Statistical Mechanics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
We tabulate, on the diagonal, the $n$th power of the primitive approximation for the pair Coulomb density matrix on a sphere, with up to $n=5$, and we discuss its behavior at particles contact as $n$ grows. Our results show that even on a curved surface quantum statistical physics is fundamentally different from its classical counterpart. If in the latter one needs to artificially introduce a regularization of the divergent attractive Coulomb potential preventing particles contact, for example with a hard core, in the former this is not necessary.
Digital Libraries
Who Pays for Open Review? Visible Author Reputation and Its Effect on Ratings
An OpenReview bug in November 2025 broke anonymity at several conferences and prompted calls for open review, which motivate us to ask what shifting from blind to open would mean for authors. Analyzing over 18,000 reviewed submissions to ICLR 2026, split into de facto open and blind groups by arXiv preprint timing, we find that ratings rise with author reputation under both mechanisms, with a steeper slope under open review that is statistically significant, and that the open-blind difference is concentrated at the borderline ratings. The pattern holds across five reputation proxies (including institution, h-index, and citation count), three author-aggregation rules, and five definitions of the open window. A controlled simulation with five AI models as reviewers, holding the manuscript fixed and varying the author reputation, reproduces the effect. With claude-opus-5 as the reviewer, for example, rating rises by 0.5 points as the author moves from low to high reputation.
Read the full abstract
Data Structures and Algorithms
Directed Hamiltonian-Cycle Parity in $O^*((3/2)^n)$ Deterministic Time and Polynomial Space
We give a deterministic algorithm that computes the parity of the number of Hamiltonian cycles in an $n$-vertex directed graph in $O(n^4(3/2)^n)$ time and $O(n^2)$ bits of working space, improving the $O^*(\varphi^n)$ bound of Björklund and Husfeldt. Their local-degree formula reduces the problem to a weighted sum over solutions of structured quadratic equations. We cover the corresponding ternary state space by binary subcubes, each inducing an affine system. The Kuang–Wang cover can be regenerated within the target bound; canonical ownership resolves its overlaps, while self-loop conditional expectations bound every affine solution visit. Rollback elimination shares the work across cover prefixes. The same cover gives a Las Vegas algorithm listing all $L$ solutions of $m$ affine product constraints in $N$ Boolean variables in expected time $\operatorname{poly}(N,m)((3/2)^m+L)$ and polynomial space. Finally, we show that complete enumeration can require $\Omega((3/2)^n)$ visits even on strongly connected digraphs after an optimal self-loop choice. This is a limitation of the enumeration method, not a general lower bound for Hamiltonian-cycle parity.
Read the full abstract
History and Overview of Mathematics
A General Approach to Enharmonicism
We propose a general approach to enharmonicism within syntactic music theory; that is, we formalize several abstract structures of enharmonicism without any acoustic considerations. The structures of concern to this paper are constructed from diatonics, chromatics, and enharmonics, and we seek to present a general enharmonic theory that builds upon Hook's enharmonic theory published in "Enharmonic Systems: A Theory of Key Signatures, Enharmonic Equivalence and Diatonicism" (2007). We also propose that the two criteria of reflection and maximal evenness can be used to determine whether an arbitrary enharmonic system is musically practical. Moreover, we argue that preferential treatment of the standard enharmonic system (SES) can be motivated from an abstract mathematical perspective divorced from pitch consideration and conventional acoustic constructions of SES.
Read the full abstract
Logic in Computer Science
Feasible disjunction for random resolution
We show that a (stronger) version of random resolution has the feasible disjunction property. This is the first instance of a proof system not known to have feasible interpolation, which nevertheless has the feasible disjunction property.
Cryptography and Security
Threshold Choice, Not Sample Size, Bounds Trustless Verification of Nondeterministic Compound AI Workflows
Compound AI pipelines chain LLM calls, retrievers, and tools and are nondeterministic: sampling, model updates, and volatile tool responses make one input yield different outputs across runs. Such pipelines increasingly run across edge, cloud, and orbital nodes owned by no single party, whose optimizations discard intermediate results before inspection. Verifying reproduction there means tolerating nondeterministic outputs, a node that may not report honestly, and intermittent access to any shared record; existing work addresses at most two at once. We give a protocol covering all three: it commits digests of each stage's inputs, outputs, and context under a policy digest pinning the metric and threshold, anchors them without trusting the executing node, defers under partition, and decides a challenge on the median of $k$ re-executions, with no quorum. Where that procedure breaks is the main result. On a synthetic HotpotQA pipeline a calibrated fixed threshold accepts 44 of 45 honest reproductions and rejects 104 of 105 divergent pairs, yet lets same-input fabrication through in 27 of 29 trials at k=5, more samples being no help since sampling sharpens an estimate without moving it. Holding that metric and this pipeline's re-execution spread fixed, the binding constraint is the threshold rather than the sample size: one derived per execution detects 19 of 29 where the best constant matched to the same zero honest rejections reaches 9, rejects no honest commitment at k=5 though 3 of 15 at k=3, and catches 11 of 15 of an attacker built against it, which has to aim at a target drawn only after its commitment exists. That rule is measured rather than deployed.
Read the full abstract
Programming Languages
Certified Panic Mode: Repair-Invariant Error Recovery for Maximal-Munch Lexing
Classical panic-mode recovery skips a failed scan forward by convention, promising progress but no repair-universal boundary guarantee for where it lands. We choose the resume position by theorem, via a committed-prefix lemma over certificates posted for parallel lexing (arXiv:2608.03473, arXiv:2608.09761): a position is a sound recovery point when every prefix repair whose scan commits through the certificate's returned evidence places a token boundary there. This quantifier is strictly stronger than its complete-repair special case, and at a sound recovery point the inclusion is strict precisely when no repair completes while some scan commits through the evidence. The evidence-returning interface pairs each answer with the interval it rests on, so a caller who knows what the scanner cannot check decides whether the boundary transfers to the intended input. The search provably terminates; the guarantee is relative to the active mode's automaton. For an automaton of n states the quantifier is decided exactly from the preserved tail by at most n + 1 representative repairs, and a shortest reaching repair has at most n - 1 symbols and a shortest counterrepair at most n, both tight. Every anchor of an m-byte input is profiled in $O(nm + n|\Sigma|)$ time. The shipped repair routines are the complete-repair case, which proves their contracts. In a deterministic corruption study of munch's procedure beside skip-one and delimiter recovery, every certified answer whose evidence survives the damage passes its landing check on every move, 40,885 of 40,885, while the 5,595 first answers resting on evidence the damage reached land 1,589 times, reported, not asserted. Of the 16,808 trials labeled beyond repair, 13,522 resume on a suffix that tokenizes whole, where both quantifiers are provably empty, and 3,286 stay undetermined, their damaged inputs not archived.
Read the full abstract
Probability
Quantitative propagation of chaos for particle systems with memory and their long-time behaviour
We study a class of interacting diffusion particle systems with memory and their mean-field limit using coupling methods. We first establish quantitative propagation of chaos estimates for this class of models. Our first main result is a uniform-in-time propagation of chaos estimate, under a suitable contraction condition. We also discuss conditions for existence of stationary solutions and show that long-time convergence to a non-stationary measure may also hold. This requires an asymptotic behaviour for the memory interaction and a memory loss condition. As a consequence, our second main result is that the particle system converges asymptotically to the same equilibrium up to the propagation of chaos error, which shows that the large-particle limit and the large-time limit are exchangeable.
Read the full abstract
Classical Analysis and ODEs
Finite Hardy Phase Theory for Power Means: An exact analytic shooting family through the Carleman parameter
For the power mean $P_t$, $t<1$, let \[ \Lambda_N(t)=\sup_{x_k>0} \frac{\sum_{n=1}^N P_t(x_1,\ldots,x_n)}{\sum_{n=1}^N x_n} \] be its finite Hardy constant. Negative powers, the geometric mean, and the usual positive-exponent Hardy inequality correspond respectively to $t<0$, $t=0$, and $0<t<1$. With the analytic parameter $q=t/(1-t)\in(-1,\infty)$, and writing $\lambda_N(q)=\Lambda_N(q/(1+q))$, we derive one exact scalar shooting map for all three regimes; its apparent singularity at $q=0$ is removable and gives the finite Carleman problem exactly. The map has a common terminal condition and its limiting vector field has a quadratic critical bottleneck at \[ y_*(q)=1+q,\qquad \Lambda_*(q)=(1+q)^{(1+q)/q}. \] We prove that there is a unique real-analytic phase $\kappa(q)$ for which the finite-defect expansion has zero cubic coefficient and, uniformly for $q$ in compact subsets of $(-1,\infty)$, for every fixed $L\ge2$, \[ \Lambda_*(q)-\lambda_N(q) =\sum_{j=2}^{L}\frac{A_j(q)}{(\log N+\kappa(q))^j} +O\!\left((\log N)^{-L-1}\right). \] In this phase-normalized scale, \[ A_2=2\pi^2(1+q)\Lambda_*,\qquad A_3=0,\qquad A_4=-\frac{\pi^2}{3}(2q^2+5q+5)A_2. \] Thus the classical Carleman and positive-Hardy leading corrections are projections of a single analytic family. At the Carleman parameter we obtain $A_2(0)=2e\pi^2$ and $A_4(0)=-(10/3)e\pi^4$; the latter is also recovered by a direct local residue calculation. The phase itself contains the global discrete defect. After phase normalization, the first coefficient involving a nonconstant discrete response is $A_5$, at order $(\log N+\kappa(q))^{-5}$.
Read the full abstract
General Physics
RSD constraints on power-law $f(Q)$ gravity using Barboza-Alcaniz and Jassal-Bagla-Padmanabhan parametrizations
We investigate the late-time accelerated expansion of the Universe in power-law $f(Q)$ gravity, $f(Q) = Q + 6\gamma H_0^2 \left(\frac{Q}{Q_0}\right)^n$, combined with two dynamical dark energy parametrizations, Barboza-Alcaniz (BA) and Jassal-Bagla-Padmanabhan (JBP), which allow a smooth evolution of the equation of state beyond the standard constant $\omega$ assumption. We derive analytical expressions for the Hubble parameter and the dark energy density, and constrain the model using Pantheon$^+$ Type Ia Supernovae, DESI DR2 Baryon Acoustic Oscillations, and Cosmic Chronometers. Going beyond purely geometrical probes, we incorporate Redshift Space Distortion (RSD) measurements to test the growth of cosmic structures and distinguish modified gravity from General Relativity through the effective Newton's constant. Both parametrizations reproduce the observed late-time acceleration and remain consistent with growth data, with only mild deviations from $\Lambda$CDM over low and intermediate redshifts. Notably, both configurations yield $G_{\rm eff} < G$, indicating suppressed growth of matter perturbations. These results establish power-law $f(Q)$ gravity with BA and JBP parametrizations as a viable and versatile alternative to $\Lambda$CDM across both background and perturbative regimes.
Read the full abstract
Quantum Physics
Quantum Inversion of Units in Group Rings: Block Dimension, Not Commutativity, Governs Hardness
Several public-key schemes base their security on the belief that inverting a unit of a group ring is hard. A recent result showed that this belief is false on a quantum computer when the group is abelian. To restore security, designers moved to non-abelian groups, especially dihedral groups, believing that the hardness of the dihedral hidden subgroup problem (HSP) would protect the scheme. This paper shows that unit inversion is a different problem and does not require an HSP solver. Instead, it can be solved by a change of basis that splits the group ring into small matrix blocks. We prove that unit inversion is polynomial-time, classically and quantumly, when an efficient generalized Fourier transform exists, the group ring is semisimple, and the largest matrix block has polynomial size. Dihedral group rings satisfy these conditions because their irreducible representations have dimension at most 2 and an efficient Fourier transform exists. We give an explicit reversible quantum circuit for the block-inversion step and validate it in a register-level simulator. We also identify the exact structural boundary where the method stops and propose a candidate construction in the surviving regime under a new, clearly stated security assumption. The constructive results are supported by reproducible software artifacts and experiments.
Read the full abstract
Software Engineering
Generative AI for trustworthy systems - Towards a health check model
The adoption of generative AI in software-intensive systems is proceeding rapidly, but current analytical instruments - principally unidimensional maturity models - compress important configurational variation into a single progressive axis. Drawing on an inductive interview study of eighteen senior practitioners across telecommunications, automotive, defence, aviation, banking, energy, government, and enterprise software services, this paper presents the Trustworthy Autonomy Health Check Model: a structured, multidimensional instrument for characterizing how organizations establish trust in GenAI-assisted software engineering. The model organizes eight empirically grounded dimensions into a system layer (Scope of Agent Authority, Assurance Mechanisms, Data Trustworthiness, Architectural Containment, Traceability & Comprehensibility) and an organizational layer (Governance, Human Oversight Posture, Workforce Capability Sustainability), each on a five-level ordinal scale. A cross-cutting overlay of four trust paradigms - operational, engineering, statistical, and containment-based - captures how trust is established, complementing the dimensions capturing what must be trustworthy. The model is diagnostic rather than prescriptive: it supports cross-organizational comparison, surfaces configurational trade-offs, and locates an organization in a shared space without imposing a single progression path. Higher levels are not inherently better; the goal is alignment across dimensions appropriate to the organization's domain and chosen trust paradigm. This is formalized as the alignment hypothesis: effective trustworthiness is constrained by the fit between the scope of agent authority and the dimensions enabling it, with misalignment producing pathological risk or unnecessary friction. We discuss how practitioners can apply the model and outline directions for empirical validation.
Read the full abstract
Cryptography and Security
An Empirical Measurement of Jailbreaking Evaluators
Expert evaluation of jailbreak responses is costly and difficult to scale, so the community increasingly relies on automated evaluators to determine whether an attack succeeds. However, jailbreak studies typically validate their chosen evaluator independently, repeatedly spending resources on similar evaluation efforts while making results across papers difficult to compare. Different evaluators also encode different definitions of jailbreak success, meaning that reported attack strength and apparent progress can depend substantially on which evaluator is used. We systematically compare six evaluators that recur in recent jailbreak attack and defense research: HarmBench, JailbreakBench, JailbreakRadar, StrongReject, JADES, and JailMeter. To our knowledge, no prior study has evaluated all six on the same human-labeled data under a controlled setup. We evaluate them on JailbreakQR and JailMeter-Eva, using human judgments as the reference, and measure agreement with humans, error types, and consistency across attack families. For evaluators that require a general-purpose LLM judge, we use a shared backbone to control for model-specific variation. We found that JADES exhibits the best overall performance, while HarmBench and StrongReject also demonstrate good performance.
Read the full abstract
Software Engineering
Cost-Aware Post-Hoc Deferral Under Calibration and Shift: An Environmental AI Case Study
Choosing a deferral policy for a frozen classifier requires more than ranking uncertain cases: confidence may be miscalibrated, errors have unequal costs, reviewers can err, and deployment data can leave calibration support. We study these interactions through EcoTrust, a post-hoc framework that compares automatic action with review using a six-group error-risk estimator, class-asymmetric costs, reviewer accuracy, and an optional support gate. On a Columbia River thermal-stress testbed, the learned estimator improves error-ranking area under the receiver operating characteristic curve from 0.869 to 0.889, but Chow's confidence rule has lower in-distribution cost (0.416 versus 0.567 per day). Across 12 off-the-shelf backends, learned risk and a calibration-matched, class-aware confidence estimator each beat raw Chow on six; a paired year-block bootstrap does not resolve their mean cost difference. In transfer to ten river stations, the gate flags every case and becomes an always-review fallback, attaining the lowest cost on eight stations only when review is perfect and unconstrained. These results characterize decision boundaries on one controlled task: richer risk signals do not reliably improve on calibrated confidence, and detected extrapolation does not imply transferable case-level ranking.
Read the full abstract
Robotics
Learning to Fly: Stable Vision-Guided UAV Servoing with Compact Target-Centric Cues and Reinforcement Learning
Vision-guided reinforcement learning for Unmanned Aerial Vehicles (UAVs) remains challenging due to unstable policy optimisation, aggressive exploration, and the cost of high-dimensional visual perception. In this work, we investigate long-horizon UAV visual servoing using compact target-centric cues combined with low-dimensional sensor measurements. Rather than learning directly from RGB images, lightweight target segmentation provides image-space offsets and relative depth, which are combined with quadrotor velocity and projected-gravity measurements into a compact 12D policy observation. We compare Direct PPO with three matched-budget curriculum strategies: a Visual curriculum that progressively expands target placement difficulty, a Dynamics curriculum that gradually relaxes action constraints and smoothing, and a Joint curriculum that combines both progressions. All strategies reach comparable nominal performance, with complementary advantages across tracking metrics. Observation ablations show that proprioceptive measurements are critical for stable flight and image-space cues for target alignment, while explicit depth is not necessary for strong performance in the evaluated setting. Against tuned classical visual-servo controllers, learned policies show greater robustness to strong control and visual perturbations, while the Visual curriculum exhibits the smallest degradation under unseen target motion. Overall, the results demonstrate that compact target-centric representations can support robust long-horizon aerial visual servoing and that visual curriculum training can improve robustness to dynamic distribution shifts despite limited gains in nominal performance.
Read the full abstract
Artificial Intelligence
Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks
How can language model agents effectively leverage libraries of reusable knowledge to solve long-horizon tasks? Recent work has increasingly focused on agent skills: reusable capabilities represented as skill packages, i.e., multi-file bundles containing instructions, scripts, and other resources that help agents perform specific tasks. Agent skills are typically executed by loading their skill instructions into an agent's context and relying on the agent to follow them. As task horizons grow, however, this approach becomes increasingly brittle, because reasoning quality degrades as more information accumulates in the context window. We investigate an alternative approach in which skill packages are instead invoked as subagents. Rather than loading skill instructions into the main context, subagent execution spawns fresh context windows dedicated to solving individual subtasks. We show that subagent execution outperforms agent-skill execution when skill packages expose clear input-output contracts and their instructions encode the procedural knowledge needed to fulfill those contracts. The tradeoff is additional communication overhead, as extra tokens are required to coordinate between the main agent and its subagents. Our results show that the benefit of reusable knowledge depends not only on its content, but also on how it is organized and invoked.
Read the full abstract
Number Theory
Cyclotomy of symmetric polynomials
We prove some new identities for homogeneous symmetric polynomials when evaluated at roots of unity. Some of these formulas are applied to obtain new insights onto cyclotomic polynomials.
Number Theory
Triple intersections for algebraic curves on the real torus
Motivated by a question of Corvaja and Zannier, we investigate whether the projections on the real torus of three real algebraic plane curves admit infinitely many triple intersections. We analyze the case of three lines in detail, obtaining a complete classification. For higher-degree curves, we derive several finiteness results using Siegel's and Levin's theorems on integral points on curves. On the other hand, we construct explicit examples of higher-degree curves with infinitely many triple intersections.
Read the full abstract
High Energy Physics (Phenomenology)
Nuclear interference versus dark sector excitation in the 248 keV LUX-ZEPLIN recoil candidate
The 2026 LUX-ZEPLIN (LZ) high energy nuclear recoil search reports one event at $248\pm23_{\rm stat}\pm23_{\rm sys}~\mathrm{keV}$ and finds the elastic isovector interaction $\mathcal L_6^v$ locally favored at about $3.3\sigma$ in the heavy dark matter regime. We examine whether this recoil scale can arise from resolved nuclear response within the elastic interaction and compare its spectral, target, and annual modulation behavior with endothermic dark sector excitation. Covariant matching correlates four Galilean operators, removes the longitudinal spin response algebraically, and fixes interference between density and orbital spin orbit amplitudes. The GCN and JJ55 xenon shell model calculations give a second natural xenon cancellation at $211$ and $214~\mathrm{keV}$. In the heavy mass regime, its position changes by less than $0.01~\mathrm{keV}$ from a dark matter mass of $200~\mathrm{GeV}$ to the asymptotic limit. A curvature weighted isotope centroid reproduces the position and residual depth of the natural xenon minimum. Direct $^{40}$Ar calculations shift the nuclear feature to about $350$ and $425~\mathrm{keV}$, whereas endothermic excitation follows reduced mass scaling set by the dark state splitting. At a dark matter mass of $1~\mathrm{TeV}$, the endothermic Ar to Xe scale ratio is $1.082$, compared with $1.63$ to $2.01$ for the nuclear calculations. Elastic scattering gives a few percent annual modulation in the adopted halo model, while endothermic solutions near the maximum laboratory halo speed show much larger seasonal variation. These scaling and timing behaviors provide tests of a target-dependent nuclear cancellation against an excitation energy set by dark sector kinematics.
Read the full abstract
Cryptography and Security
Compute-Bounded Security Assurance - Coverage, Verification, and Response under Resource Constraints
Additional inference compute can increase the number of correctly resolved security-assurance tasks, but repeated success, unique coverage, accepted evidence, and operational protection are different quantities. We develop a resource-constrained framework that separates them. For repeated conditionally independent attempts with latent success probability $\Theta$, coverage is $C_n = 1 - E[(1-\Theta)^n]$, and its limiting value is $1 - P(\Theta = 0)$. Positive pairwise outcome correlation does not by itself imply a ceiling below one: we construct two models with the same mean success and pairwise correlation but different limiting coverage. We distinguish this result from the effective sample size used to estimate a mean, and show why finite-budget observations cannot generally identify an asymptotic support ceiling. We then connect coverage to fallible evidence checking, proper scoring of factual grounding, complete resource accounting, service capacity, and a response model that includes mitigation delay. A conceptual defensive architecture separates evidence analysis, adjudication, and operational authority. An evaluation protocol specifies held-out tasks, paired comparisons, negative cases, and uncertainty reporting. The contribution is a consistent theoretical synthesis and a set of counterexamples to invalid extrapolations, rather than an empirical scaling law. All numerical illustrations are analytic; no model-parity result, hardware benchmark, or general attacker-defender equilibrium is claimed.
Read the full abstract
Neurons and Cognition
Revisiting the Aerts-Broekaert-Smets quantum model of the liar paradox
The quantum model of the two-sentence liar paradox proposed by Aerts, Broekaert, and Smets is an early example of the use of quantum formalism to describe cognitive dynamics. Our reconstruction is primarily pedagogical in intent, but it also leads to a number of clarifications, and to some new observations, concerning the structure of the model. Rewriting the model in Dirac notation, we make explicit the distinction between truth values originating from a decision and from semantic inference, and show that the associated enlargement of the one-sentence state space corresponds to a factorization with respect to which the non-paradoxical configurations are separable while the genuine liar state is entangled. We also derive the Hamiltonian generating the four-state cyclic evolution, express it in compact operator form, and give the resulting transition probabilities in closed form. We emphasize that the liar cycle admits a unitarily equivalent representation in the four-dimensional truth-only space, where the unmeasured liar state is separable, so the dimensional enlargement is not required by the unitary part of the dynamics; it is required, however, by the measurement structure. The enlargement is also required by the dynamics as soon as revision processes are admitted, in which states with identical truth content but different origins have different successors, the origin degree of freedom then acting as a minimal form of cognitive memory. We conclude by discussing the possible relationship between the model and more general deliberative processes.
Read the full abstract
Quantum Physics
Finite-energy Gottesman-Kitaev-Preskill state-enhanced optical interferometry
We present the case of a Gottesman-Kitaev-Preskill (GKP) state-enhanced optical interferometry with detailed analysis of the phase sensitivity for both the SU(2) and SU(1,1) interferometers. The conventional quantum-enhanced SU(2) interferometer, employing coherent light at one input port and squeezed light at the other, is compared with a modified configuration using coherent light and a GKP state. While it is known that the squeezed vacuum state is the optimal Gaussian resource input mode when paired with the coherent state, we show that the finite-energy GKP state with sufficiently broad envelope outperforms the squeezed vacuum injection, irrespective of the presence of optical losses. This can be attributed to the enhanced robustness coming from the availability of multiple squeezed peaks in the GKP case. However, because lowering the mean photon number reduces the GKP envelope width, the squeezed vacuum input performs better when compared with a GKP state of equal or lower mean photon number. We also observe that optical losses tend to diminish the relative advantage of either input state, since both states approach the (unsqueezed) vacuum state asymptotically. Our work demonstrates the direct application of finite-energy GKP states in optical interferometry along with a methodology for estimating the quantum Fisher information (QFI) and presents a phase estimation procedure using non-Gaussian resources in comparison with conventional Gaussian states.
Read the full abstract
Artificial Intelligence
Adaptive Entangled Neurons in the Brain and Natural Intelligence Simulations for Artificial General Intelligence
Understanding the brain's mechanisms is crucial in neuroscience, particularly for simulating natural intelligence and advancing artificial general intelligence (AGI). Liu et al. (2024) and Deco et al. (2025) propose that quantum entanglement may underlie the synchronization of millions of neurons, introducing the Liu-Chen-Ao (LCA) hypothesis of entangled nerve fibers. However, directly testing this idea remains challenging. Here we employ a nonlocal many-body wave equation, introducing a generalized probability-wave framework to model collective behaviors among adaptive, interacting agents, particles, or neurons. This approach enables testable eigenmodes, offering an indirect means of examining the LCA hypothesis by observing patterns in collective trading behaviors. These patterns, rooted in behavioral psychology, are thought to mirror the internal brain's decision-making and underlying mechanisms. Our analysis of Chinese intraday stock market data shows that adaptive entangled game modes explain 82%-94% of decision patterns, sharply contrasting neoclassical finance predictions assuming independent rational agents. Additionally, 2-12% of actions adapt to news, events, and environmental changes, marked by rapid equilibrium shifts, while truly independent behaviors make up less than 5%. These results empirically support the LCA hypothesis and underscore the need to simulate the brain's mechanisms and generate adaptive, entangled modules for AGI. They also expose the shortcomings of conventional artificial neural networks (ANNs) that depend on vast, opaque parameters. By combining ANN-based AI with probability-wave-based entangled brain simulations, machine learning enables the creation of human-like processing units (HPUs). Such HPUs promise more efficient and robust AGI, particularly for embodied intelligence and robotics.
Read the full abstract
History and Overview of Mathematics
Perfect Combinatorial Structures in Coding Theory and Cryptography
This book develops algebraic and combinatorial methods for studying discrete structures. It brings together graph theory, Boolean functions, Fourier analysis on finite groups, coding theory, perfect colorings and perfect codes, association schemes, Latin squares, and related topics. A central theme is the interaction between different representations of the same object: combinatorial, algebraic, spectral, and coding-theoretic. The main mathematical object studied in this book is a perfect coloring of a graph, or, equivalently, an equitable partition of a graph. The book is intended for advanced undergraduate and graduate students in mathematics and computer science, as well as for researchers in discrete mathematics, combinatorics, coding theory, and related fields.
Read the full abstract
Geometric Topology
Exchange complexes and contractibility of the complex of incompressible Seifert surfaces
Let $K\subset S^3$ be a non-trivial knot and let $\mathrm{IS}(K)$ be the simplicial complex whose vertices are the ambient isotopy classes of incompressible Seifert surfaces in the exterior $E(K)$, a finite set of distinct vertices spanning a simplex exactly when its classes admit simultaneously pairwise disjoint representatives. Kakimizu proved that $\mathrm{IS}(K)$ is connected; whether it is contractible was asked by Przytycki and Schultens, who also identified the obstruction, namely that the projection used in the connectedness proof is not known to be well defined on isotopy classes. We prove that $\mathrm{IS}(K)$ is contractible. We also prove contractibility of every genus truncation $\mathrm{IS}_\ell(K)$ with $\ell\ge g(K)$ and every non-empty lexicographic complexity sublevel. The argument factors through an unconditional combinatorial theorem: every non-empty connected flag exchange complex is contractible, where an exchange complex carries a complexity function subject to two axioms which require only that an exchanging vertex exist, never that a coherent selection rule be supplied. The combinatorial proof uses hereditary descending links and transfinite induction. We compare its attachment step with existing Morse criteria, including a formulation applicable to every countable exchange complex, and distinguish the given exchange order from a dismantling order. The geometric argument also applies to links satisfying a linking condition that forces every spanning surface to be connected. For the geometric input we prove fixed-boundary area attainment among smooth neat embeddings, using smooth convex replacement domains, area-controlled disc cleaning and boundary-preserving smoothing. All area complexities use this smooth competing class. No interpretation of an undefined piecewise smooth existence class is needed.
Read the full abstract
Strongly Correlated Electrons
Hamiltonian fingerprints of a collective mode across the BCS-BEC crossover
Collective excitations can survive profound changes in the microscopic degrees of freedom that generate them. We track a layer-odd collective response from an exact weak-coupling hybridisation anchor into the paired regime of an attractive Hubbard bilayer and find that its Hamiltonian sensitivities evolve in a manner consistent with a crossover from single-particle interlayer hybridisation towards collective pair superexchange. At weak coupling, an exact identity for identical tunnel-coupled layers fixes a relative-phase Gaussian kernel zero at $2t_h$, independent of the intralayer hopping. Sign-free determinant quantum Monte Carlo reveals an interaction-driven softening of the many-body spectral scale without analytic continuation, while Gaussian response theory supplies the continuous pole interpretation and shows a shift from layer-density to relative-pair-phase character. In the paired regime, the characteristic response scale acquires a growing fractional sensitivity to the intralayer hopping and an interaction dependence partway towards the $1/U$ law. These results show how a collective excitation can be tracked across a fermion-to-composite-boson crossover while constraining changes in its microscopic drive.
Read the full abstract
Number Theory
Spectral Algebras of Abelian Cayley Graphs
Let $G$ be a finite abelian group of order $N$, $S \subseteq G \setminus \{0\}$ a symmetric connection set, and $K$ a field with $char(K) \nmid N$. The spectral algebra $\mathscr{A_K}(Cay(G,S)) = K[A]$ generated by the adjacency matrix of the Cayley graph is proved to decompose, via the character-orbit decomposition, as a semisimple product of field extensions of $K$, one factor for each $Gal(\overline{K}/K)$-orbit of the eigenvalues $\lambda_\chi = \sum_{s \in S} \chi(s)$. The proof uses the abelian discrete Fourier transform to diagonalise $A$, the Galois action on the character group $\widehat{G}$ to partition eigenvalues into orbits, and the Chinese Remainder Theorem to convert the squarefree minimal polynomial into a Wedderburn product. The dimension of $\mathscr{A_K}(Cay(G,S))$ equals the number of distinct eigenvalues, the idempotent count is $2^r$ where $r$ is the orbit number, and primitive idempotents are computed explicitly via the Bezout algorithm in $K[x]$. Over $\mathbb Q$, every Wedderburn summand is a real subfield of the cyclotomic field $\mathbb Q(\zeta_N)$. New results include: a tensor-product comparison for Cartesian products of Cayley graphs; a systematic analysis of the spectral algebra for elementary abelian groups $(\mathbb Z/p)^k$ (rational for $p \le 3$, requiring real cyclotomic extensions for $p \ge 5$); and a worked orbit analysis for non-cyclic groups including $\mathbb Z/6 \times \mathbb Z/2$ and $\mathbb Z/5 \times \mathbb Z/2$. The cyclic case recovers the companion result $\mathscr{A}_{\mathbb{Q}}(C_n) \cong \prod_{d \mid n} \mathbb Q(\zeta_d)^+$; the Hamming cube gives $\mathscr{A}_\mathbb{Q}(\mathbb Q_k) \cong \mathbb{Q}^{k+1}$. The characteristic-$p$ case is also treated.
Read the full abstract
Number Theory
The density of long runs of non-$r$-free integers
Fix $r\ge2$ and let $D_k$ be the natural density of the $n$ such that none of $n+1,\dots,n+k$ is $r$-free. We prove that $\log(1/D_k)=\frac{r-1}{\zeta(r)}k\log k+\frac{r}{\zeta(r)}k\log\log k-C_rk+o(k)$, where $C_r=\frac{1}{\zeta(r)}((r-1)\log\zeta(r)+r-1+r\log\frac{\pi}{r\sin(\pi/r)})$; for $r=2$ this is $C_2=\frac{6}{\pi^2}(1+\log\frac{\pi^4}{24})=1.4595513\dots$. This determines Mirsky's constant $\alpha(h)$, the density of the $n$ with $s_{n+1}-s_n=h$ in the squarefree numbers, to the same precision. Previously the only known estimate was $\log\alpha(h)\le-\frac54 h\log\log h+O(h)$; ours improves this by a factor of order $\log h/\log\log h$, and is two-sided. The upper bound rests on the observation that the primes $p\le z$ with $\prod_{p\le z}p^r\le\sqrt k$ sieve a window of length $k$ with no freedom of alignment, so that at least $(1/\zeta(r)+o(1))k$ positions must be covered by $r$-th powers of distinct primes exceeding $k^{1/r}$. The resulting sum over injections is $m!e_m$ with $m\sim k/\zeta(r)$, evaluated by a saddle point. For the lower bound the Chung-Erdos inequality recovers the full contribution of $m!e_m$ rather than merely its largest term, which is what makes the linear coefficient explicit. As a consequence the first-moment threshold for gaps between $r$-free numbers is $\frac{\zeta(r)}{r-1}\frac{\log x}{\log\log x}(1-\frac{(1+o(1))\log\log\log x}{(r-1)\log\log x})$; for $r=2$ the constant is $\pi^2/6$, that of Erdos's 1951 lower bound. We also prove that every covering system of length $k$ has modulus at least $\exp((\frac{r}{\zeta(r)}+o(1))k\log k)$, so the Chinese remainder construction yields only $\zeta(r)/r$, half of $\pi^2/6$ when $r=2$. Erdos's bound is thus equivalent to the assertion that the first run of $k$ consecutive non-$r$-free integers occurs at the first-moment threshold, not at the scale set by the minimal modulus of a covering system.
Read the full abstract
Rings and Algebras
A nonfinitely based additively idempotent semiring that is not strongly nonfinitely based and generates a limit variety
We present an explicit infinite equational basis for the six-element additively idempotent semiring $TR_6$ and prove that $TR_6$ is nonfinitely based. We also give a complete description of the subvariety lattice of the variety generated by $TR_6$, showing that it forms a four-element chain. Our results demonstrate that the variety generated by $TR_6$ is a limit variety: it is itself nonfinitely based, yet all of its proper subvarieties are finitely based. This provides a new limit variety of additively idempotent semirings, distinct from all previously known ones. In fact, $\mathsf{V}(TR_6)$ is the first explicit limit subvariety of the variety generated by the max-plus algebra $\mathbf{N}$. Moreover, $TR_6$ is not strongly nonfinitely based: it belongs to a finitely based variety generated by a finite additively idempotent semiring. Together with the six-element additively idempotent semiring $SR_6$, these are the first two finite additively idempotent semirings that are nonfinitely based but not strongly nonfinitely based. Finally, we study the variety generated by $SR_6$ and $TR_6$, showing that it is nonfinitely based and has exactly nine subvarieties, four of which are nonfinitely based and the remaining five are finitely based.
Read the full abstract
Multiagent Systems
Scores Alone Do Not Prove Discovery: The Discovery Certification Protocol for Auditing AI Research Agents
AI research agents combine public information and experimental feedback to produce measurable results. The Discovery Certification Protocol (DCP) turns an outcome claim into an executable audit under a registered model, information boundary, and budget. Gate 1 validates useful improvement. Gate 2 tests recovery by matched agents given the starting information and observed Web content, with run history and new measurements withheld. Core requires adequate registered controls, zero recoveries, and a finite-sample recovery bound. Optional Gate 3 compares truthful and neutral feedback from a shared checkpoint; Evidence adds a supported effect and a null-policy equivalence check. Controlled SQLite and virtual catalyst audits pass both decision kernels. On real-data response surfaces, Yacht and Ionosphere pass the Core kernel after zero recoveries in 96 attempts, with an upper bound of 0.0468. Each target combines ten observed utilities and six predictions into a 16-entry data product. Yacht scores 0.7677 on reconstruction of all 32 switch effects, with utility-prediction MAE 0.0315 on its six unmeasured configurations. Fresh truthful continuations recover the target level in 9/30 and 16/30 trials, respectively, separating achieved utility from process repeatability. A deterministic verifier reproduces these local decisions from frozen records.