling meaning as a we the experience of the mathematical of the This that when coherent is a resource for meaning not just to be our quantum model to have mathematical it must with the human We that the such as interference, and in our analysis, and describe the and within our This is with the dynamic et cognitive of its in the While classical linguistics often treats words as for words in the meaning a can be a discrete and dynamic this exists through evidence the of et that narrative the linguistic in This creates a of where semantic are in These are to and the coherent quantum this the mathematical of by the in our can be by and between in the The through and research that on within specific et & the a narrative context The Old in with the a more effectively at and are more to This is a of are or it or This the interference in our conflicting semantic to on a the from to which we "state vector at the cognitive can be as a of representations in our et a quantum probabilistic model of Here, and interference of cognitive how our of semantic traditional and classical probability the we this "collapse" is the shift from a more to a more that a specific interpretation. This shift be by dynamic in and a interpretation. on and literary reading also that complex meaning interaction between systems and This suggests that meaning requires the and of & our this of is like a from a of to one this theory in serves a current limitations of Artificial Large Language Models are but rely on static vector to understand with the nature of meaning. Our approach suggests AI must the we to a computational model that ambiguity not as to out, but as a fundamental resource for meaning construction. This framework also for neurophysiological how can how semantic are theoretical framework for dynamic meaning literary or cognitive It significant implications for how classical probabilistic models to capture the and interference in human meaning. This research AI a practical is in the major semantic that Generative Artificial from classical to isn't just an an paradigm shift for to meaning with the and as the human generative models, such as LLMs, linguistic as using classical probability to the This approach However, it context-sensitive of meaning into As research traditional vector models often meanings into one This makes it to or and language modeling proposes a it treats linguistic or as within a Hilbert This formalism at and and widely it to model mixtures of semantic and over meanings et 2025). In this a like a a semantic A by a captures a probabilistic of semantic classical these semantic models have this et from They treat each in as a over semantic on and that these representations to this to using and to and probabilistic language This in semantic et on these a semantic offers to analyze literary meaning. the in The Old Man and the it can be a biological and a symbol or In a these can be as or within a semantic Hilbert the importance of each or between them. This allows the system to a dynamically semantic rather than it to a with theoretical and empirical in which to handle linguistic ambiguity and dynamically meaning in contexts. like and and to and meanings within In these approaches, significant over vector in inherent ambiguity and semantic representations from to which model this modeling to semantic correlations, & this the is more than a complex for It a an semantic system to semantic like conflicting or as its captures between semantic for Crucially, the also meaning through with models of and These features a mathematical for modeling and meaning. they a for semantic to the limitations of static generative models but or that in language isn't a simple It from a of and a lack of or et 2025). a quantum can be as the of semantic within a complex language models propose a can be as quantum This of words and and similar to quantum interference et et 2025). In these models, meaning and or between or is by quantum on These that through complex or can express complex semantic and dynamic on these we can a method for generative quantum to or supported in a semantic a have We can with to these that the context and are will have to interference in or will have This will in the probability current models are not generative in and models, where quantum is to and stability intelligent for quantum and to useful and in complex et the and in AI from an and "semantic that from semantic in probabilistic models and suggests the need for that this these one can a In this the semantic of a or narrative be as a in a Hilbert space, often as a over semantic in a classical or then be into this semantic as or A the and of these on and effectively of and interference such a no be a it an of the semantic While this a theoretical it is with empirical This evidence suggests that models can capture semantic and the dynamic evolution of meaning more effectively than classical and semantic et et et we how intelligent systems are through embodied and photonic a major in current Artificial becomes and meaning from our dynamic physical AI models as are of physical a and with to In contrast, deep the of and This significant and inherent the of "semantic states" that can be in photonic and for semantic and to on photonic integrated et et These systems are for for traditional such as and photonic have for deep models, with et 2025). This that photonic can more the and In the quantum integrated and on and quantum et that meaning is not just an abstract vector but a within an In a photonic semantic could be by These in and interference and In this a semantic to a specific of and in a and by the While quantum and language models interference, and for semantic at the et photonic and a way to these representations into the of of on not for such dynamic this an Artificial need to mimic biological Instead, it requires a physical can and semantic Photonic and with of Hilbert interference and a clear toward such embodied we as within a complex space, human into a where and interpret these the of meaning between a and a can be as a of the true how to how to or how to While classical theory effectively it at each are and Quantum It allows to be and within Hilbert This the decision-making and the of that even the of and between these can significant coherent mixtures of by have to classical in 2025). quantum to to model each In this model, to fixed in these This mathematical framework is in like and where quantum and quantum have as for over et a semantic these theoretical that ambiguity and the of not be as but rather as valuable a AI an that allows for even It then from the to and the semantic state. language and models and to and capture complex between linguistic that are for classical methods to theory extends this to the interaction can be as or dynamically through They an that not only but also of and et semantic this paradigm a fundamental AI not or as to be Instead, they or incompatible They then to a of state. Quantum and such interaction models in which the computational is not a but a distribution over and these into AI the of system AI focus on meaning through This will a for it from current with and study a in cognitive science: how the human with its and to and such fluid, and often We argue that most current computational by words as mere static or are rigid to capture this dynamic nature of meaning. Therefore, we propose a the concept of "superposition" from classical cognitive with Quantum to our this a is no a of fixed in the but a dynamic system wave-particle A in its exists as a "wave" of it only into a meaning when by a or of Hemingway's The Old Man and the this We mathematically the tension between and in the The that quantum probability and interference, effectively and that classical probability as a classical model a probability when with conflicting semantic but our quantum model a probability of coherent meaning This mathematical for the that context just it the semantic features to and of each other Crucially, this mathematical is in biological the between the quantum and suggests that the is not a but a of the this framework offers a theoretical we must its current which also research our in a simple Hilbert While for the interference in a semantic of semantic features within a more research will need to to model these computational we must be the of the We are not the is a quantum at Instead, we for that the as a biological to using probability because these are more for these our practical implications literary a for psychology and Artificial cognitive science, this model suggests a using to specific of semantic interference meaning Artificial Intelligence, our the vector as the fundamental of in current Large Language Models the Quantum a for AI These could to understand and meaning in complex We that intelligent systems of the deep meaning of not just requires a shift to Quantum Language and Photonic that to model semantic just as meaning is more than a mathematical the novel's "Big Fish" is more than a simple It a dynamic the of into a of understanding. the quantum nature of this allows to with the of human to the of the mind.
ABSTRACT To resolve the conflict between privacy protection and compliance verification of commercial contracts in the digital economy, this study aims to propose the application of a commercial contract generation technology based on verifiable encryption signature algorithms in civil and commercial law. This technology innovatively combines Attribute‐Based Encryption (ABE) with Zero‐Knowledge Succinct Non‐Interactive Argument of Knowledge (zk‐SNARKs) to construct a new electronic contract framework. Methodologically, ABE is used to encrypt contract clauses, enabling refined “on‐demand decryption” based on the roles and permissions of participants. Meanwhile, zk‐SNARKs are utilized to generate proofs for clauses requiring verification (e.g., transaction amount), allowing third parties (such as regulatory authorities) to complete “privacy‐preserving verification” without accessing the original contract text. Experimental results show that when the model processes a contract containing 500 clauses, the size of the generated data package is only 64.7 KB, which is approximately 3.8 times smaller than that of the pure ABE encryption model (247.8 KB) and 7.4 times smaller than that of the traditional digital signature model (480.1 KB). At the same time, the key size for users to access 500 clauses is only 5.2 KB, which is much smaller than that of the baseline models. Therefore, this technical solution effectively resolves the long‐standing “privacy‐verification” dilemma in commercial activities. This also proves its feasibility and economy in practical applications with the advantages of low storage and transmission costs, providing secure and reliable technical support for civil and commercial legal practices in the digital era.
The global digital identity landscape is undergoing an unprecedented crisis. Approximately 1.1 billion individuals worldwide lack any verifiable form of digital identity, while existing identity systems face existential threats from the industrialization of deepfake technology with injection attacks targeting biometric verification surging 900% since 2022 and occurring at a rate of once every five minutes in 2024. Simultaneously, conventional blockchain-based identity proposals that store biometric templates on-chain introduce critical privacy vulnerabilities incompatible with emerging regulatory frameworks including the EU AI Act (2024) and GDPR. This paper presents ZKP-GDIS (Zero-Knowledge Proof Global Decentralized Identity System), a novel, privacy-by-design identity architecture that fundamentally departs from prior work in three key dimensions. First, ZKP-GDIS never stores raw biometric data on-chain; instead, it employs zk-SNARK (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge) cryptographic commitments that allow identity verification without any disclosure of underlying biometric features. Second, we introduce a Hybrid Deepfake-Resistant Liveness Pipeline (HDRLP) — a multi-modal anti-spoofing layer that fuses passive CNN-based texture analysis, photoplethysmography (PPG) heart-rate detection, and hardware-attested device fingerprinting to defeat both presentation and injection attack vectors. Third, the system adopts W3C Decentralized Identifier (DID) standards and implements a federated governance model, enabling cross-jurisdictional interoperability while respecting national digital sovereignty. We provide formal security proofs under the computational Diffie-Hellman hardness assumption, evaluate the system against the ISO/IEC 30107-3 Presentation Attack Detection benchmark, and report experimental results demonstrating 99.87% genuine acceptance rate, 0.004% false acceptance rate under deepfake attack, and 94% reduction in on-chain gas costs versus Ethereum mainnet through zkEVM Polygon deployment. ZKP-GDIS establishes a reproducible, standards- compliant, and audit-ready framework for the next generation of global digital identity infrastructure.
The book&s;s emphasis is shifted to the future in Chapter 8 , “Emerging Trends and Advanced Topics,” which offers a survey of the cutting-edge ideas and revolutionary technologies that are set to completely alter the logistics industry. The chapter makes the case that logistics is developing into a hyperconnected, intelligent, and autonomous ecosystem rather than a collection of distinct tasks. The convergence of multiple important technological domains, each supported by complex mathematical and engineering principles, is what is driving this evolution. In order to enable strong predictive analytics, demand forecasting, and real-time optimization, artificial intelligence and machine learning are positioned as the brains of logistics operations in the future. This is where the story starts. The idea of “digital twins,” which produce virtual representations of entire supply chain networks in real time, expands on this theme. These virtual models, mathematically grounded in dynamical systems and Markov Decision Processes, allow companies to simulate complex scenarios, assess risks, and test optimization strategies in a virtual environment before physical implementation. From optimization, the chapter moves to the critical issues of trust and security, highlighting Blockchain for Supply Chain Transparency. It explains that blockchain&s;s function will transcend simple record-keeping, creating a decentralized and tamper-proof ledger for all transactions. The text delves into the mathematical foundations securing this trust, including cryptographic hash functions, Zero-Knowledge Proofs (ZKP), and secure consensus algorithms. The physical implementation of these trends in warehouse automation and robotics is finally examined in this chapter. Multi-agent reinforcement learning and graph-based optimization models are used to coordinate swarms of autonomous mobile robots and cooperative “cobots” in hyper-automated warehouses of the future, achieving previously unheard-of levels of efficiency and adaptability. This idea is expanded to the scale of advanced manufacturing and smart factories, where material flows are coordinated in real-time to satisfy changing manufacturing demands and logistics becomes a deeply integrated, cyber-physical part of the production system itself. According to the chapter&s;s conclusion, the combination of these technologies will result in intelligent, self-adjusting, and extremely resilient logistics networks, offering businesses that adopt this technological change a major competitive edge.
We know how to document an AI system. We know how to test it, log what it did, and report when something goes wrong. What current governance practice does not clearly tell us is how to verify which model is actually computing. This is not a hypothetical gap. When an organization says "this is the model we evaluated," that claim is typically supported by a model card, a registry entry, or a hash of a weight file — evidence about a *file*, not about the system that is running. A neural network is not a static document. A weight file stores the network; the model is what appears when that file is loaded and begins transforming inputs into outputs. The file and the running model are related, but they are not the same thing — and current governance practice rarely distinguishes between them. This paper proposes a framework for doing so. It identifies three kinds of evidence that can support model identity claims, each answering a different question. Structural evidence — drawn from the model's internal computations during live operation — can verify which specific model is running, and is the most resistant to tampering. Thermodynamic evidence — drawn from the model's output statistics — can verify that the system is a genuine neural network rather than a substitute, but cannot distinguish one model from another. Functional evidence — drawn from patterns in the model's outputs over an API — can detect whether a model was copied from another, but this signal fades quickly: routine model updates can erase it within days to weeks of continued training. The paper shows that inspecting the model's files alone is insufficient for verifying which specific model is running. The identity-bearing signal cannot be recovered from the tested static properties of those files; it is most reliably established by observing the model while it operates. The paper formally proves that these three kinds of evidence cannot substitute for one another. Verifying that a system is genuine does not tell you which specific model it is. Detecting that a model was copied does not tell you the identity of the copy. The practical consequence is a standard for identity claims: any claim should declare which kind of evidence supports it, because borrowing evidence from the wrong category produces unreliable conclusions. The framework maps directly to compliance questions raised by current AI governance obligations, including those under the EU AI Act. It provides the missing evidentiary specification for model identity claims: which kind of evidence is admissible for which identity question. Supplementary Material This paper is accompanied by EvidenceSufficiency.v, a Coq proof file that formally verifies the cross-layer inadmissibility results described in §4. The proof mechanically checks each logical step of the observation-limited verification impossibility theorem and its three directional corollaries. The file contains no unresolved obligations (Admitted) and compiles cleanly under the Rocq Prover 9.1.1 (the current release of the Coq proof assistant, compiled with OCaml 5.4.0). It is available for download as a supplementary file attached to this record. Technical Note: Artifact Identity Is Not Runtime Identity — Trustfall Lite and the Boundary of File-Level Model Verification (DOI: 10.5281/zenodo.20019127) The Neural Network Identity Series — Mathematical foundations, empirical validation, and governance frameworks for verifying which model is running Newest addition: Technical Note: The Disappearing Window — AI Logprob Access Withdrawal and the Structural Verifiability of Frontier Model Contracts (DOI: 10.5281/zenodo.20362098) Paper 1: The δ-Gene: Inference-Time Physical Unclonable Functions from Architecture-Invariant Output Geometry (DOI: 10.5281/zenodo.18704275) Paper 2: Template-Based Endpoint Verification via Logprob Order-Statistic Geometry (DOI: 10.5281/zenodo.18776711) Paper 3: The Geometry of Model Theft: Distillation Forensics, Adversarial Erasure, and the Illusion of Spoofing (DOI: 10.5281/zenodo.18818608) Paper 4: Provenance Generalization and Verification Scaling for Neural Network Forensics (DOI: 10.5281/zenodo.18872071) Paper 5: Beneath the Character: The Structural Identity of Neural Networks — Mathematical Evidence for a Non-Narrative Layer of AI Identity (DOI: 10.5281/zenodo.18907292) Paper 6: Which Model Is Running?: Structural Identity as a Prerequisite for Trustworthy Zero-Knowledge Machine Learning (DOI: 10.5281/zenodo.19008116) Paper 7: The Deformation Laws of Neural Identity (DOI: 10.5281/zenodo.19055966) Paper 8: What Counts as Proof? — Admissible Evidence for Neural Network Identity Claims (DOI: 10.5281/zenodo.19058540) Paper 9: Composable Model Identity — Formal Hardening of Structural Attestations in the Enterprise Identity Stack (DOI: 10.5281/zenodo.19099911) Paper 10:Where Identity Comes From: Path Sensitivity and Endpoint Underdetermination in Neural Network Training (DOI: 10.5281/zenodo.19118807) Paper 11: Post-Hoc Disclosure Is Not Runtime Proof: Model Identity at Frontier Scale (DOI: 10.5281/zenodo.19216634) Paper 12: Family-Dependent Response to Reasoning Distillation Across Structural and Functional Identity Layers (DOI: 10.5281/zenodo.19298857) Paper 13: Safety-Alignment Removal as a Model-Identity Failure — Structural Evidence from Published Weight-Level Mutation Checkpoints (DOI: 10.5281/zenodo.19383019) Technical Note: Agent Identity Is Not Model Identity (DOI: 10.5281/zenodo.19240883) Technical Note: Gap Invariance: Why PPP Measurements Are Domain-Independent by Construction (DOI: 10.5281/zenodo.19275524) Technical Note: Measured Model Substitution Under Valid Agent Credentials (DOI: 10.5281/zenodo.19342848) Technical Note: Artifact Identity Is Not Runtime Identity — Trustfall Lite and the Boundary of File-Level Model Verification (DOI: 10.5281/zenodo.20019127) Formal Verification Stack for Neural Network Structural Identity (IT-PUF Coq Proofs) (DOI: 10.5281/zenodo.18930621) Copyright (c) 2026 Anthony Ray Coslett / Fall Risk AI, LLC. All Rights Reserved. Confidential and Proprietary. Patent Pending (Applications 63/982,893, 63/990,487, 63/996,680, 64/003,244).
<sec> <title>BACKGROUND</title> The rapid integration of deep learning into nuclear medicine promises to revolutionize precision oncology but faces a critical "trust gap." As AI models become "black boxes," clinicians struggle to verify the integrity of individual diagnostic inferences, leaving systems vulnerable to adversarial attacks and silent model drift. </sec> <sec> <title>OBJECTIVE</title> This formative evaluation proposes and validates an in-silico proof-of-concept for a blockchain-agnostic Proof of Inference (PoI) protocol. The objective is to establish a standard of Computational Integrity for AI-assisted workflows in nuclear medicine without exposing proprietary model weights or patient privacy. </sec> <sec> <title>METHODS</title> he PoI protocol leverages Zero-Knowledge Succinct Non-Interactive Arguments of Knowledge (zk-SNARKs), specifically the Groth16 proof system. An in-silico feasibility study was conducted using a 1.2-million-parameter U-Net model on synthetic 128×128 Ga-68 PSMA-11 PET slices. Proof generation and verification latencies were benchmarked using an NVIDIA A100 GPU and a standard CPU, respectively. </sec> <sec> <title>RESULTS</title> The architectural analysis demonstrates that the protocol successfully offloads computational burden to the prover (cloud server). In our empirical benchmarking, cryptographic proof generation required 28.81 seconds per inference. Crucially, client-side verification of the proof was completed in 448.59 milliseconds, demonstrating that cryptographic attestation can be integrated into existing PACS viewers with sub-second, clinically acceptable latency. </sec> <sec> <title>CONCLUSIONS</title> The proposed PoI protocol provides a feasible forensic support layer for medical AI. By shifting clinical trust from institutional reputation to deterministic cryptographic assurance, this infrastructure creates a tamper-evident audit trail essential for algorithmic accountability in decentralized healthcare environments. </sec>
Open access
Artificial Intelligence in Healthcare and Education
This paper introduces the Quantum Tensor Sequence (.qtsq) format, a universal file format built to compress any kind of data — whether it's images, audio, video, text, structured data, or raw binary. Instead of treating everything as a generic stream of bytes like most compressors do, .qtsq looks at what the data actually is before deciding how to compress it. An internal component we call the "Spaghettification Engine" analyzes the input, figures out which of 16 data types it belongs to, and picks the best compression strategy for it: Iterated Function Systems (IFS) for images, Discrete Fourier Transform (DFT) for audio and signals, dictionary-schema encoding for text and structured data, and procedural seed generation for binary data. Everything is packed into a single 80-byte header that supports 11 features: type-aware compression, lazy partial decompression, AES-256-GCM encryption, Schnorr zero-knowledge proofs, Reed-Solomon error correction, inter-file deduplication through wormhole links, compressed-domain differential updates, multi-resolution quality scaling, and adaptive size thresholds. The whole thing is organized around a five-layer architecture inspired by black hole physics — each region of the black hole maps to a stage of the compression pipeline. As far as we know, no existing file format brings all of these capabilities together in one place. Based on the theoretical properties of the algorithms involved, we expect compression ratios somewhere between 18:1 and 55:1 on mixed-type data. Real-world benchmarks using the reference implementation will follow in a separate paper.
Filippo Scaramuzza, Marco Tonnarelli, Damian A. Tamburri, W.J.A.M. van den Heuvel
The challenge of achieving auditable, GDPR-compliant consent management while preserving true data subject privacy persists, as current blockchain-based solutions compromise anonymity through public ledgers. This paper addresses this by introducing a novel, privacy-by-design IT artefact built on the Ethereum platform that employs Zero-knowledge Succinct Non-Interactive ARgument of Knowledge (zk-SNARKs) to enable truly anonymous and irrefutable validation of data processing permissions. Implemented and evaluated through a Design Science Research (DSR) approach, the artefact demonstrated a high degree of functional and GDPR compliance, confirming its capacity to generate trustless, auditable on-chain proof of permission status. The proof of concept successfully implements core privacy-by-design principles through anonymity and encryption, with technical benchmarks indicating acceptable performance for the artefact's intended use despite the computational intensity of proof generation.
Secure cross-domain UAV authentication is challenging because identity verification alone is insufficient to guarantee safe operation. In many UAV applications, it is equally critical to verify that a UAV is currently located within an authorized geographic region. Existing approaches often expose precise GPS coordinates, rely on static identifiers that enable tracking, or fail to guarantee the freshness and authenticity of location evidence. These weaknesses allow replay, location spoofing, and trajectory inference attacks, especially in multi-domain environments. To address these limitations, we propose PrivLocAuth, a zero-knowledge-based cross-domain UAV authentication protocol that enforces geofence restrictions without revealing actual locations. In PrivLocAuth, UAVs encode their current coordinates into fresh Pedersen commitments, which are attested by the home Local Domain Server (LDS) using short-lived Schnorr signatures. Based on these attested commitments, UAVs generate Bulletproof range proofs to demonstrate compliance with cross-domain server-defined geofences. This design ensures that UAVs operate within authorized airspace while preserving strong location privacy. PrivLocAuth further incorporates a lightweight elliptic curve cryptography (ECC) and Schnorr signature-based credential framework that enables unlinkable authentication across-domains, preventing session correlation and identity tracking. Formal security analysis demonstrates resistance to impersonation, replay, geofence-bypass, and linkage attacks. Experimental evaluation shows low computational latency and minimal communication overhead, confirming the protocol’s suitability for resource-constrained UAV platforms operating in dynamic cross-domain environments.
We present NanoZK, a zero-knowledge proof system for verifiable LLM inference: clients and third-party auditors check that a provider executed the advertised model on a committed input without learning weights or activations. NanoZK introduces a layerwise proof framework that decomposes transformer inference into independently provable layers linked by a SHA-256 commitment chain, yielding constant-size sub-circuit proofs (3.5-3.7 KB; about 83 KB total at L=12), comparable in total size to and substantially more parallelizable than prior ZKML's monolithic 101-126 KB proofs. We prove compositional soundness and zero-knowledge under standard assumptions, design 16-bit lookup-table approximations for softmax, GELU, and normalization with measured perplexity degradation below 1e-4 across six model/dataset combinations, and add a Fisher-information-guided audit-budget triage as an efficiency tool (full soundness still requires verifying every layer). On CPU the MLP sub-circuit proves in about 6.3 s prove-only (about 43 s setup plus prove) with about 22 ms verification at any width; attention prove-only time scales from 0.9 s (d=16) to 184 s (d=256); full-block end-to-end proofs are measured to d=128, with a projected GPU time of about 68 s per block at d=768 from measured O(d^2) MSM scaling and a conservative 15-30x GPU-MSM speedup range based on Icicle's published 30x result for n >= 2^20 and extrapolated to the smaller-n regime. Privacy scope: NanoZK hides weights and activations from verifiers and auditors but does not hide the prompt from the prover; this is complementary to HE/MPC.
We present NanoZK, a zero-knowledge proof system for verifiable LLM inference: clients and third-party auditors check that a provider executed the advertised model on a committed input without learning weights or activations. NanoZK introduces a layerwise proof framework that decomposes transformer inference into independently provable layers linked by a SHA-256 commitment chain, yielding constant-size sub-circuit proofs (3.5-3.7 KB; about 83 KB total at L=12), comparable in total size to and substantially more parallelizable than prior ZKML's monolithic 101-126 KB proofs. We prove compositional soundness and zero-knowledge under standard assumptions, design 16-bit lookup-table approximations for softmax, GELU, and normalization with measured perplexity degradation below 1e-4 across six model/dataset combinations, and add a Fisher-information-guided audit-budget triage as an efficiency tool (full soundness still requires verifying every layer). On CPU the MLP sub-circuit proves in about 6.3 s prove-only (about 43 s setup plus prove) with about 22 ms verification at any width; attention prove-only time scales from 0.9 s (d=16) to 184 s (d=256); full-block end-to-end proofs are measured to d=128, with a projected GPU time of about 68 s per block at d=768 from measured O(d^2) MSM scaling and a conservative 15-30x GPU-MSM speedup range based on Icicle's published 30x result for n >= 2^20 and extrapolated to the smaller-n regime. Privacy scope: NanoZK hides weights and activations from verifiers and auditors but does not hide the prompt from the prover; this is complementary to HE/MPC.
Title: Compositional Transfer in Neural World Models via Symbolic Law Discovery Core Thesis This research establishes a mathematically grounded paradigm for Modular World Modeling, where physical invariants are recovered as additive vector fields rather than monolithically memorized. We prove that by framing learning as Tangent-Space Residual Superposition, neural networks can internalize isolated physical laws that compose zero-shot to predict complex, unseen multi-physics environments. Key Breakthroughs & Upgrades The Compositional Scaling Law (2D to 12D) Our experiments reveal a fundamental divergence in high-dimensional scaling. While monolithic models suffer from "Baseline Washout" and entanglement, our modular ensembles maintain physical integrity across 12-dimensional manifolds. In chaotic triple-force environments, the modular framework achieves a 6.5× reduction in trajectory MSE ($71.5 \times 10^{-4}$ vs. $470.2 \times 10^{-4}$ for the monolith). Causal Discovery: Active Gradient Conflict ($\rho \approx -0.99$) We provide the first empirical proof identifying the causal driver of monolithic failure. Gradient alignment analysis reveals that monolithic models are trapped in a state of Active Gradient Sabotage, where the update required for one force (e.g., Gravity) almost perfectly cancels out the update for another (e.g., Spring). Our framework bypasses this bottleneck by isolating gradients in tangent space, ensuring 100% of task-specific knowledge is preserved. Benchmark vs. Physics-Informed Neural Operators (PINO) A head-to-head comparison with the PINO paradigm reveals a fundamental Compositional Utility Gap. While PINOs are powerful solvers for specific partial differential equations, they fail the Zero-Shot test because solution operators are inherently non-additive. Our framework is not only capable of additive operator composition but is also 2.5× faster at inference and provides a direct path to SINDy Symbolic Discovery (99.25% recovery accuracy). Hierarchical Physical Discovery The framework scales hierarchically, enabling the unsupervised decomposition of environments into continuous dynamics (gravity) and discrete contact events. This allowed the discovery of hidden physical constants, such as the coefficient of restitution ($\epsilon=0.80$), without target-domain supervision. Scientific Impact These results transform world modeling from a holistic storage problem into a sparse algebraic retrieval problem. By bridging the gap between black-box simulation and verifiable symbolic laws, this work provides a scalable roadmap toward interpretable Artificial General Intelligence (AGI) that respects the structural symmetries of the physical universe.
South Africa’s experience shows how participatory sustainability models can translate climate goals into locally led action across very different contexts. This chapter synthesizes evidence from urban Tshwane and rural Limpopo to examine how community empowerment, inclusive governance, and co-creation improve resilience while advancing multiple Sustainable Development Goals (SDGs). In Tshwane, collaborative planning through the Integrated Development Plan, a climate action plan to 2050, and programmes such as the Pretoria East Urban Biosphere Reserve and school-based climate initiatives combined citizen science, green infrastructure, and nature-based solutions to restore ecosystems, strengthen water and energy security, and mainstream net-zero building standards. In Limpopo, the community-led Multiple-Use Services model and the Giyani Local-Scale Climate Resilience Programme paired Indigenous knowledge with technical support to co-design low-cost water systems, deploy solar-powered boreholes and small treatment units, and build local capacity for monitoring and maintenance. A six-step participatory framework underpinned both cases, from joint problem definition through implementation and skills transfer, with Innovation Forums enabling representation of women and youth and improving accountability. Comparative analysis highlights common enablers (trusted intermediaries, transparent budgeting, citizen science, and blended finance) and barriers (fragmented mandates, uneven data, limited digital access, and funding continuity). Policy recommendations include institutionalising community-led mechanisms in municipal processes, scaling nature-based and MUS approaches where appropriate, investing in skills and open data, and creating fit-for-purpose finance that matches local capabilities. The South African lessons offer scalable, context-sensitive pathways for empowering communities to climate-proof urban and rural systems.
Sustainability and Climate Change Governance
Innovative Approaches in Technology and Social Development
Let $ξ$ be the stationary occupation field generated by a Poisson system of independent simple symmetric random walks on $\mathbb Z$ in space--time dimension $1+1$. For a finite set $A\subset\mathbb Z$, we consider the classical fixed-region observables $W_N(A)$, the cumulative occupation of $A$ up to time $N$, and $D_N(A)$, the number of distinct particles visiting $A$ up to time $N$. We prove quantitative central limit theorems for both observables, with Wasserstein rate of order $N^{-1/4}$. In addition, we introduce an independent nearest-neighbour random walk $S=(S_n,\,n\ge 0)$ on $\mathbb Z$ with non-zero drift and sample the field along this ballistic path. For a fixed polynomial observable $φ(x)=\sum_{j=0}^k β_j x^j, β_k\neq 0$, of degree $k\in \mathbb N$, we consider the partial sums $Y_{N,φ}=\sum_{n=1}^N φ(ξ(n,S_n)).$ We prove a Wasserstein bound of order $N^{-1/2}$ for the normal approximation of the standardized $Y_{N,φ}$. To the best of our knowledge, this is the first quantitative normal approximation result for polynomial functionals of the Poisson occupation field sampled along a random walk path. The drift induces an effective decorrelation of the sampled environment, leading to a substantial improvement over fixed-region sampling. The proofs rely on a representation of $ξ$ as a Poisson functional on path space and on the Malliavin--Stein method for Poisson functionals.
Palm leaf manuscripts have rich sources of knowledge and information reflecting cultural, historical, and linguistic knowledge. Extracting information from palm leaf manuscripts poses significant challenges for preservation and access, as they are fragile in nature. We propose an advanced multimodal deep learning framework for the digitization, character reconstruction, and decentralized federated learning of palm leaf manuscripts. The proposed approach integrates Transformer-based OCR models (TrOCR, LayoutLM), Vision Transformers (ViTs), and Contrastive Language-Image Pre-training (CLIP) to enhance character recognition for damaged and missing characters in the manuscripts. Natural Language Processing Algorithms are implemented to restore incomplete or faded characters while preserving the originality of the manuscripts. To ensure secure and decentralized access, we employ a blockchain-based federated learning system where metadata, translations, and reconstructed text are securely stored on a Zero-Knowledge Proof (ZKP) blockchain ledger. Federated learning across distributed nodes minimizes the centralized dependencies while enabling real-time collaborative OCR model updates. Scalability is ensured by Docker and Kubernetes where real-time processing is done across distributed nodes. Experimental results demonstrate superior OCR accuracy (96.3%), improved character restoration fidelity (92.7%), and enhanced blockchain security with minimal overhead (4.2%), outperforming traditional methods. The proposed method stores the manuscripts in a digitized form, providing easy access for researchers and scientists globally. The proposed work unlocks the hidden treasures, knowledge and information from the cultural treasures. Results obtained show that the efficiency and scalability of the proposed approach paves a path into the digital era by enhancing cultural preservation.
We present COMET-MPC (Commitment-Oriented Multi-Party Computation with Equality Testing), a novel MPC framework that fundamentally reimagines multi-party computation as a commitment verification problem rather than a traditional secret-sharing problem. COMET-MPC achieves dealerless, one-round setup by reducing all MPC coordination tasks—input binding, consistency verification, and replay prevention—to homomorphic equality testing over elliptic-curve commitments. Unlike classical MPC protocols (Shamir’s secret sharing, FROST) that require multi-round polynomial verification or interactive zero-knowledge proofs, COMET-MPC performs verification through a single algebraic check: testing whether a group element equals the identity. This zero-detection paradigm exploits the structural identity between EC-Pedersen commitments and EC-ElGamal ciphertexts, enabling decrypt-free verification with perfect correctness and computational hiding under the Decisional Diffie-Hellman (DDH) assumption. We provide complete algorithmic specifications, formal security proofs including binding, hiding, replay resistance, and a rigorous security level analysis demonstrating ≈ 128-bit security for standard 256-bit elliptic curves. COMET-MPC is particularly suited for real-world ledger systems, authentication protocols, confidential databases, and privacy-preserving regulatory compliance where minimal interaction, minimal trust, and minimal leakage are paramount.
Ensuring atomic execution of cross-shard transactions is a fundamental challenge for sharded blockchains, particularly in scenarios demand coordination across multiple shards. However, existing solutions either rely on on-chain coordination, leading to high communication overhead, or leverage secure hardware for off-chain execution, imposing strong trust assumptions and reducing general applicability. To this end, we propose RollShard, a sharded blockchain that integrates stateless off-chain mechanism to efficiently process multi-shard transactions (MSTs). In RollShard, each MST is abstracted into a transaction DAG by the Sequencer Shard to ensure the authenticity of the transaction content and the correctness of its execution order. Batched MSTs are dispatched to off-chain executors, each of which simulates transaction logic using a virtual zero-state model integrate with a hierarchical state-delta tree (HSDT). The HSDT employs a Merkle Sum tree to precisely capture batched MSTs’ impact on per-shard account states. Based on the HSDT, the executor generates the zero-knowledge proof to attest the correctness of each shard’s state changes and global value conservation. The resulting net state deltas are then optimistically committed to the relevant shards without cross-shard coordination, reducing intra-shard coordination. We design a game-theoretic incentive mechanism to ensure rational behavior of off-chain executors, showing that honest execution forms a Nash equilibrium under collateral staking. Experimental results based on a prototype deployed in a local area network demonstrate that ROLLSHARDsignificantly outperforms two baseline coordination models proposed in ByShard, namely the Linear and Distributed designs. Specifically, under high workload, RollShard improves throughput by 44.9% and 158%, and reduces cross-shard latency by 38.9% and 42.1%, compared to the Linear and Distributed models, respectively.