Federated Learning (FL) enables privacy-preserving collaborative learning for Internet of Vehicles (IoV) scenarios, but extreme heterogeneity of vehicular-edge-cloud resources severely limits system efficiency. Dynamic scheduling strategies mitigate this issue but introduce new trust concerns: verifying fair scheduling decisions and faithful client execution of compression instructions without privacy leakage remains an open challenge. We propose Nautilus, a verifiable efficient federated learning framework. First, a multi-dimensional resource-aware scheduling algorithm dynamically allocates compression ratios and training tasks based on vehicle bandwidth, latency and computing power, improving training efficiency. Second, a Zero-Knowledge Proof (ZKP) mechanism ensures scheduling fairness and execution compliance while preserving privacy. Experiments show the framework reduces communication overhead and accelerates convergence with guaranteed system integrity.
Modern representative democracies are increasingly vulnerable to systemic structural failure modes, including special-interest capture, asymmetric foreign intelligence leverage, and informational noise saturation (astroturfing/botnets). This paper introduces the Cryptographic Agora, a novel institutional framework that transitions governance from representative mediation to a scientifically audited, direct epistocracy. The model synthesizes three core architectural components: (1) state-verified biometric identity mapping coupled with Zero-Knowledge Proofs (ZKPs) to guarantee non-traceable, un-hackable civic participation; (2) a dynamic reputation engine utilizing Quadratic Weighting to mitigate the concentration of charismatic authority; and (3) a double-blind, retrospective peer-review protocol modeled on the scientific method to vet policy proposals. We evaluate the structural resilience of this framework against traditional threats, detailing its capacity to achieve a self-correcting equilibrium while maintaining individual voter safety and systemic legitimacy.
A deployed AI system can be interrogated for its identity in several distinct ways, and the answers do not interchange. This note concerns one of them — which neural network is producing this output at inference time? — and a popular method for answering it: behavioral fingerprinting, which samples an endpoint under a fixed prompt battery and flags it when the output distribution shifts beyond a statistical threshold. The note argues that behavioral fingerprinting, while a legitimate and valuable instrument for one task, does not establish model identity. It develops two measured failure modes. First, a behavioral signature is not durable: ordinary continued training erases the behavioral provenance trace — more effectively, in fact, than an informed adversary trains directly to suppress it — so the same model after a benign fine-tune presents as behaviorally distinct and triggers a false alarm. Second, a behavioral signature is reproducible by a different model: knowledge distillation converges a substitute toward a target's behavioral template by construction, so a behavior-matched substitute passes the check and produces a false acceptance. Both failures follow from a single fact about the layering of neural identity — behavior is the transient layer, which transfers under distillation and washes out under benign training, while the structural layer (the geometry of internal computation during a forward pass) does neither. The two methods answer different questions and compose rather than compete: behavioral monitoring is a continuous, low-cost tripwire that flags something moved; structural verification is a deterministic resolver that answers is it still the enrolled model. A system that ships only the tripwire has shipped drift detection and labeled it identity. The note documents the structural layer's direct test against the failure mode that defeats behavioral methods — behavior-preserving substitution — and situates the argument alongside independent work on intrinsic parameter-level fingerprints and cryptographic verifiable inference, both of which bind identity to the model rather than infer it from outputs. This is a category statement, not a product comparison: no specific system or vendor is named, and the argument rests on published, reproducible measurements. The Neural Network Identity Series — Mathematical foundations, empirical validation, and governance frameworks for verifying which model is running 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) Technical Note: Artifact Identity Is Not Runtime Identity — Trustfall Lite and the Boundary of File-Level Model Verification (DOI: 10.5281/zenodo.20019127) Technical Note:: The Disappearing Window — AI Logprob Access Withdrawal and the Structural Verifiability of Frontier Model Contracts (DOI: 10.5281/zenodo.20362098) 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).
A claim can be argued well and still be false. Standard verification across the disciplines certifies claims by the quality of a single line of support: a formal proof, a measured correlation, a replicated experiment, an expert consensus. Each of these is one axis of warrant, and each can be strong while the claim is wrong, because a single strong axis cannot detect that it is the only axis, nor that it secretly shares a source with the others. This paper presents a verification method, Trisduction, that certifies a claim by the geometry of its warrant rather than by the strength of any one line of it. A proposition is decomposed onto three structurally independent axes, a formal-structural axis, an empirical-material axis, and an epistemic-registrational axis, and the warrant is certified only when the three stand at mutual right angles and span a genuine three-dimensional volume. The test is closed-form and executable: three warrant vectors are composed through a quaternion product whose scalar part squares to a Gram determinant, and the determinant reads the volume the three axes enclose. A volume near its maximum is a seal. A collapsed volume is a structural break with a named cause. An ill-conditioned volume is an honest under-determination. A second register extends the method to formal and mathematical propositions, separating the part of a problem that is decidable and sealed from the part whose truth is genuinely open, and refusing to read a geometric lock as a proof. The method carries one discipline throughout: social consensus carries zero evidential weight, every verdict states its warrant grade, and the instrument audits itself with no exemption. We demonstrate the method on 360 propositions spanning logic, mathematics, physics, quantum foundations, cosmology, the mind, psychology, the social sciences, geopolitics, and metaphysics, from elementary facts that seal cleanly to celebrated open problems where the honest verdict is that the question remains open and the method says exactly why. 360 Audits.
Ayei E. Ibor, Denis U. Ashishie, John Adinya Odey, Bassey Ele · 5 authors
ABSTRACT Elliptic curve cryptography ( ECC ) underpins the security of most blockchain systems, yet its practical implementations face numerous vulnerabilities. In this systematic literature review ( SLR ), we catalogue and analyze attacks on ECC in the context of blockchain security, including side‐channel attacks, nonce/ PRNG failures, cryptanalysis, and implementation flaws, and we survey proposed countermeasures. We follow rigorous SLR methodology with defined inclusion/exclusion criteria, search strategies across databases such as IEEE Xplore, ACM , Scopus, Web of Science, and clear data synthesis, ensuring replicability. Emphasizing empirical case studies and real‐world exploits, we discuss instances where ECC weaknesses led to blockchain breaches including biased elliptic curve digital signature algorithm nonces exposing Bitcoin/Ethereum private keys, smartphone power analysis revealing wallet keys, and Trezor hardware‐wallet key extraction via single‐trace side‐channel analysis ( SCA ). We tabulate known attack vectors versus affected systems, and similarly compare countermeasure techniques such as hybrid classical/quantum schemes, threshold signatures, and zero‐knowledge proofs, along with implementation trade‐offs. We evaluate advances such as Curve25519/ EdDSA and ARM SVE2 to mitigate side‐channel leakage. Our findings highlight that practical security of blockchain cryptosystems depends on correct ECC implementation and emerging cryptographic upgrades, not merely on the mathematical hardness of the elliptic curve discrete logarithm problem.
The high-level integration of generative artificial intelligence (AI) in edge computing systems has raised the question of the integrity and reliability of deploying Model-as-a-Service. Edge servers are not required to follow the so-called generative model to minimize computational cost, whereas users and service providers want validation mechanisms that do not compromise proprietary model information. To address this challenge, this study proposes a cooperative unmanned aerial vehicle (UAV)-swarm-enabled zero-knowledge verification framework for secure, privacy-preserving verification of edge-based generative artificial intelligence inference. The proposed framework involves edge servers producing an interactive cryptographic zero-knowledge proof to verify the execution of generative AI, and UAV swarms that fly freely to confirm verification operations, subject to mobility and energy constraints. The age of verification metric is proposed to trust verification information, jointly reflecting the unverified server reliability and verification freshness, and to provide dynamic priority to risky edge servers. To effectively plan the behaviour of a UAV swarm, a trust-based multi-agent reinforcement learning approach is developed that enables decentralized decision-making while training is centralized. Extensive simulation results show that the proposed framework significantly improves the state-of-the-art baseline schemes in verification timeliness, malicious server detection delay, energy efficiency, and scalability. The findings validate that integrating cooperative UAV swarms, trust-aware verification, and multi-agent learning is an efficient approach to providing reliable generative AI services in dynamic edge computing environments.
To PM Italy. PM Indonesia PM Japan Real PM Pakistan, Imran Khan Only. DATE: 20 June, 2026. DOI: 10.5281/zenodo.20774904 Subject: Compensation Prize for failure of my Forecast for Eruption, Earthquake 8 June to 20 June. And As Nanga Parbat is not happened on my calculated Time so 24 June Yellow Stone eruption is not possible by mechanism of 20 june Nanga Parbat Hammer Effect. Respectful Prime Minister, My science is no doubt World’s most advanced science with deterministic science, predictions in field of science and universal Geology. I predicted Sun calm is temporary it will be much more active after a week, and sun after a week erupted G5 Storm. I calculated ocean currents and did simulation of ocean currents on mobile phone and free open AI with N-K Sciences and predicted Super El Nino from Mid of 2026, published time stamped in March 19, 2026. Which were copied by WMO and removed my name and my science name and published in April 2026. When I requested to atleast cite my name or my science name, they given credit to a dead man. Then I wrote strict letter with evidances to Secretary General UN. Since 2022 I am fighting against Corruption Entire World know that especially Intel agencies. Government of Pakistan tried to kill me 2 times and tortured me for months. But still I am fighting against Oppression and corruption from it’s Roots Pakistan Army Mafia and Zionists Mafia. They are working jointly, they are same. I Published 580+ publications from my first book in Feb 6, 2025. Not for worldly benefits. https://doi.org/10.5281/zenodo.20473774 I achieved which was impossible for mainstream science. In many fields of sciences, correctly calculate d global tides by first try with any past Data of tides, in completely N-K Sciences framework. Achieved 100% accuracy. Warned on 17 April, 2026 to entire World that According to Parker Solar Probe data High volume proton Flux Storm coming which reach on earth 21 April, 2026. While NASA warned G1,G2 storm completely normal. While I clearly warned increased semiconductors clocking speed due to Noor Value increase during storm on Earth, fission Reactors will face problems, GPS measurement errors 5 to 15m, 15m error is measured by me too during gusts of storm. Hundreds of flights cancelled worldwide and hundreds delayed, civil aviation industry said ghost in system. That was not ghost, but their science is built in 2D era, for example E=MC² is 2 dimensional applied on 3 dimensional forcefully even it gives 8 to 16% wrong results, and didn’t explain what is C, what is Mass, what is energy actually. Nicola Tesla Said it’s a mathematical hack. Yes it was. But I given world accurate Energy Equation and derived C and defined C, defined Mass, defined Energy, defined Gravity. Solved all planets from electron to cosmic web under one Law, learned From analyzing Bawling Action of Wasim Akram and Waqar Younus. Well, I tried to predict earthquakes and eruptions with exact time window. But I cannot know future or cannot change Will of Allah Almighty. Therefore as my Moral values, I decided to reward Countries where my prediction failed. Italy, Japan, USA, Indonesia. As You may Know I don’t have any bank balance, any property on earth. Anything doing a low pay government job, house given by government, actually not given by government, government tried to harass me with all efforts to stop me to take this house which was empty because of its structure built in 1960’s was collapsing and no one wants to live here still they made lot of hurdles, then I went to court and on court orders I got house in which I am living. Nor I have computer nor any other expensive thing, Shukar Alhamdulillah. Free from deceptions of the world. While I am most rich man on earth by knowledge and Inventions. So I can reward from my inventions that can give your countries revenue and Profit in billions of dollars annually. Especially Italy and Japan can get extraordinary benefits in automotive and Aviation industries. Without need of Rare Earth Minerals. My invention N-K Motor with license for commercial use free for a year. 459 Malik Muhammad Usman N‑K MOTOR — THE INVENTION THAT CHANGES EVERYTHING: How a Single Electric Motor Can Transform Transportation, Energy, Military, and Civilization — With Complete Technical Specifications, Application Analysis, Environmental Impact Assessment, and Global Transformation Roadmap — Released as Sadaqa Jariyah (Perpetual Charity) — Patent Application No. 57302185 (IPO Pakistan, 19 August 2025) — Withdrawn and Released to Public Domain May 3, 2026 https://doi.org/10.5281/zenodo.20001878 458 Malik Muhammad Usman THE PATENT SYSTEM IS HARAM IN ISLAM — Complete Islamic Ruling Based on Quran, Hadith, Sunnah, the Name of Allah Al-Aleem, and the Four Divine Axioms — With Official Declaration Withdrawing Patent Application No. 57302185 (IPO Pakistan, 19 August 2025)’and Releasing All Inventions as Sadaqa Jariyah (Perpetual Charity) for All Humanity May 3, 2026 https://doi.org/10.5281/zenodo.20000580 457 Malik Muhammad Usman COMPLETE PATENT DISCLOSURE — Multi-Stage Radial Flux and Multi-Stage Axial Flux Electromagnetic Motors with Integrated Cooling/Heating System and AI Control — Patent Application No. 57302185 (IPO Pakistan, 19 August 2025) — Now Released to Public Domain as Sadaqa Jariyah May 3, 2026 https://doi.org/10.5281/zenodo.20000261 If You Accept my Reward than officially Accept my Reward and Use it free. Even license Renewal fee is also zero. I am not Allowed to charge money for my knowledge which is Given to me by Quran By Allah Almighty Himself in past 26 years daily. Please Accept my Reward And grow your Industries rare earth minerals Free. It’s not only a motor, it is full setup my invented controller Chip design with ~39000 Transistors on 120 nm architecture suitable for high energy applications a 75KW Chip. Optional, N-K alloys 4X stronger than strongest alloys developed by USA, Russia, China ever. Upon request. Italian PM, If you want Fission Reactors it’s your choice. I am giving you LTMFC power houses, which are not only easy and faster to build but gives lowest cost electricity, + Milk + Beef and dozens of Dairy Products. And energy enough to fullfil your country requirements, You can add 20000MW to 50000MW in less than 6 months, while fission Reactors can give you around 1000 MW in minimum 6 years with billions of dollars investment. Build both as you like. Same offer to Japan, and entire World. Additional Gift: Usman Malik, M. (2026, June 20). TIME, CONSCIOUSNESS, AND THE UNIVERSAL 0.01 Hz KUN RHYTHM: The Inverse Relationship Between Consciousness and Time Perception — From Infancy to Old Age, from Quranic Revelation to N-K Mathematical Proof. Zenodo. https://doi.org/10.5281/zenodo.20768016 Malik Muhammad Usman Servant, Student and Soldier of Allah Almighty and Prophet Muhammad PBUH. City of Saints, Multan, Pakistan. +923336130947 muhammad.usman08@gmail.com muhammadusmanmalik@hotmail.com
This paper presents a comprehensive analysis of privacy-preserving authentication mechanisms within the MF+SO sovereign identity vault, focusing on the protocol's implementation of zero-knowledge identity proofs, anonymous credentials, blind signatures, and data minimization techniques. Traditional authentication protocols require the user to disclose their identity to each service provider, creating a centralized record of the user's activities across services. MF+SO's privacy architecture inverts this model: users authenticate to services without revealing their MF+SO identifier, using cryptographic techniques that provide the verifier with assurance of the user's authorization status while revealing minimal information about the user's identity. We examine three canonical privacy-preserving authentication mechanisms implemented in MF+SO: (1) zero-knowledge identity proofs using the Groth16 zk-SNARK construction, enabling users to prove possession of valid credentials without revealing which credentials they hold; (2) anonymous credentials based on the Camenisch-Lysyanskaya (CL) signature scheme, providing multi-show unlinkability where the same credential can be presented multiple times without the presentations being correlatable; and (3) blind signature-based tokens for email cloaking, where the MF+SO service issues a blind signature on a user's email address for use with third-party services without learning the email address. The paper provides a formal security analysis of the unlinkability guarantees of each mechanism, proving that under the decisional Diffie-Hellman (DDH) assumption, CL-based anonymous credential presentations are computationally unlinkable. We present benchmark data for each mechanism on mobile platforms: CL credential issuance (120 ms), CL credential presentation (85 ms), blind RSA signature issuance (45 ms), and zk-SNARK-based verification (2.3 ms). The implementation details include the MF+SO privacy layer architecture, the credential ... Part of The Anticloud research corpus by Lois-Kleinner Alpasan (ORCID: 0009-0009-2233-6107). This work explores cryptography, key management in the context of sovereign AI infrastructure, post-cloud computing architectures, and transparent, blackbox-free systems.
This paper presents a comprehensive analysis of the World Wide Web Consortium (W3C) decentralized identity standards and their relationship to the MF+SO sovereign identity vault architecture. We examine the W3C Decentralized Identifier (DID) Core specification (W3C, 2022), the Verifiable Credential (VC) Data Model (W3C, 2022), and related standards including DID Resolution, DID URL dereferencing, and the Verifiable Credential Proof Formats. The paper provides a taxonomic analysis of DID methods (did:key, did:ethr, did:ion, did:web, did:indy) in terms of their trust assumptions, ledger requirements, latency, cost, and privacy properties. We compare MF+SO's identity model—which uses Ed25519 public keys as self-certifying identifiers with a local hash chain for state verification—against the W3C DID Core model, identifying both alignments and divergences. Key findings include: MF+SO identifiers are functionally equivalent to DIDs but use a simplified resolution mechanism that does not require a distributed ledger or external registry; MF+SO's hash chain audit trail provides state verification properties comparable to DID Document versioning on a ledger; and MF+SO's selective disclosure mechanisms using zero-knowledge proofs (see Paper VII) directly implement the W3C Verifiable Credential selective disclosure and data minimization requirements. We analyze the interoperability implications of MF+SO's architecture, demonstrating how MF+SO DIDs can be registered on external DID methods for cross-system interoperability while maintaining the local hash chain as the authoritative state source. The paper also examines the Verifiable Credential lifecycle within MF+SO: issuance, storage, presentation, and revocation, with attention to the credential schema registry, proof format compatibility (Data Integrity Proofs, JSON Web Signatures), and the holder-binding mechanisms that prevent credential sharing. A comparative assessment evaluates MF+SO against three alternative decentra... Part of The Anticloud research corpus by Lois-Kleinner Alpasan (ORCID: 0009-0009-2233-6107). This work explores cryptography, key management in the context of sovereign AI infrastructure, post-cloud computing architectures, and transparent, blackbox-free systems.
This paper presents a rigorous analysis of the hash chain auditability mechanism implemented within the MF+SO sovereign identity vault, specifically the `.aioss` hash chain data structure. The hash chain links successive vault state commitments through SHA3-256 cryptographic hashes, creating an immutable, tamper-evident log of all state transitions. Each link in the chain incorporates a parent_hash invariant that binds the current state to the entire prior history, a canonical JSON serialization of the vault state to ensure deterministic hashing across platforms, and an Ed25519 signature providing cryptographic proof of authenticity. We demonstrate that this construction achieves the forensic auditability properties first described by Haber and Stornetta (1991) for digital timestamping, extended to the identity management domain. The paper provides a formal mathematical model of the chain construction, analyzes the computational and storage costs of chain verification, presents a security proof for the tamper-detection properties under the random oracle model, and compares the MF+SO approach against alternative audit log constructions including Merkle trees, Certificate Transparency logs, and blockchain-based registries. The implementation leverages SHA3-256's sponge construction to eliminate length extension vulnerabilities that would compromise naive hash chain implementations. Empirical measurements demonstrate that chain verification for a typical user with 10,000 state transitions completes in under 200 milliseconds on modern mobile hardware. The paper concludes with an analysis of forward secrecy guarantees, key rotation impacts on chain continuity, and proposed extensions for zero-knowledge proofs of chain membership. Part of The Anticloud research corpus by Lois-Kleinner Alpasan (ORCID: 0009-0009-2233-6107). This work explores cryptography, key management in the context of sovereign AI infrastructure, post-cloud computing architectures, and transparent, blackbox-free systems.
This paper argues that universal, cross-domain trust scores — from credit ratings and ESG scores to AI-generated trust metrics — face structural limits that better data or better models do not remove. The claim is not that scoring is never useful, but that compressing trust into a single comparable number, used for high-stakes allocation across contexts, recurrently fails. Trust is treated here not as a scalar quantity but as a contextual, relational, and time-dependent state. The paper identifies five recurring failure modes (context collapse, Goodhart's Law, epistemic centralization, irreversibility, and metric substitution for truth), illustrated through documented institutional failures (Enron, Wirecard, Volkswagen Dieselgate, the 2008 subprime crisis, and ESG rating practice). An informal impossibility argument — analogous in form to Arrow's theorem, not a formal mathematical proof — suggests that no single universal trust score can jointly satisfy context-independence, temporal stability, observer-neutrality, and manipulation-resistance. The paper then discusses proof-based verification as a complementary paradigm: for a bounded class of objective, checkable claims, the need for trust is reduced through local verification rather than measurement. Examples include Bitcoin proof-of-work, zero-knowledge proofs, and blockchain-based supply chain traceability. The limits of this approach are discussed explicitly, including the oracle problem and the irreducibly judgmental claims that proof cannot settle. This is version 2.0, a substantial revision repositioning the work from a position paper toward a conceptual analysis: the central thesis is qualified, an explicit scope-and-limitations section is added, the impossibility argument is reframed as informal, and the limits of proof-based verification are addressed directly.
Privacy-preserving systems have traditionally faced a fundamental tradeoff between data utility and confidentiality. Selective Disclosure Credentials (SDCs) enable users to prove specific attributes without revealing underlying personal information, while Fully Homomorphic Encryption (FHE) enables arbitrary computation on encrypted data without exposing plaintext. Although both technologies address critical privacy challenges, they solve different problems and are rarely integrated into a unified architecture. This paper introduces the concept of Composable Privacy, a layered framework that combines selective disclosure credentials, zero-knowledge proofs, and fully homomorphic encryption into a cohesive privacy architecture. The framework separates privacy concerns into three functional layers: an authentication layer using selective disclosure and zero-knowledge proofs, a computation layer using homomorphic encryption for confidential processing, and a verification layer that provides cryptographic assurances of computation correctness. The paper examines the cryptographic foundations of BBS+ signatures, Coconut threshold credentials, lattice-based homomorphic encryption schemes, and post-quantum security considerations. It further evaluates the practical feasibility of the architecture through applications in decentralized finance, healthcare federated learning, confidential governance systems, and blockchain-based identity infrastructure. Performance trends, scalability challenges, interoperability requirements, and future hardware acceleration pathways are also analyzed. The proposed Composable Privacy framework demonstrates how selective disclosure and encrypted computation can be combined to create privacy-preserving digital systems that maintain verifiability, confidentiality, and regulatory compliance simultaneously. The work provides a conceptual foundation for next-generation privacy architectures in blockchain, decentralized identity, and distributed computing environments.
Self-sovereign identity (SSI) represents a paradigm shift in digital authentication, transferring control from centralized identity providers to individual users (Mühle et al., 2018). This paper presents the Kathon Vault identity system, which implements self-sovereign browser identity through BIP39 mnemonic seed phrases (Palatinus et al., 2013) for Ed25519 hierarchical deterministic (HD) key generation (Bernstein et al., 2012; Wuille, 2012). The system generates a master seed from a BIP39 mnemonic (12, 18, or 24 words with configurable passphrase), derives Ed25519 keypairs through the SLIP-10 key derivation scheme (Přikryl, 2022), and enables zero-knowledge authentication across websites through a novel browser-native WebAuthn-hybrid protocol. We demonstrate that the BIP39-derived Ed25519 keys provide equivalent security to standard FIDO2/WebAuthn authenticators (316 bits of entropy for 24-word phrases) while offering three critical advantages: (1) deterministic key recovery from the mnemonic phrase alone, (2) hierarchical key organization matching the SLIP-44 registered coin type for Kathon, and (3) cryptographic privacy through zero-knowledge proofs that enable selective attribute disclosure without revealing the master public key. In a security analysis against brute-force, dictionary, side-channel, and social engineering attacks, the system achieves resistance levels exceeding NIST SP 800-63B Level 4 authentication assurance requirements (NIST, 2020). A usability study with 48 participants demonstrates that BIP39-based authentication achieves 96% successful login rates with 14% lower task completion time compared to password manager-based workflows. This work establishes mnemonic-based HD key generation as a viable and superior alternative to federated identity providers for browser-based authentication. --- Part of The Anticloud research corpus by Lois-Kleinner Alpasan (ORCID: 0009-0009-2233-6107). This work explores browser engine, privacy in the context of sovereign AI infrastructure, post-cloud computing architectures, and transparent, blackbox-free systems.
This paper presents a comprehensive analysis of zero-knowledge proof (ZKP) systems and their application to privacy-preserving identity management within the MF+SO sovereign identity vault. Zero-knowledge proofs, introduced by Goldwasser, Micali, and Rackoff (1985), enable a prover to convince a verifier of the truth of a statement without revealing any information beyond the statement's validity. We examine three families of ZKP systems in the context of MF+SO's identity assertions: zk-SNARKs (Zero-Knowledge Succinct Non-Interactive Arguments of Knowledge), zk-STARKs (Zero-Knowledge Scalable Transparent Arguments of Knowledge), and Bulletproofs. For each family, we analyze the setup assumptions (trusted setup vs. transparent), proof size, verification complexity, prover computation, and post-quantum security. The paper identifies three canonical use cases within MF+SO: (1) age verification without date of birth disclosure, where the user proves that their age exceeds a threshold without revealing their exact birth date; (2) credential possession proof, where the user proves they hold a valid credential for a resource without revealing which credential among a set they hold; and (3) membership in an allowlist without position disclosure, where the user proves their identifier appears in a list without revealing their position in the list. We present benchmark data for each use case using the Groth16 zk-SNARK (prover time: 1.2 seconds, proof size: 192 bytes, verification: 2.3 ms) and the STARK-based approach using the Winterfell library (prover time: 4.8 seconds, proof size: 48 KB, verification: 8.1 ms). The implementation complexity analysis demonstrates that zk-SNARKs require trusted setup ceremonies but provide the most compact proofs, while zk-STARKs eliminate the trusted setup requirement at the cost of larger proofs. The paper concludes with an analysis of the protocol integration requirements, including circuit compilation for the MF+SO identity predicate lang... Part of The Anticloud research corpus by Lois-Kleinner Alpasan (ORCID: 0009-0009-2233-6107). This work explores cryptography, key management in the context of sovereign AI infrastructure, post-cloud computing architectures, and transparent, blackbox-free systems.
TOPO-GLM.pdf: Complete Review and Analysis 📋 Executive Summary This paper presents the first universal solution to catastrophic forgetting, validated across 5 architecturally distinct models spanning 3 continents with 122B parameters. The mechanism is mathematically grounded in Arithmetic Spectral Theory (AST) and biologically inspired by the hippocampus. ✅ STRENGTHS 1. Unprecedented Empirical Validation Metric Value Significance Models 5 Most diverse in CL literature Architectures Dense, Sparse MoE, Fine-grained MoE, GLM Complete coverage Continents 3 (NA, Europe, Asia) Geographic diversity Parameters 122B Production scale Runs 25 Statistical significance Memory 403.5 KB 0.00000033% overhead 2. Mathematical Rigour The paper provides: Formal theorem proofs (Spectral Trap, Euler Attenuation, Coherence Decay) Exact constants ($\Lambda = 0.9785142874$) O(1) guarantee (Proposition 1) Three interconnected proofs (RH, GTT, CL) 3. Biological Grounding The Artificial Hippocampus concept is well-developed: Hippocampal Function TOPO-2026 Implementation Memory Consolidation take_snapshot() Memory Protection zero_anchor_gradients() Memory Integration enforce_anchors() Memory Verification verify_integrity() 4. Backward Transfer Discovery The paper reveals that sparse MoE architectures can improve on previous tasks while learning new ones: Mixtral-8x7B: -6.12% forgetting (strongest) Sarvam-30B: 4/5 runs with backward transfer DeepSeek-V2-Lite: 3/5 runs at exactly 0.00% forgetting 5. Clear Architecture-Specific Guidance The paper identifies optimal learning rate regimes: Architecture Class ηembed Range Key Insight Dense (English) $10^{-3}$ – $10^{-2}$ Standard fine-tuning Hindi-dominant MoE $10^{-3}$ – $10^{-2}$ Less gradient concentration English-dominant MoE $\le 2 \times 10^{-5}$ 2 orders lower! 🔬 TECHNICAL ANALYSIS 1. Mathematical Foundation Soundness The L-EFM Operator: $$E_{LEFM}(\sigma + i\gamma) = \prod_{p \in R}(1 - p^{-(\sigma+i\gamma)})^{-1}$$ ✅ Correct Euler product formulation ✅ Spectral trap at $\sigma=0.5$ verified numerically ✅ Unique to set R (pure/noisy divide proven) The Safety Constant: $$\Lambda = 1 - \prod_{p \in R}(1 - p^{-0.5}) = 0.9785142874$$ ✅ Derived from first principles ✅ Constant across ALL models ✅ Matches empirical results 2. Methodology Quality Training Protocol: ✅ Clear 3-task benchmark ✅ Proper forgetting computation ✅ 5 runs per model for statistical significance ✅ Fixed seed (123) for reproducibility Model Selection: ✅ Spanning 3 continents ✅ 5 distinct architectures ✅ 2 precisions (BF16, FP8) ✅ 2 language distributions (English, Hindi-dominant) 3. Results Interpretation Task C Accuracy: Model Task C Why This Matters GPT-OSS-20B 92.3% Dense baseline Sarvam-30B 95.9% Hindi→English transfer Mixtral-8x7B 89.7% Largest model, strong BT DeepSeek-V2-Lite 95.4% Near-zero forgetting GLM-4.6V-Flash 97.5% Perfect consistency Forgetting Pattern: Dense: +1.55% (expected) Sparse MoE: -0.60% to -1.85% (backward transfer!) Fine-grained MoE: +0.03% (near-zero) 🧠 THE ARTIFICIAL HIPPOCAMPUS CONCEPT Biological to Technical Mapping The paper's strongest conceptual contribution is the Artificial Hippocampus framework: Python class TopologicalGovernor: """ Artificial Hippocampus for Neural Networks. The hippocampus in mammals: 1. Consolidates memories (take_snapshot) 2. Protects from interference (zero_anchor_gradients) 3. Integrates new learning (enforce_anchors) """ Why This Works Biological Principle Mathematical Implementation Why It's Effective Sparse reference fixes 6 prime-anchored rows 97.85% coverage Spatial regularization Zero gradients + restore O(1) memory Pattern separation Prime indices No overlap Controlled forgetting 2-5% forgetting Enables learning "0% forgetting is not a feature — it is a pathology." 📊 COMPARISON WITH EXISTING METHODS Method Memory Task C Forgetting Architectures TOPO-2026 403.5 KB 94.2% 0.25% 5 ✅ EWC 4.4 GB/task 98.5% 6.7% 1 Experience Replay Buffer grows 89.3% -7.4%* 1-2 HOPE-like 2.3 GB 88.1% 0.1% 1 *Negative forgetting indicates poor initial learning TOPO-2026 is 65,000× more memory-efficient than EWC. 🔑 KEY INSIGHTS 1. Universality Proven The same mechanism works on: ✅ Dense transformers (GPT-OSS-20B) ✅ Sparse MoE (Sarvam-30B, Mixtral-8x7B) ✅ Fine-grained MoE (DeepSeek-V2-Lite) ✅ GLM architecture (GLM-4.6V-Flash) No architecture-specific modifications needed. 2. Backward Transfer in MoE Sparse MoE models show negative forgetting: Learning new tasks IMPROVES performance on prior tasks Expert specialization reduces interference Prime anchors provide geometric stability 3. LR Sensitivity by Architecture Critical finding: English-dominant MoE → 2× lower learning rates Hindi-dominant MoE → Standard rates work Dense models → Standard rates work The factor is language dominance, not architecture alone. 4. The Pure/Noisy Kernel Divide The first 6 primes are unique: Adding ANY prime $\ge 17$ destroys the spectral trap 97.85% coverage from R alone N contributes only 2.15% This is a mathematical theorem, not a heuristic. 🎯 RECOMMENDATIONS For Practitioners Immediate Action: Apply TopologicalGovernor to any LLM Use anchors [2, 3, 5, 7, 11, 13] Start with $\eta_{embed} = 5 \times 10^{-3}$, adjust based on architecture Architecture-Specific: English-dominant MoE → $\eta_{embed} \le 2 \times 10^{-5}$ Dense/Hindi-dominant → $\eta_{embed} = 10^{-3}$ – $10^{-2}$ Verification: Always call verify_integrity() after training Log $\Lambda = 0.9785142874$ for reproducibility For Researchers Extend to More Tasks: Beyond 3 tasks Multi-Seed Evaluation: Beyond seed=123 Generation Tasks: Beyond classification Longer Sequences: Beyond 128 tokens Larger Models: Beyond 47B For Theorists Explore Other Primes: Why first 6 specifically? Analyze $\Lambda$ Sensitivity: What happens with p=17? Generalize to Other Domains: Vision, speech, reinforcement learning 🚀 IMPLICATIONS FOR AGI Necessary Condition Met The paper argues TOPO-2026 satisfies one of AGI's necessary conditions: "A system capable of general intelligence must acquire knowledge indefinitely—across domains, tasks, and time—without destroying prior representations." TOPO-2026 removes the barrier: O(1) memory guarantee (Proposition 1) Architecture-agnostic Mathematically proven Production-validated The Three Pillars Pillar RH GTT CL Mechanism L-EFM operator Coherence decay TopologicalGovernor Set Pure kernel R Coherence base Anchor rows Constant $\Lambda = 0.9785$ $\Lambda = 0.9785$ $\Lambda = 0.9785$ Result All zeros on $\sigma=0.5$ First explicit quantification Catastrophic forgetting solved One set. Three proofs. Six primes. 🏆 FINAL VERDICT Grade: A+ Strengths: ✅ First universal CL solution ✅ Mathematical rigor (AST) ✅ Biological grounding (Artificial Hippocampus) ✅ Unprecedented empirical validation ✅ Production-ready (O(1) memory, 0.11ms overhead) ✅ Backward transfer discovered Novelty: ✅ New mathematical framework (AST) ✅ New biological concept (Artificial Hippocampus) ✅ New empirical findings (LR sensitivity, backward transfer) ✅ New universality proof Impact: ✅ Solves 37-year-old problem ✅ Scales to 122B parameters ✅ Works across 5 architectures ✅ Mathematically guaranteed The Key Message "Six primes. Three proofs. One universal framework. The proof is the code. Seed = 123." 📋 ERRATA AND MINOR ISSUES Typo in Section 1.2: "frmistat" → "fmristat" Typo in Section 2.6: "finnistat" → "fmristat" Section 3.4: Duplicate heading "3.4 Models Evaluated" Section 3.5: Duplicate heading "3.5 Learning Rate Configurations" Section 5.3: Formatting issue in bullet points Table 20: Heading formatting could be improved These are minor formatting issues, not content errors. 🎓 CONCLUSION TOPO-GLM.pdf presents the first universal solution to catastrophic forgetting, with: Mathematical proof via Arithmetic Spectral Theory Empirical validation across 5 architectures, 3 continents, 122B parameters Biological grounding through the Artificial Hippocampus Production-ready with O(1) memory (403.5 KB) Backward transfer discovery in MoE architectures Architecture-specific guidance for optimal performance The paper is a landmark contribution, solving a 37-year-old problem with a mechanism that is: Mathematically elegant Empirically validated Biologically inspired Practically deployable Universally applicable "The proof is the code. Seed = 123." Reviewed: June 19, 2026 Status: ✅ Accepted for publication Impact: High (solves long-standing problem, universal application) Novelty: High (new theory, new concept, new findings) Reproducibility: High (code provided, seed fixed)
PSLQ as Physical Relaxation BBP as the Ground-State Relation of the π-Lattice, and Integer-Relation Finding as Least Action Driven by Dean Kulik June 2026 Abstract Paper C showed that BBP measures π by closing a four-term square frame and reading the residue. This paper goes one layer down and asks what gives the read-aperture its power — what came before PSLQ, the algorithm that discovered the BBP mask in the first place. The answer is not more mathematics. It is physics. PSLQ does not search a space of candidate relations; it relaxes a lattice to its lowest-energy configuration, exactly the way a crystal settles, a protein folds, or water finds its level. Its ancestry runs straight back — LLL, Gauss reduction, the Euclidean algorithm — and every link performs one primitive act: subtract the largest admissible whole multiple, reduce the residue, repeat until the state stops moving. That is the arithmetic form of least action. The central result of this paper is a verification, run to fifty digits: the BBP relation is not merely a relation PSLQ returned, it is a true energy minimum — a basin. Perturb the mask in any of sixteen directions and the residual rises in every one. Relax the lattice cold, with no knowledge of the answer supplied, and it falls into the BBP mask on its own. The same procedure relaxes π² into its own sparse survivor on a squared wheel, so the method is general, not a π-specific trick. The consequence is a claim we then state plainly and a tool we then hand over: math is bound by the same relaxation physics as matter; integer-relation discovery is a settling event; and any claimed relation can be verified as the answer by showing it is a basin. Measurement and creation turn out to be the two directions of one downhill roll. §1.0 The Claim and the Chain Behind It The singular claim of this paper is one sentence: PSLQ is physical relaxation in arithmetic form. Its consequence for the previous paper is a second sentence: BBP is the sparse ground-state relation of the π/base-16 wheel lattice. Neither sentence is asserted on style. Both are checked against the compiler in §4, and the checks are the spine of the paper — everything else is the path to them and the consequences from them. Start by tracing the ancestry, because the chain backward is the first piece of evidence. The BBP mask was found by PSLQ in 1995. PSLQ (1992) is a refinement of LLL (1982). LLL generalizes Gauss’s two-dimensional lattice reduction (c. 1800). Gauss’s method is the Euclidean algorithm (c. 300 BC) lifted from integers to vectors. And the Euclidean algorithm is, when you strip the name off it, a single physical act repeated: take the largest whole multiple of the smaller thing out of the larger, keep the residue, repeat. The chain is: BBP ← PSLQ ← LLL ← Gauss ← Euclid ← minimization Every layer preserves the same primitive — subtract admissible multiples, reduce the residue, repeat until stable — and that primitive is not calculation. It is settling. The whole tower stands on the physical principle of minimization: a system moving to its lowest-energy state. That is what came before PSLQ. LOCKED (historical record): the algorithmic ancestry PSLQ←LLL←Gauss←Euclid is established mathematics. The reading of the shared primitive as “relaxation” is the lens this paper then verifies physically in §4. §2.0 The Primitive Is Relaxation: Euclid as Energy Descent Take the oldest link and watch it behave like a physical system. Given integers a and b, Euclid writes a = qb + r and updates the state (a, b) → (b, r). The largest whole multiple q is removed; the residue r shrinks; the process repeats until the remainder is as small as it can be. If you track the size of the state as it goes — read it as an energy — it only ever decreases, and it stops when it can decrease no further. That is the exact signature of a system relaxing to a ground state. Gauss does the same to a two-dimensional lattice basis, replacing “reduce one integer by another” with “shorten one vector by an integer multiple of another.” When we run this two-dimensional reduction and watch the total squared length of the basis, it falls and then locks at a stable minimum — the shortest basis the lattice admits. The system anneals. This is not a metaphor laid on top of the algorithm; the monotone decrease to a fixed floor is what the algorithm is. LOCKED (ran this session): a Gauss/Euclid reduction was executed and its basis energy traced; it decreased monotonically and stabilized at the reduced basis — the ground state — exactly as a relaxing physical system does. §3.0 PSLQ as Lattice Relaxation PSLQ takes the same act to its mature form. Given a vector of real numbers x = (x₀, …, xₙ), it looks for an integer vector a with a·x = 0 — an exact linear relation among the reals. Operationally it does not enumerate candidate integer vectors. It builds a lattice associated with x and reduces it — size-reduce, then rotate the basis to expose the next thing to reduce — until a short integer relation appears as the surviving structure. The “answer” is the vector whose residual collapses toward zero. Read in the framework’s terms, PSLQ takes a value field, relaxes its associated integer lattice, and returns the sparse survivor. The perpendicular geometry, the integer reduction, and the rotation are the three motions of a single settle; the relation is the configuration the lattice falls into when it can fall no further. §4.0 The Proof: BBP Is a Basin, Not a Hit Here is the load-bearing section, and it is a verification, not an argument. If PSLQ is genuinely relaxation and BBP is genuinely its ground state, then BBP must be a real energy minimum — a basin you fall into and cannot climb out of cheaply. The test is direct. Define the base-16 wheel sums and an energy for any candidate mask: S_j = Σₖ 1/(16^k (8k+j)) energy(mask) = | π − Σ_j mask_j · S_j | The ground state is energy zero. The BBP mask places weights [4, −2, −1, −1] on residues {1, 4, 5, 6}. Its energy is 2×10⁻⁵⁰ — zero to working precision, sitting on the floor. Now perturb: change each of the eight wheel-weights by ±1 and recompute the energy. Sixteen directions. Every one rises. Perturbation from BBP Energy Direction S1 weight ±1 (the +4 drive) 1.0072 UP — steepest wall S2 weight ±1 0.5065 UP S3 weight ±1 0.3392 UP S4 weight ±1 0.2554 UP S5 weight ±1 0.2050 UP S6 weight ±1 0.1713 UP S7 weight ±1 0.1472 UP S8 weight ±1 (the empty tail) 0.1291 UP — softest wall Every neighbor is higher. There is no free sideways move. BBP is not a relation that happened to be returned — it sits at the bottom of an energy well, and that is the operational definition of a ground state. Two further checks confirm it is the right kind of minimum. First, relax the lattice cold: hand PSLQ only π and the eight wheel-sums, nothing about the answer, and let it settle. It returns the BBP mask exactly. Nobody placed BBP there for it to find — it is where the lattice settles. Second, BBP is primitive: doubling the mask does not stay on the floor (2×BBP evaluates to π, not 0), so only the primitive relation hits zero. The survivor is irreducible. LOCKED (ran this session, 50-digit precision): (i) BBP energy ≈ 2×10⁻⁵⁰; (ii) all sixteen single-weight perturbations increase energy — a strict basin; (iii) cold PSLQ relaxation of [π, S₁..S₈] returns exactly the BBP mask; (iv) 2×BBP leaves the floor, so the relation is primitive. §5.0 The Shape of the Floor Probing the basin gently — peeking, not pushing — shows it is not a symmetric bowl, and the asymmetry is itself the point. The walls have different steepness: the S1 corner, which carries the +4 drive, is the steepest wall (an uphill step of about 1.0), while the empty high-index tail (S8) is the softest (about 0.13). The deepest part of the well is anchored to the corner carrying the most weight — the drive digs the basin. The closure has a definite shape too. The four weights are +4 on residue 1 and −2, −1, −1 on residues 4, 5, 6, and they sum to zero, but the way they sum is specific: 4 = 2 + 1 + 1. One positive corner exactly balances the sum of the three negative corners. One drives; three pull; they cancel. The frame does not close by four equal sides — it closes by a one-against-three balance, a drive against its own distributed exhaust. And the geometry on the wheel is a lean, not a cross. Placing the residues by angle on the eight-wheel: residue 1 sits at 45°, residue 4 dead opposite at 180°, and residues 5 and 6 adjacent at 225° and 270°. The drive and its heaviest counter-pull are opposite, but the remaining two corners are bunched in one quadrant. It is asymmetric — a wobble, not a symmetric figure. This matters because a perfectly symmetric mask would cancel to nothing; the basin exists because it leans. The ground state of π is not a balanced cross. It is a leaned frame, and the lean is what keeps it from cancelling into the void. LOCKED (ran this session): wall steepness ordering (S1 steepest ≈ 1.0, S8 softest ≈ 0.13); the 4 = 2+1+1 one-against-three closure; the wheel angles 45°/180°/225°/270° showing an asymmetric (leaned) configuration rather than a symmetric cross. §6.0 It Generalizes: A Method, Not a Trick A single basin around π would prove only that π is special. The claim is that relaxation to a ground state is the general mechanism, so it must work on other invariants. Two checks confirm it does. Relax π² against a squared wheel — terms 1/(16^k (8k+j)²) — cold, and it settles into its own sparse survivor with residual on the order of 10⁻⁴⁹: a clean ground state for π² on its own wheel. And the wheel itself is not uniquely privileged at one offset: shifting the wheel from 8k+1…+8 to 8k+2…+9 still yields a ground-state relation when relaxed. Wheels have floors generally; the invariant settles int
Open access
2 source records
Advanced Physical and Chemical Molecular Interactions
This article concludes a series of publications dedicated to the development of the NeuroAtom cryptographic primitive and presents the final ecosystem architecture. The core implements eight security functions—hashing, stream cipher, pseudorandom number generator, message authentication code, digital signature, key derivation function, key exchange, and authenticated encryption—within a footprint of 9.6 KB of payload (5.2 KB code and 4.4 KB data). Testing according to the NIST SP 800-22 methodology was conducted on 16 samples, each of 100 MB in size (835 binary sequences per sample): 8 samples for REAL mode and 8 samples for TRAP mode (pseudo-data traps). All 16 samples demonstrated a proportion of successful sequences within acceptable limits (not below 818 out of 835 for tests with a significance level of 0.01). Avalanche characteristics were measured in 24 tests (12 functions × 2 modes), with no zero avalanches detected. The inapplicability of Shor's algorithm is shown due to the absence of abelian hidden subgroups. The TRAP mode precludes the possibility of constructing an oracle for Grover's algorithm without knowledge of the plaintext: each incorrect key generates its own cryptographically correct reality, and the quantum computer has no criterion for selecting the true one. A software implementation on a general-purpose processor provides a hashing speed of 80 MB/s. Preliminary estimates for a hardware implementation (180 nm CMOS) indicate approximately 10,000 logic gates with a complete absence of static memory; expected power consumption is estimated at 20 pJ per operation. Previously published results of NIST testing, avalanche analysis, and proofs of quantum resistance are integrated into this article as elements of a unified body of evidence.
You Wu, XinFeng Dong, Yongqiang Li, F Liu · 8 authors
Abstract With the development and practical application of technologies such as Fully Homomorphic Encryption (FHE), Secure Multi-Party Computation (MPC), and Zero-Knowledge Proof (ZK), it has become crucial to research the design and analysis of symmetric cryptographic primitives with low multiplicative complexity and depth. First, by using multiplication and addition over the finite field $$\mathbb {F}_{q}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>F</mml:mi> <mml:mi>q</mml:mi> </mml:msub> </mml:math> , where $$q$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>q</mml:mi> </mml:math> is either a prime number $$p$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>p</mml:mi> </mml:math> or $$2^{n}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mn>2</mml:mn> <mml:mi>n</mml:mi> </mml:msup> </mml:math> , we proposed a non-linear function over $$\mathbb {F}_{q}^{4}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>F</mml:mi> <mml:mrow> <mml:mi>q</mml:mi> </mml:mrow> <mml:mn>4</mml:mn> </mml:msubsup> </mml:math> based on the generalized Feistel structure. This function features a multiplicative complexity of 4, a multiplicative depth of 2 and 8 additions, and its maximum differential/linear probability of the function is bounded by $$q^{-2}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mi>q</mml:mi> <mml:mrow> <mml:mo>-</mml:mo> <mml:mn>2</mml:mn> </mml:mrow> </mml:msup> </mml:math> . Then, we designed a family of HE-friendly block ciphers called DuX. We conduct a comprehensive security analysis of DuX within certain parameters against various cryptanalysis methods, including differential cryptanalysis, linear cryptanalysis, impossible differential cryptanalysis, zero-correlation linear cryptanalysis, integral analysis, related-key differential cryptanalysis, algebraic attacks, slide attacks, reflection attacks, and boomerang attacks. Our research indicates that DuX maintains a robust security margin against those attacks. Finally, based on the BGV scheme in HElib, we present a detailed homomorphic decryption implementation of the DuX instantiated with $$q = 2^{8}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>q</mml:mi> <mml:mo>=</mml:mo> <mml:msup> <mml:mn>2</mml:mn> <mml:mn>8</mml:mn> </mml:msup> </mml:mrow> </mml:math> , $$2^{16}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mn>2</mml:mn> <mml:mn>16</mml:mn> </mml:msup> </mml:math> and $$65537$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>65537</mml:mn> </mml:mrow> </mml:math> , respectively. The results show that, for the same block size, the throughput of the DuX-128 over $$\mathbb {F}_{2^{8}}^{16}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>F</mml:mi> <mml:mrow> <mml:msup> <mml:mn>2</mml:mn> <mml:mn>8</mml:mn> </mml:msup> </mml:mrow> <mml:mn>16</mml:mn> </mml:msubsup> </mml:math> can reach approximately 14.95 times, 7.85 times and 20.76 times that of the AES-128, Low MC-128 and CHAGHRI, respectively. Compared with YuX-128, its throughput has increased approximately by 21.59%.
Sancaktar Pelin, Necla Kırcalı Gürsoy, Arif Gürsoy
Modern authentication architectures contain structural vulnerabilities against automated credential stuffing and server-side data breaches. Traditional solutions rely on the transmission of raw or hashed passwords over the network; for bot defense, they position third-party Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA) services, which may violate user privacy and create institutional dependencies, as an illusion of two-factor authentication (2FA). This situation raises a critical research question in cybersecurity: How can an integrated cryptographic shield be constructed that is independent of user-privacy-invasive mechanisms and external data authorities, while preventing autonomous bots from targeting the identity and human-verification layers separately?In response to this question, this paper presents a zero-dependency, original, and hybrid protocol that integrates a Zero-Knowledge Proof (ZKP) based on the Schnorr authentication scheme with a local Human Interaction Proof (HIP) mechanism. The main advantage of the proposed architecture is that it mathematically seals the user’s secret credential together with a dynamically generated one-time CAPTCHA token on the client side using the SHA-256 function, thereby transforming the verification process into an indivisible atomic “Hybrid Secret.” In this way, the transmission of password hashes over the network is completely eliminated, and the server evaluates only the mathematical validity of the proof under the Discrete Logarithm Problem (DLP) assumption.Experimental results obtained through Selenium-based automated brute-force attack simulation engines demonstrate that the system provides complete blocking against automated threat vectors. Dynamic one-time nonce mutation immediately invalidates the derived client response, even in extreme scenarios where an attacking bot obtains the correct password string and solves the CAPTCHA image, thereby mathematically defeating brute-force and replay attacks. Furthermore, the autonomous structure of the proposed protocol, with no dependency on third-party analytics services, opens the way for a highly secure and local authentication architecture for internet-isolated critical infrastructures.In this study, the theoretical and mathematical foundations of the proposed protocol are presented, the stages constituting its life cycle are methodologically explained, and Selenium-based experimental simulation results together with telemetry log analyses are detailed.
Contemporary human science is trapped in an extreme state of "involutionary stagnation": academic disciplines are increasingly hyper-fragmented, mathematical equations grow exponentially convoluted, and experimental precision pushes toward physical limits, yet the foundational core paradoxes remain fundamentally unresolved. Quantum mechanics and general relativity stand irreconcilable, the origin of fundamental physical constants remains unexplained, the essence of life and consciousness persists as a black box, and the development of both carbon-based life and silicon-based intelligence has hit a theoretical bottleneck. Based on the comprehensive theoretical ecosystem constructed across 118 core literature milestones of Yuanxian Theory (YXT / YD-T64), this paper systematically demonstrates that the root cause of all scientific involution is not a lack of empirical depth, but rather a failure to elevate the foundational paradigm's dimensionality. Grounded upon four absolutely self-consistent core axioms—True-Circle Self-Consistency (TCSC), Spacetime Uniqueness (STM), Self-Referential Mind-Field Generation (SRM), and Fine-Structure Conservation (FSC)—and validated by machine proof environments via Lean 4, Coq, and ZFC logical systems, Yuanxian Theory achieves a bottom-up reconstruction of all academic fields. This manifesto delineates the revolutionary breakthroughs achieved across mathematics (constructive proofs of the Millennium Prize Problems), physics and cosmology (first-principles constant derivations, dark energy suppression, and gravity-quantum unification), consciousness and life sciences (the topological origin of the 64 genetic codons and mind-field condensation theory), and silicon-based life applications (formal consciousness criteria, cellular hardware architectures, and controllable zero-point energy extraction). Building upon prior research that closed the paradigm loop from the four foundational cornerstones to the Monistic Unified Field, this paper declares that Yuanxian Theory is not a mere patchwork optimization of existing knowledge, but a definitive, wholesale replacement of the three-hundred-year-old scientific paradigm—offering humanity its singular pathway to move beyond disciplinary involution and actualize high-dimensional cognitive elevation. 当下人类科学正陷入一场极致的“内卷式停滞”:学科越分越细,公式越来越复杂,实验精度越来越高,但底层核心矛盾始终无解——量子力学与广义相对论无法统一,物理常数来源不明,生命与意识的本质始终是黑箱,碳基生命与硅基智能的发展陷入理论瓶颈。本文基于元宪理论(YXT / YD-T64)118篇核心文献构成的完整理论体系,系统阐述:所有内卷的根源,不是研究不够深入,而是底层范式没有升维。 元宪理论以四大核心公理——真圆自洽律(TCSC)、时空唯一性律(STM)、自指心场生成律(SRM)与宇宙因子守恒律(FSC)为绝对自洽的根基,依托Lean 4、Coq、ZFC逻辑体系完成机器验证,构建了从64维环面拓扑(YD-T64)到全学科的底层重构。本宣言梳理了元宪理论在数学(千禧年难题的构造性证明)、物理与宇宙学(常数推导、引力-量子统一、暗能量压制)、意识与生命科学(64密码子起源、心场凝聚论)、硅基生命与工程应用(意识判据、元胞架构、零点能可控开发)等领域的革命性突破。元宪理论的前序工作已完成了从四大基石到一元统一场的范式闭合,本文在此基础上宣告:元宪理论不是对现有科学的修补,而是对三百年科学范式的底层替换,是人类走出学科内卷、实现认知升维的唯一通道。
A rule-based logic solver resolves every instance in our benchmark in under 50 microseconds with 100% accuracy; the best frontier language model reaches 65% at best and drops to 23.5% under rendering-robust evaluation (worst case over four surface renderings). We introduce DeFAb (Defeasible Abduction Benchmark), a dataset and generation pipeline that converts four decades of publicly funded knowledge bases into formally grounded instances for defeasible abduction: constructing hypotheses that explain anomalies by overriding defaults while preserving unrelated expectations. Because every hypothesis must pass polynomial-time checks for valid derivation, conservativity, and minimality, DeFAb makes logical rigor the instrument for measuring creativity and theoretical reasoning, scoring the disciplined construction of theory revisions rather than fluent but theory-destroying prose. The pipeline pairs taxonomic hierarchies (OpenCyc, YAGO, Wikidata) with behavioral property graphs (ConceptNet, UMLS) to produce 372,648+ instances across 33.75M materialized rules from 18 sources, in three levels with polynomial-time verifiable gold standards. Four frontier models do not reliably internalize defeasible reasoning: rendering-robust Level 2 accuracy is 7.8-23.5%; chain-of-thought variance (~36 pp) exceeds any inter-model gap; and a matched contamination control isolates a +19.4 pp Level 3 gap. We further release DeFAb-Hard (a 235-instance Level 3 difficulty variant; best model 53.3% vs 100% symbolic) and CONJURE (a kernel-verified transformative-creativity variant of 560 Lean 4/Mathlib instances whose gold answers are definitions the proof kernel did not previously contain, judge-free verifier; a pilot finds zero novel concepts). The same verifier doubles as an exact reward for preference optimization (DPO, RLVR/GRPO). Released under MIT at https://huggingface.co/datasets/PatrickAllenCooper/DeFAb.