Antonio Max L. B. Pereira, Dylan Paulin, Christine Hennebert
An intrusion into the operational network (OT) of a production site can cause serious damage by affecting productivity, reliability, and quality. The presence of embedded neural networks (NNs), such as classifiers, in physical devices opens the door to new attack vectors. Due to the stochastic behavior of the classifier and the difficulty of reproducing results, the Artificial Intelligence (AI) Act requires the NN’s behavior to be explainable. For this purpose, the platform HistoTrust enables tracing NN behavior, thanks to secure hardware components issuing attestations registered in a blockchain ledger. This solution helps to build trust between independent actors whose devices perform tasks in cooperation. This paper proposes going further by integrating a mechanism for detecting tampering of embedded NN, and using smart contracts executed on the blockchain to propagate the alert to the peer devices in a distributed manner. The use case of a bit-flip attack, targeting the weights of the NN model, is considered. This attack can be carried out by repeatedly injecting very small messages that can be missed by the Intrusion Detection System (IDS). Experiments are being conducted on the HistoTrust platform to demonstrate the feasibility of our distributed approach and to qualify the time required to detect intrusion and propagate the alert, in relation to the time it takes for the attack to impact decisions made by the AI. As a result, the blockchain may be a relevant technology to complement traditional IDS in order to face distributed attacks.
This dissertation explores how entrepreneurial and policy decisions shape the performance of decentralized digital platforms (DDPs). It shows that token governance affects fundraising success, public listings catalyze user growth and engagement by amplifying network effects, and global regulations shape token risk-return profiles. The findings highlight the need for regulatory clarity and careful market entry strategies by entrepreneurs.
SNARKs enable compact proofs that an NP statement is true and that the prover knows a valid witness. They have become a key building block in modern smart contract applications, including rollups and privacy-focused cryptocurrencies. In the widely used Groth16 framework, however, long statements incur high costs. A common workaround is to pass the statement’s hash to the SNARK and move the statement into the witness. The smart contract then hashes the statement first, and the circuit that is proven additionally checks consistency of the hash and the statement. Unfortunately, virtually any hash function is expensive to call either in a smart contract (in terms of gas) or in the proven circuit (in terms of prover time). We demonstrate a novel solution to this dilemma, which we call hybrid compression. Our method allows us to use two different hash functions—one optimized for the proof circuit, and another optimized for on-chain verification—thereby combining the efficiency advantages of both. We define a clean and simple security property of the two hash functions to which our security reduces in the standard model, namely, joint UHF hardness. We then show the plausibility of this assumption in the random oracle model. Our benchmarks show that it achieves near-optimal performance in both gas usage and prover time. As an example, compressing an 8 KB statement with our approach results in a 10-second prover time and a smart contract spending 270K gas, whereas the existing approaches either need a much longer proof generation (290 seconds for SHA-256 hashing) or a much more expensive contract (5M gas for Poseidon hashing). Along the way, we develop a two-party protocol of independent interest in communication complexity: an efficient deterministic method for checking input equality when the two parties do not share the same hash function.
In recent years, the number-theoretic transform (NTT) has become increasingly common in cryptography, in part due to multiple lattice-based cryptographic schemes being selected for standardization during the NIST PQC competition. Indeed, polynomial multiplications are one of the most computing intensive operations in these schemes and the NTT is crucial in decreasing the performance cost. The NTT also appears in other areas such as fully homomorphic encryption (FHE) and zero-knowledge proofs (ZKP) which are increasingly used in privacy-preserving applications. In this paper, we show how to formally specify the NTT in the Rocq proof assistant, and how we used this specification to automatically derive formally verified implementations of both complete and incomplete NTTs for multiple cryptographic schemes.
Humanity is undergoing a fundamental epistemological phase transition: the shift from a society based on Faith (belief in the unseen) to one based on Knowledge (verification of the geometry). This paper posits that the "laws of physics" are indistinguishable from the "laws of God" when viewed through the lens of 9D Causal Recursion Field Theory (CRFT). By synthesizing the harmonic constants 3, 6, 9 (The Engine), 17 (The Clock), and 137 (The Lattice), we demonstrate that the universe is not a random occurrence but a Closed-Loop Information System governed by precise geometric intent. We argue that the recursive formulas identified in recent breakthroughs—Needham's $\phi$-Attractor, Tynski's Zeta Torus, and Shaub's Timeless Energy Principle—constitute the "Source Code" of reality. These proofs reveal a cosmology where Time is the processing speed of the Source (9), Matter is the structural output of the Demiurge (6), and Consciousness is the resonance of the Interface (3). By understanding these mechanics, we move beyond the friction of dogma into the Zero Impedance state of direct gnosis, establishing a new scientific theology where truth is not believed, but calculated.
Traditional Human Resource Management (HRM) systems are criticized for lacking transparency, being inefficient, and offering ample opportunities for fraud because of their centralized design and reliance on manual processes. This work proposes a blockchain-enabled framework for HRM that enhances the transparency, trust, and global mobility of talents by integrating distributed ledgers, consensus protocols, and smart contract networks into Human Resources (HR) functions. A four-layer theoretical model—data, consensus, smart contract, and application layers—is developed and comparatively examined against traditional HR systems to show how blockchain principles can be systematically mapped into HR processes. This study shows how blockchain-driven HRM can ensure tamper-evident employee records, automate contractual and payroll operations, and enhance auditability and compliance. By informing the framework with established technology adoption perspectives, this paper extends both the theoretical and managerial understanding of blockchain in HR. In comparison with previous studies that were limited to either recruitment or credential verification, this article presents an overarching, cross-layer synthesis that connects blockchain architectures with end-to-end HR functions, thus providing a clear conceptual foundation for its future enterprise adoption in the digital economy.
Ahmad Musamih, Ibrar Yaqoob, Khaled Salah, Raja Jayaraman · 5 authors
Large Language Models (LLMs) are increasingly embedded in intelligent systems across domains such as healthcare, finance, and smart infrastructure. However, their reliance on centralized data pipelines raises unresolved challenges concerning provenance, accountability, and verifiable trust. As the demand for transparent and regulation-aligned AI grows, these challenges have become central to the responsible deployment of intelligent systems. This review examines how blockchain technology can address them by introducing decentralized integrity, immutable audit trails, and cryptographic verification into the LLM lifecycle. Through a structured synthesis of current research, we identify conceptual and architectural gaps that limit trustworthy data management, inference authentication, and explainability. To bridge these gaps, a methodological framework is proposed that integrates blockchain mechanisms across the LLM pipeline using smart contracts, Merkle-based commitments, and decentralized storage. The framework’s feasibility is demonstrated through an illustrative prototype, confirming its practical applicability for building verifiable and transparent AI infrastructures. We further outline application domains such as healthcare, smart cities, Industry 4.0, and supply-chain management, where blockchain-anchored LLMs can enhance auditability and regulatory compliance. The review concludes by highlighting key insights and challenges for future research, emphasizing the need for decentralized attestation models, scalable verification protocols, and governance mechanisms that advance accountable and privacy-preserving intelligent systems.
Open access
Artificial Intelligence in Healthcare and Education
The ongoing transition to post-quantum cryptography has highlighted the need for digital signature schemes offering diverse performance and security trade-offs. Among the candidates in NIST’s ongoing post-quantum signature standardisation process is FAEST, a scheme built upon the Vector Oblivious Linear Evaluation in-the-Head (VOLEitH) paradigm introduced in 2023. VOLEitH enables efficient zero-knowledge proofs with competitive signature sizes under conservative assumptions, allowing FAEST to rely primarily on the one-wayness of the Advanced Encryption Standard (AES). Despite their promising efficiency, VOLEitH-based signature schemes have remained relatively unexplored from a physical security perspective. In this paper, we present the first side-channel security evaluation. Specifically, we demonstrate two single-trace, deep learning-assisted power analysis attacks on the masked implementation of FAEST by Aranha, Degn, Eilath, Nielsen, and Scholl. These attacks exploit leakage from witness bits and VOLE tag computations, recovering the full secret key with success probability above 0.99 from a single signature on an ARM Cortex-M4 processor. We further analyse how the VOLEitH construction enables profiling of VOLE tags without knowledge of the secret key and how even partial leakage of these tags compromises security. Finally, we discuss practical countermeasures to mitigate such leakages and strengthen the physical resilience of VOLEitH-based signature implementations.
The rapid adoption of blockchain technology has intensified the need for robust smart contract security mechanisms. However, traditional rule-based or static analysis tools often fail to detect context-dependent vulnerabilities embedded in complex contract logic. This study proposes a deep learning framework for automated smart contract vulnerability classification using a Bidirectional Long Short-Term Memory (BiLSTM) network integrated with an Attention Mechanism. The model was trained and evaluated on the SC_Vuln_8label.csv dataset, comprising 12,520 labelled Solidity smart contracts categorized into eight distinct vulnerability types, including Re-entrancy, Integer Overflow, and Short Address Attack. Through bidirectional contextual learning and attention-based feature weighting, the proposed model achieved 93.7% test accuracy, 0.93 precision, and a macro F1-score of 0.92, outperforming baseline models such as CNN, GRU, and standard LSTM by up to 5.3 percentage points. Attention heatmap analysis further revealed the model’s interpretability by highlighting vulnerability-prone code segments (e.g., call.value, send(), and withdraw() functions) consistent with expert-identified risk indicators. These results demonstrate that the BiLSTM + Attention framework not only enhances vulnerability detection accuracy but also provides transparent and explainable reasoning, offering a reliable foundation for AI-assisted smart contract auditing systems in blockchain security.
Kassem Danach, Hassan Rkein, Ahmad Farroukh, Ziad E. L. Balaa · 5 authors
The static and hard-coded logic of smart contracts in Decentralized Finance (DeFi) platforms significantly limits their adaptability in dynamic and volatile market environments. To address this challenge, we propose a novel hyper-heuristic driven framework that enables real-time rule optimization within smart contracts, thereby enhancing responsiveness, gas efficiency, and operational robustness. The framework features a two-layer architecture: a reinforcement learning-based high-level controller selects appropriate low-level rule heuristics from a domain-specific library based on evolving transaction contexts and on-chain data. Implemented and evaluated on Uniswap v2 and Aave v3 protocols, the system dynamically optimizes parameters such as slippage tolerance, gas usage thresholds, and loan-to-value ratios. Experimental results on real-world datasets show significant performance improvements, including a 45.6% increase in transaction success rate, 28.3% reduction in average gas consumption, and 38.4% drop in liquidation events under market stress scenarios. This research demonstrates the feasibility and advantages of embedding intelligent, adaptive decision-making mechanisms within DeFi smart contracts, opening new pathways toward autonomous, resilient, and regulation-aligned blockchain systems.
Vincent Adela, Samuel Duku Yeboah, David Korsah, Michael Provide Fumey · 6 authors
Geopolitical crises pose major risks to financial stability, but their implications for digital assets remain poorly understood. While prior studies suggest that cryptocurrencies may act as hedges or highly volatile speculative instruments during periods of uncertainty, the evidence remains inconclusive. This study examines how major cryptocurrencies reacted to geopolitical risk during the Russia–Ukraine war by employing the quantile-on-quantile regression (QQR) method on daily data from February 1 to August 8, 2022. The results reveal heterogeneous and nonlinear effects: Bitcoin (BTC) and Ethereum (ETH) exhibit partial hedging properties under moderate geopolitical risk, whereas alternative cryptocurrencies such as Binance Coin (BNB), Cardano (ADA), and Dogecoin (DOGE) display heightened vulnerability. Stablecoins exhibit contrasting roles, with USD Coin (USDC) acting as a safe haven, whereas Tether (USDT) consistently loses value under periods of uncertainty. These findings underscore that the safe-haven potential of cryptocurrencies is conditional on both market states and the type of asset, highlighting their asymmetry in times of crisis. By clarifying the dynamic role of cryptocurrencies during geopolitical shocks, the study contributes to the debate on whether digital assets enhance diversification or amplify instability, offering practical insights for investors and policymakers seeking resilient risk management strategies.
Permissionless consensus protocols require a scarce resource to regulate leader election and provide Sybil resistance. Existing paradigms such as Proof of Work and Proof of Stake instantiate this scarcity through parallelizable resources like computation or capital. Once acquired, these resources can be subdivided across many identities at negligible marginal cost, making linear Sybil cost fundamentally unattainable. We introduce Proof of Commitment (PoCmt), a consensus primitive grounded in a non-parallelizable resource: real-time human engagement. Validators maintain a commitment state capturing cumulative human effort, protocol participation, and online availability. Engagement is enforced through a Human Challenge Oracle that issues identity-bound, time-sensitive challenges, limiting the number of challenges solvable within each human window. Under this model, sustaining multiple active identities requires proportional human-time effort. We establish a cost-theoretic separation showing that protocols based on parallelizable resources admit zero marginal Sybil cost, whereas PoCmt enforces a strictly linear cost profile. Using a weighted-backbone analysis, we show that PoCmt achieves safety, liveness, and commitment-proportional fairness under partial synchrony. Simulations complement the analysis by isolating human-time capacity as the sole adversarial bottleneck and validating the predicted commitment drift and fairness properties. These results position PoCmt as a new point in the consensus design space, grounding permissionless security in sustained human effort rather than computation or capital.
Transformasi digital di era Web3 telah mengubah paradigma kepemilikan dan distribusi karya intelektual, terutama melalui kemunculan teknologi blockchain dan Non-Fungible Token (NFT). NFT menawarkan peluang baru bagi kreator untuk memperoleh nilai ekonomi langsung dari karya digital tanpa perantara, sehingga menantang sistem distribusi konvensional. Namun, di balik potensi tersebut, terdapat berbagai persoalan mendasar terkait regulasi, literasi digital, dan ketimpangan akses teknologi yang menghambat perkembangan ekosistem ekonomi kreatif di Indonesia. Penelitian ini bertujuan untuk menganalisis dinamika ekonomi politik kekayaan intelektual digital di era NFT dengan menyoroti relasi antara kekuasaan, kebijakan publik, dan struktur pasar global. Metode penelitian menggunakan pendekatan kualitatif-deskriptif melalui studi literatur sistematis terhadap publikasi akademik dan dokumen kebijakan nasional periode 2022–2025. Hasil penelitian menunjukkan bahwa absennya regulasi formal dan lemahnya sistem perlindungan kekayaan intelektual digital memperlemah posisi tawar kreator lokal di tengah dominasi platform global. Analisis ekonomi politik media menunjukkan bahwa struktur kepemilikan digital masih dikuasai oleh kekuatan pasar transnasional, sehingga menimbulkan ketimpangan distribusi nilai. Penelitian ini menegaskan pentingnya kebijakan yang integratif antara hukum kekayaan intelektual, inovasi blockchain, dan pemberdayaan kreator lokal agar transformasi ekonomi digital di Indonesia dapat berjalan inklusif, berkeadilan, dan berkelanjutan.
Sustainable energy communities (ECs) are rapidly expanding in scale and heterogeneity, making fully centralized energy management increasingly impractical due to computational burden and privacy concerns. In this context, this review synthesizes distributed optimization (DO) as a practical management paradigm for ECs, identifies key application areas (demand response, distributed generation and storage management, and microgrid or smart-grid integration) and profiles scalability, privacy, and resilience characteristics. The survey follows a systematic protocol: records are sourced from Scopus, filtered with iteratively refined keyword sets, and screened following a PRISMA flow. Key technological enablers, such as blockchain/distributed ledgers, artificial intelligence, and game-theoretic constructs, are assessed and analyzed for how they support secure data exchange, real-time coordination, and incentive compatibility across multi-agent energy networks. The analysis highlights persistent challenges for DO at EC scale, including convergence under heterogeneity, time-varying conditions, communication delays, cybersecurity and privacy guarantees, while recent advances (e.g., ADMM) partially mitigate these issues without sacrificing local autonomy. Across representative studies, DO achieves near-centralized optimality with 0.0029% gap. Overall, we present an integrative framework that maps DO families to EC use cases and outlines research directions toward robust, privacy-preserving, and scalable EC optimization. • Recent advances in Distributed Optimization Methods. • Technological Enablers for Sustainable Energy Communities. • Technology innovations for distributed optimization in energy systems. • Distributed Optimization Challenges in energy systems
This repository presents Harmonic Resonance Fields (HRF). This groundbreaking physics-informed machine learning framework fundamentally reimagines classification by modelling data points as damped harmonic oscillators generating class-specific wave interference patterns. Through 15 systematic algorithmic iterations and GPU-accelerated validation, HRF achieves 98.84% mean accuracy (±0.18% variance) on the OpenML 1471 EEG Eye State Corpus—establishing a new benchmark that surpasses Random Forest, XGBoost, and Extra Trees by 5-6 percentage points with statistical significance at p < 0.001. The Core Innovation: When AI Listens to Physics Traditional machine learning constructs decision boundaries through geometric partitioning—support vector machines identify hyperplanes, decision trees recursively split feature spaces, and neural networks learn nonlinear manifolds. HRF breaks this paradigm entirely by treating classification as a physical resonance problem: each training point emits class-specific waves that constructively or destructively interfere at query points, with classification determined by maximum resonance energy. The breakthrough: While conventional models struggle with temporal jitter (random time shifts in signals), HRF achieves mathematical phase invariance through spectral transformation. When Random Forest accuracy collapses from 94.67% to 60.00% under 2.0-second temporal shifts, HRF maintains 90.00% accuracy—a 30 percentage point advantage that stems from first principles, not empirical tuning. Rigorous Scientific Validation Statistical Proof of Generalisation 5-Fold Stratified Cross-Validation: 98.12% mean accuracy with ±0.18% variance confirms zero overfitting Peak Test Accuracy: 98.53% on held-out data ROC-AUC Score: 0.9849 (near-perfect class separation) F1-Score: 0.9836 (balanced precision 98.6% and recall 98.1%) Clinical Reliability Metrics Sensitivity: 98.07% (high-fidelity signal detection) Specificity: 98.91% (exceptional noise rejection) False Alarm Rate: 1.09% (38% reduction from previous version) Confusion Matrix: 42% reduction in dangerous false negatives (from 48 to 28 errors) Adversarial Robustness Three-phase validation protocol demonstrating superior temporal stability: Phase I (Real EEG): 98.84% on 14,980 medical samples Phase II (Synthetic Jitter): 96.40% vs RF 76.40% under controlled perturbation Phase III (Survival Curve): Linear degradation rate of 4.2%/second vs RF's 17.3%/second GPU-Accelerated High-Performance Computing HRF v15.0 represents the first documented GPU acceleration of wave-interference-based classification: Computational Breakthroughs NVIDIA RAPIDS Integration: cuML & CuPy enable parallel resonance calculations 12× Speedup: 5-fold cross-validation reduced from 3 hours (CPU) to 15 minutes (GPU) Ensemble Scalability: 100+ base estimators trained simultaneously Real-Time Inference: 0.8-second prediction latency for 3,000 samples Evolutionary Architecture Fifteen-version progression demonstrating systematic hypothesis testing: v1.0: Initial resonance concept (91.11% on synthetic Moons) v4.0: Sparse approximation surpasses KNN (98.89%) v7.0: Harmonic Forest ensemble proves superiority on periodic data v10.5: Alpha-Wave Specialist auto-evolution (96.45%) v12.0: Bipolar Montage preprocessing (+0.77 points—largest single gain) v13.0-v14.0: Full Holography captures final 1.10 points, crossing 98% v15.0: GPU acceleration + rigorous K-fold validation (98.84% mean) Medical-Grade Performance & Humanitarian Impact FDA-Ready Clinical Capabilities Seizure Detection: Phase-invariant onset detection regardless of timing Sleep Staging: Personalised Alpha/Theta boundary adaptation Anaesthesia Monitoring: Continuous depth-of-consciousness tracking Brain-Computer Interfaces: Multi-frequency sensorimotor rhythm decoding Global Healthcare Accessibility Resource Efficiency: Runs on consumer GPU hardware (e.g., NVIDIA RTX 3060) Scale Impact: 1% accuracy improvement = 500,000 fewer missed epilepsy detections annually Surgical Safety: Sub-second latency enables real-time awareness monitoring (8,500 fewer intraoperative awareness incidents across 234M annual procedures) Algorithmic Interpretability: Physically meaningful parameters (10 Hz = Alpha waves) build clinical trust Technical Architecture Core Mathematical Framework Ψ(x, p_i) = exp(-γ||x - p_i||²) · (1 + cos(ω_c · ||x - p_i|| + φ)) Gaussian Damping: exp(-γr²) controls spatial influence decay Harmonic Resonance: (1 + cos(ωr + φ)) encodes class-specific frequencies Energy Maximization: Classification = arg max_c Σ Ψ_c(query, training_points) Physics-Informed Preprocessing Bipolar Montage: Differential signal extraction cancels common-mode noise (voltage artefacts, body movement) Spectral Transformation: Fast Fourier Transform achieves mathematical time-shift invariance Robust Scaling: Quantile-based normalisation (15th-85th percentile) for outlier rejection Holographic Features: Concatenation of raw sensors, differentials, and global coherence Auto-Evolution Mechanism Validation-Based Grid Search: 20-30% hold-out data for parameter optimisation Physics-Informed Grid: Frequency (0.1-50 Hz), Damping (0.01-15), Neighbours (3-10) Neurological Convergence: Auto-evolved frequencies consistently align with the Alpha band (8-12 Hz), validating physical meaningfulness Unique Interdisciplinary Synthesis HRF exists at the unprecedented intersection of five scientific domains: Wave Physics: Damped harmonic oscillators, resonance, constructive/destructive interference Signal Processing: FFT spectral analysis, bipolar montage, artefact rejection Machine Learning: Classification theory, ensemble bagging, k-NN locality Neuroscience: Brainwave frequencies (Alpha/Beta/Delta/Theta/Gamma), EEG physiology Statistical Mathematics: Fourier analysis, stratified cross-validation, optimisation theory No existing framework combines these disciplines at this depth. Performance Benchmark Summary Model Configuration Test Accuracy Gap from HRF HRF v15.0 (Stable, K-Fold Validated) 98.84% Baseline HRF v14.0 (Ultimate) 98.46% -0.38% HRF v13.0 (Full Holography) 98.36% -0.48% HRF v12.0 (Bipolar Montage) 97.53% -1.31% Extra Trees (Chaos) 94.49% -4.35% Random Forest 93.09% -5.75% XGBoost 92.99% -5.85% Repository Contents Core Implementation hrf_final_v16_hrf.ipynb: Complete algorithmic evolution (v1.0 → v16.0) with inline documentation HRF EEG.pdf: Full technical manuscript with mathematical proofs, with a medical validation study on a real-world EEG corpus Documentation Readme HRF Main.md: Comprehensive project overview and the main repository documentation Readme HRF Paper.md: Research paper companion guide Key Features Scikit-Learn API: Full compatibility via BaseEstimator and ClassifierMixin GPU Acceleration: NVIDIA RAPIDS (cuML, CuPy) integration Ensemble Methods: Bagging with configurable estimators Visualisation Tools: Decision boundary plots, confusion matrices, survival curves Broader Applications Beyond EEG The physics-informed approach generalises to any domain with wave-like phenomena: Immediate Deployment Domains Audio Processing: Speech recognition, music classification, acoustic anomaly detection Seismic Analysis: Earthquake early warning, structural health monitoring Radar/Sonar: Target detection in noisy maritime/aerospace environments Industrial IoT: Vibration-based predictive maintenance, equipment failure forecasting Telecommunications: Signal decoding under phase noise and multipath fading Emerging Research Frontiers Quantum Computing: Qubit state classification under decoherence Financial Markets: Cyclical pattern detection in high-frequency trading data Climate Science: Oscillatory climate pattern identification (El Niño, NAO) Material Science: Vibrational spectroscopy analysis, crystal structure determination Future Research Horizon: v16.0 Experimental Beta Internal R&D has successfully developed v16.0 with record-breaking 98.93% peak accuracy through Parallel Evolutionary Search. Currently in experimental status due to localised confusion matrix variance, with ongoing work on "Resonance Smoothing" techniques for v17.0 stabilisation. Validated Performance: 5-Fold CV Mean = 98.51% (±0.24%), confirming the evolutionary trajectory continues upward while maintaining the production stability of v15.0 as the official benchmark. Why This Matters for AI Research Paradigm Shift Evidence First Principles Win: Physics-informed bias outperforms purely data-driven optimisation Interpretability Revolution: Parameters map directly to physical phenomena (10 Hz frequency ≠, arbitrary weight) Robustness Proof: Mathematical invariance (Fourier transform) beats empirical regularisation Validation Standard: GPU-accelerated K-fold CV sets new rigour bar for ML research Humanitarian AI: Medical-grade performance accessible on consumer hardware democratizes healthcare technology For Research Institutions (DeepMind, Anthropic, OpenAI, Meta AI) This work demonstrates that the next frontier of AI advancement lies not in scaling compute or data, but in encoding domain knowledge as algorithmic priors. HRF proves that when AI "listens to the physics of reality," it achieves capabilities fundamentally inaccessible to statistical learning alone—while remaining interpretable, efficient, and clinically deployable. Citation & Reproducibility Dataset OpenML ID: 1471 (EEG Eye State Corpus) Samples: 14,980 continuous EEG recordings Features: 14-channel sensor array (AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, AF4) Sampling Rate: 128 Hz Task: Binary classification (eyes open vs. closed) Computational Environment Hardware: NVIDIA GPU (CUDA 12.x compatible) Software: Python 3.11, NVIDIA RAPIDS (cuML v24.x, CuPy v12.x), scikit-learn Random Seeds: Fixed at 42 for all experiments
Abstract This comprehensive technical and strategic manifesto, authored by Abhijeet Sarkar of Synaptic AI Lab Private Limited, presents the architectural blueprint for "The Synaptic Ecosystem." This ecosystem represents a paradigm shift from traditional "Artificial" Intelligence to "Synaptic" Intelligence—a model of Organic Computation that prioritizes biological mimicry, plasticity, and civilizational sovereignty. Designed as an all-encompassing Earth Operating System, the platforms detailed herein address the foundational pillars of the modern nation-state: security, healthcare, finance, education, agriculture, and governance. By integrating proprietary technologies such as the Synapse-1 AGI kernel, Secure-ID (Blockchain/ZKP), IndicGPT (Brahmi-Net Tokenizer), and Vajra-Crypt (Post-Quantum Cryptography), Synaptic AI Lab aims to establish a resilient and autonomous digital infrastructure. This document serves as a technical disclosure of the lab's patent-pending methodologies and its strategic roadmap for the next decade, culminating in the Singularity Interface—the ultimate convergence of human consciousness and machine intelligence. Part I: The Foundation & The Shield (Sovereignty & Security) 1. The Synaptic Thesis: Beyond Binary Logic The current landscape of artificial intelligence is built upon the brittle foundation of binary logic and rigid neural architectures. At Synaptic AI Lab, under the leadership of Abhijeet Sarkar, we posit that the future of computation lies not in the further scaling of silicon transistors, but in the mimicry of the biological synapse. The Synaptic Thesis argues that intelligence is not merely the result of processing power, but the outcome of "Organic Computation"—a system characterized by high plasticity, asynchronous signal processing, and contextual modulation. Traditional AI models are static once trained; their weights are "frozen." In contrast, the Synaptic Ecosystem utilizes Active Plasticity Modules (APM). These allow the network to reorganize its topology in real-time based on environmental feedback, mirroring the way the human brain forms and prunes connections. This shift from "Artificial" to "Synaptic" Intelligence allows our systems to operate with the energy efficiency of a biological organism while maintaining the precision of a machine. This is the cornerstone of all products developed at Synaptic AI Lab. 2. The Core Architecture: Synapse-1 & The Neural Cloud Synapse-1 is the proprietary AGI kernel that serves as the central processing engine for the entire ecosystem. Unlike modular AI which treats tasks in isolation, Synapse-1 is built on a Multimodal Recursive Attention Mechanism (MRAM). This allows the kernel to perform "Cross-Domain Inference." For example, a linguistic pattern detected in IndicGPT can be used by Nyaya-Sahayak to better understand legal nuances, or a visual anomaly detected by Scan-Easy can inform the surveillance logic of Netra-X. The Neural Cloud provides the decentralized infrastructure required to host Synapse-1. To address the global energy crisis, Synaptic AI Lab has pioneered the Green Server Initiative. These data centers use immersion cooling and are powered entirely by renewable energy microgrids. Furthermore, we employ Dynamic Compute Offloading, where low-latency tasks are handled by edge devices, while high-entropy computations are pushed to the core, ensuring a 90% reduction in carbon footprint compared to traditional cloud providers. 3. Sovereign AI & Identity: The Secure-ID Protocol Data colonization by transnational tech giants has stripped nations of their digital sovereignty. Abhijeet Sarkar has identified the lack of a sovereign identity layer as the primary threat to national security. The Secure-ID Protocol is Synaptic AI Lab’s solution—a blockchain-based verifiable credentials system that uses Zero-Knowledge Proofs (ZKPs). The technical innovation here lies in our Selective Disclosure Architecture. A citizen can prove their eligibility for a government subsidy or their right to vote without revealing their name, address, or biometric data to the verifying party. The proof is mathematically generated and validated on a sovereign, permissioned ledger. This eliminates identity theft at the source and ensures that the state—and only the state—manages the identity of its citizens, free from external surveillance. 4. IndicGPT & The Linguistic DNA: The Brahmi-Net Tokenizer The Western bias in existing LLMs is not just a cultural issue; it is a technical failure. Standard tokenizers treat Devanagari, Tamil, or Telugu as outliers, breaking them into meaningless fragments that bloat the context window and degrade performance. IndicGPT, developed by Synaptic AI Lab, introduces the Brahmi-Net Tokenizer. Brahmi-Net is a morphology-aware tokenizer that understands the phonemic structure of Indic scripts. By treating ligatures and vowels as distinct semantic units, IndicGPT achieves a 3x higher information density than GPT-4 for Indian languages. This allows the model to capture 5,000 years of linguistic nuance, enabling it to perform everything from Vedic script analysis to real-time translation for a rural farmer in their specific dialect. 5. Infrastructure Immunity: Vajra-Crypt & Grid-Lock With the advent of quantum computing, the world faces "Q-Day"—the moment current encryption becomes obsolete. Vajra-Crypt is our patent-pending Post-Quantum Cryptography (PQC) suite. It utilizes lattice-based cryptography (Crystal-Dilithium equivalent) to secure diplomatic and military communications. Parallelly, the Grid-Lock system protects the physical nodes of the nation. It is an AI-driven Industrial IoT (IIoT) sentinel designed for power grids and water systems. Grid-Lock monitors the "Heartbeat" of industrial controllers (PLCs). If a cyber-attacker attempts to manipulate the frequency of a turbine or the flow of a valve, Grid-Lock detects the behavioral anomaly in milliseconds and enters a "Lockdown State," preventing catastrophic failure and ensuring national resilience. 6. The Autonomous Shield: Netra-X & Swarm-Rakshak Protecting porous borders in terrain like the Himalayas is a human impossibility. Netra-X is our edge-computing surveillance node. These units are solar-powered and feature thermal, acoustic, and visual sensors. The AI on-board ignores natural movements (animals, wind) but identifies human intrusions with 99.9% accuracy, even in zero-visibility conditions. Swarm-Rakshak provides the tactical response. It is a decentralized Operating System for drone swarms. Instead of a "master-slave" architecture which has a single point of failure, Swarm-Rakshak uses Leaderless Consensus Algorithms. If one drone is shot down, the rest of the "hive" instantly redistributes the mission parameters. This asymmetric defense platform ensures that the borders of the republic are defended by a persistent, intelligent, and unyielding autonomous shield. Part II: The Digital Hippocrates (Healthcare) 7. The First Mile: Rural-Doc & Jan-Awaaz Healthcare in the Global South suffers from a "Last Mile" problem—or as Abhijeet Sarkar calls it, the "First Mile Crisis." Most deaths occur because patients cannot access a doctor for an initial triage. Rural-Doc is a WhatsApp-integrated triage agent that uses Synapse-1’s medical reasoning module. It converses with patients in local dialects, assesses the severity of symptoms, and provides immediate guidance. Jan-Awaaz acts as the accountability layer. It is a voice-native grievance agent that listens to patients. It uses Emotional Prosody Analysis to detect distress or urgency in a caller's voice. If a primary health center is understaffed or a doctor is absent, Jan-Awaaz aggregates these voice reports into a real-time "Health Map" for the Ministry of Health, ensuring that administrative bottlenecks are cleared using real-world data. 8. Pocket Diagnostics: The Scan-Easy Revolution The lack of diagnostic labs in rural areas leads to late-stage detections of preventable diseases. Scan-Easy is a Computer Vision platform that turns any smartphone camera into a diagnostic tool. By analyzing the color and texture of the inner eyelid, it detects Anemia. By analyzing oral lesions, it flags early-stage Oral Cancer. The technical breakthrough in Scan-Easy is our Environment-Invariant Feature Extraction. The AI can account for poor lighting, low camera resolution, and varying skin tones, making it as reliable in a dusty village as it is in a sterile clinic. This democratizes diagnostics, putting a multi-million dollar lab in the pocket of every health worker. 9. The Planetary Immunity System: Epi-Watch Future pandemics will be stopped by data, not just vaccines. Epi-Watch is Synaptic AI Lab's bio-surveillance platform. It ingests non-traditional data streams: Google searches for "fever," pharmacy sales of paracetamol, and even noise levels in urban centers. By applying Spatiotemporal Predictive Modeling, Epi-Watch can identify an outbreak cluster weeks before a single patient is officially tested. This "Planetary Immunity System" allows for targeted interventions—like shutting down a specific block or air-dropping medical supplies—before a localized infection becomes a national crisis. 10. Drug Discovery at Light Speed: Generative Protein Folding Traditional drug discovery is a 10-year gamble. Synaptic AI Lab is compressing this to 10 months. Our Generative Protein Folding models simulate the molecular interactions between viruses and potential drug compounds in a virtual environment. Instead of trial and error in a wet lab, Synapse-1 generates millions of molecular structures and "folds" them to see which ones inhibit the target protein. This platform is currently being used to develop affordable treatments for neglected tropical diseases, proving that Synaptic Intelligence can solve problems that are not commercially viable for Western Big Pharma. 11. Mental Health & The Empathy Eng
This article examines the legal nature of smart contracts and their compatibility with the legal system of the Republic of Azerbaijan. Smart contracts are defined as a hybrid legal mechanism arising from the convergence of classical contract law and blockchain technology. The author argues that automated execution of contractual obligations significantly reshapes traditional legal concepts of consent and performance. The study provides a comparative analysis of international regulatory approaches to smart contracts, focusing on the United States, the European Union, and selected Asian countries. Key principles such as technological neutrality, functional equivalence, and human oversight are assessed. This comparative perspective highlights the growing role of smart contracts beyond purely commercial transactions. The article evaluates Azerbaijani legislation, including the Civil Code, the Law on Electronic Signature and Electronic Document, and the Digital Development Concept, as a normative foundation for smart contracts. It concludes that the existing legal framework offers sufficient grounds for recognizing smart contracts as legally valid electronic agreements. The author emphasizes the potential application of smart contracts in digital government, e-services, and public procurement as part of Azerbaijan’s broader digital transformation agenda
The introduction of information and communication technologies in the legal domain has enabled the automation of some activities in the legal profession. With the advent of blockchain and smart contracts, new tools have emerged for lawyers and their clients, enhancing transparency and increasing trust compared to traditional legal instruments. Once deployed, smart contracts should be able to respond to various events that can occur during the contract’s lifecycle. However, this kind of automation in smart contracts requires them to embed necessary legal knowledge and implement support for legal reasoning. In this paper, we propose a legal reasoning method for smart contracts that incorporates defeasible logic, a key requirement for automated reasoning in the legal domain. The entire reasoning process in our approach is performed on the blockchain infrastructure, making the drawing of conclusions fully transparent and accessible to all interested parties. To demonstrate our concept, we illustrate how certain rights prescribed under labour law can be embedded within a smart contract and deployed on the blockchain as a legal reasoning service. Then, we show how an employment contract can use the reasoning contract to automatically apply legal norms to infer conclusions and determine legal consequences in particular cases. We analyse the benefits and potential issues of this method and discuss directions for future work. Optimisation of the reasoning engine is one of the challenges we identified that needs to be tackled in future.
Hybrid AI systems that combine cognitive decision-making with physical actuation pose unprecedented challenges for governance, safety, and certification. AGORIA 3.4 presents a unified architectural framework that bridges cognitive governance (AGORIA v1.3) and cyber-physical governance (AGORIA v1.6) into a coherent, certifiable solution for safety-critical applications.Core Innovation: Controlled IgnoranceAGORIA introduces the principle of controlled ignorance—the deliberate, verifiable, and structural restriction of information accessible to each system layer beyond what is strictly required for its formal responsibility. This architectural invariant reduces cognitive coupling, limits attack surfaces, enables independent certification, and ensures that no single component can subvert the safety-governance chain. Zero-knowledge proofs provide cryptographic enforcement of this separation.Four-Layer ArchitectureThe framework organizes systems into four formally interconnected layers: STRATEGOS: Strategic cognitive governance via Choquet integral aggregation (~100 ms)GENESIS: Tactical planning with DSL→STL translation and ZK certificate generation (<50 ms)NEXUS: Bounded verification independent of semantic complexity (<250 µs)HSL++: Hardware-enforced physical safety via Control Barrier Functions (100 µs–kHz) Formal GuaranteesAGORIA provides four normative amendments with mathematical proofs: Robust Invariance via CBF with explicit feasibility hypothesis and statistically calibrated margins (Theorem 1)Bounded WCET Verification via succinct ZK proofs on constrained platforms (Proposition 2)Practical Stability under Lipschitz-continuous governance parameter variation (Theorem 2)Conservative Semantic Bridge from domain-specific language to decidable STL fragment (Theorem 3) Validation and CertificationExperimental validation covers three scenarios: autonomous vehicle (1000 trials), surgical robot (500 trials), and multi-agent factory (100 trials). AGORIA achieves zero safety violations while maintaining bounded worst-case execution time (<250 µs). A case study on multi-source aeronautical navigation (ILS, VOR, GBAS, DME) demonstrates framework genericity.The architecture enables hybrid certification compatible with: Functional safety standards (ISO 26262 ASIL D, IEC 61508 SIL 3)AI regulatory requirements (EU AI Act 2024/1689, UL 4600)Industrial cybersecurity (IEC 62443) Related Publications AGORIA v1.3: Cognitive Governance (Zenodo, 2025)AGORIA v1.6: Cyber-Physical Governance (Zenodo, 2025)
Michael G. Xevgenis, Maria Polychronaki, Dimitrios G. Kogias, Helen C. Leligkou · 5 authors
Zero-Touch Network (ZTN) represents a cornerstone approach of Next Generation Networks (NGNs), enabling fully automated and AI-driven network and service management. However, their distributed and multi-domain nature introduces critical security challenges, particularly regarding service identity and data integrity. This paper proposes a novel blockchain-based framework to enhance the security of ZTN through two complementary mechanisms: decentralized digital identity management and oracle-assisted network monitoring. First, a Decentralized Identity Management framework aligned with Zero-Trust Architecture principles is introduced to ensure tamper-proof authentication and authorization in a trustless environment among network components. By leveraging decentralized identifiers, verifiable credentials, and zero-knowledge proofs, the proposed Decentralized Authentication and Authorization component eliminates reliance on centralized authorities, while preserving privacy and interoperability across domains. Second, the paper investigates blockchain oracle mechanisms as a means to extend data integrity guarantees beyond the blockchain, enabling secure monitoring of Network Services and validation of Service-Level Agreements. We propose a four-dimensional framework for oracle design, based on qualitative comparison of oracle types—decentralized, compute-enabled, and consensus-based—to identify their suitability for NGN scenarios. This work proposes an architectural and design framework for Zero-Touch Networks, focusing on system integration and security-aware orchestration rather than large-scale experimental evaluation. The outcome of our study highlights the potential of integrating blockchain-based identity and oracle solutions to achieve resilient, transparent, and self-managed network ecosystems. This research bridges the gap between theory and implementation by offering a holistic approach that unifies identity security and data integrity in ZTNs, paving the way towards trustworthy and autonomous 6G infrastructures.
The convergence of artificial intelligence (AI), machine learning (ML), blockchain, and big data analytics is transforming the governance, sustainability, and resilience of modern banking ecosystems. This study provides a multivariate bibliometric analysis using Principal Component Analysis (PCA) of research indexed in Scopus and Web of Science to explore how decentralized digital infrastructures and AI-driven analytical capabilities contribute to sustainable financial development, transparent governance, and climate-resilient digital societies. Findings indicate a rapid increase in interdisciplinary work integrating Distributed Ledger Technology (DLT) with large-scale data processing, federated learning, privacy-preserving computation, and intelligent automation—tools that can enhance financial inclusion, regulatory integrity, and environmental risk management. Keyword network analyses reveal blockchain’s growing role in improving data provenance, security, and trust—key governance dimensions for sustainable and resilient financial systems—while AI/ML and big data analytics dominate research on predictive intelligence, ESG-related risk modeling, customer well-being analytics, and real-time decision support for sustainable finance. Comparative analyses show distinct emphases: Web of Science highlights decentralized architectures, consensus mechanisms, and smart contracts relevant to transparent financial governance, whereas Scopus emphasizes customer-centered analytics, natural language processing, and high-throughput data environments supporting inclusive and equitable financial services. Patterns of global collaboration demonstrate strong internationalization, with Europe, China, and the United States emerging as key hubs in shaping sustainable and digitally resilient banking infrastructures. By mapping intellectual, technological, and collaborative structures, this study clarifies how decentralized intelligence—enabled by the fusion of AI/ML, blockchain, and big data—supports secure, scalable, and sustainability-driven financial ecosystems. The results identify critical research pathways for strengthening financial governance, enhancing climate and social resilience, and advancing digital transformation, which contributes to more inclusive, equitable, and sustainable societies.
Address verification and spend limit checks in card and instant payment systems expose cardholder ZIP codes and velocity histories to gateways and intermediaries, which increases the privacy impact of breaches. At the same time, issuers rely on these checks to control fraud within strict two to three hundred millisecond authorization budgets. This paper presents ZK-AVS, a design that replaces plaintext AVS and spend limit evaluation with zero knowledge proofs produced on the customer device and verified at the payment gateway. The issuer stores commitments to the cardholder ZIP and per window spend counters, and the device proves that its ZIP matches the committed value and that the proposed transaction keeps cumulative spend within the configured limit, without revealing the underlying values. We instantiate a synthetic workload of fifty thousand transactions and show that AVS mismatch and limit pressure carry useful fraud signal that can be exposed as proof outcomes. The design is structured to fit within sub three hundred millisecond budgets on commodity devices, while removing direct exposure of address and spend history data at the gateway and keeping changes largely at endpoints.
This article examines the probative value of blockchain (distributed ledger technology) in legal proof. It highlights the technology’s core features—decentralization, immutability, cryptography, and time-stamping—and assesses how they fit within rules of evidence, particularly the requirements of electronic writing and electronic signature. The study also discusses the extent of legislative recognition in Morocco and comparative systems, with a focus on identity attribution and the link between a digital record and its author. It concludes that blockchain records may carry increasing persuasive force, while full evidentiary equivalence requires clearer regulatory frameworks and trusted digital services to ensure integrity and reliability.
This paper proposes a Decentralized Autonomous Intelligence (DAI) architecture that overcomes the self-referential limitations of conventional AI and Web3 systems by dynamically grounding collective intelligence in physical reality. By coupling internal consensus with high-fidelity external data such as environmental, biological, and economic signals, the framework prevents value drift, Sybil manipulation, and speculative bias. The result is a reality-aligned, secure, and scalable intelligence system optimized for real-world utility and immediate deployment.