Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Sunday 6 September

Computation and Language

A Ticket from Marginals to Joints: Coupled-Noise Distillation for One-Step Block Generation in Diffusion Language Models

Can a diffusion language model generate a coherent token block in one forward pass? Masked models already predict every position at once, but each prediction is the marginal distribution given the visible context, so the tokens can be mutually inconsistent and later steps revise those already committed. We introduce CONDOR (Coupled-Noise Distillation for One-Step Readout), trained from scratch to map different noise samples to different coherent blocks. Initially, random noise is not naturally paired with a target. Winner-take-all supervision lets different samples specialize, and self-distillation trains the one-pass output to match the refined coherent block. TinyStories experiments show diverse, coherent continuations over successive blocks, one forward pass each. Qualitative MNIST experiments show that the same approach can extend to multimodal generation, such as text-to-image and unconditional text-and-image generation.
Read the full abstract

Distributed Computing

Water-network decisions share one hydraulic gradient, and it can now be computed exactly

Calibration, leak localisation and sensor placement on water distribution networks (WDNs) are decisions about continuous parameters, yet the hydraulic engine that defines the physics returns a solution and no derivatives, so practice falls back on derivative-free search or on surrogates whose error the answer inherits. We make the global gradient algorithm itself exactly differentiable: the forward pass reproduces the reference engine's discrete devices, status switching and low-flow linearisation included, and the backward pass solves the implicit adjoint by reusing the forward pass's terminal factorisation, so one extra sparse solve returns every parameter's gradient at once, batched over scenarios on one graphics processor. Across 52 public, synthetic and operational networks and 8,140 simulation frames, every network meets the acceptance criterion, the largest head deviation from EPANET 2.2 is 1.137e-13 ft and 25 agree exactly. One adjoint solve replaces the 906 simulations a finite-difference roughness Jacobian costs on the 905-pipe L-TOWN benchmark, and a leak-inversion training loop runs at 463-470 ms per optimiser step for 256 scenarios, 191 times the prior pipeline. Gradient calibration reaches its endpoint within a median 595 model calls, where the strongest of five tuned metaheuristics needs 8,060 to match it on the training loss and two never do within 20,000. On a 554-link operating network, one adjoint pass audits, pipe by pipe, which roughness parameters the installed sensors can constrain and which sensors to add, on the model the utility already operates.
Read the full abstract

All authors Tianwei Mu, Yue Wang, Mingzhe Yuan, Wenhong Wang, Qing Luo, Min Xiao, Jun Li, Hui Yang, Manhong Huang

Cryptography and Security

Inevitability of Encrypted Traffic Side-Channel Leakage in the Multi-Class Setting

The Side-Channel Existence Theorem proves $I(X;Y)>0$ in the binary, undefended setting, but is confined to pairwise arguments and ignores active defenses. We extend it to $k$ classes via the per-class decomposition $I(X;Y)=\sum_i\pi_i D_{\mathrm{KL}}(P_{Y|i}\|P_Y)$, with defense cost modelled by per-class Wasserstein-1 constraints $\sup_x W_1(Q_x^D,P_x)\le B$. Three results follow: (1) a summation-form MI lower bound over all active classes; (2) a cascade critical cost theorem and a per-class budget corollary, nonzero where the uniform-budget bound vanishes; (3) an accuracy corollary $\mathrm{Acc}^*\ge 2^{I_0}/k>1/k$. On a 95-class website fingerprinting dataset the measured MI has a strictly positive $95\%$ confidence lower bound under every defense tested. Against the strongest pairwise baseline—a convex program over all $\binom{k}{2}$ triangle constraints, also $\Theta(1)$ in $k$ under the same non-vanishing-gap conditions—the summation form is only $1.45\times$ stronger, so the case for the per-class decomposition is structural: only it gives each class a critical cost and a cascade. FRONT's apparent $122\times$ gap is inflated mainly by threshold exclusion rather than the inequality chain: on the active classes it is $21\times$, within $1.4\times$ of the $15\times$ measured undefended. Measuring the chain's two steps separately bounds the collapse onto one Lipschitz statistic below by $28\times$, against a divergence step measured at $1.5\times$. Undefended OVR distinguishability predicts post-defense per-class leakage at Spearman $\rho=0.62$–$0.77$, the transfer the certification procedure relies on. The framework carries over unchanged to a 100-class QUIC/TCP pair.
Read the full abstract

Plasma Physics

Experimental realization of a dusty plasma rocking ratchet with current reversal

A single dust particle confined in an asymmetric ratchet potential is periodically driven by two oppositely directed laser beams, forming an underdamped dusty plasma rocking ratchet. We experimentally investigate the transport dynamics of the particle under varying driving amplitudes and frequencies. Depending on the driving conditions, the particle exhibits positive, zero, or negative net currents, and current reversal is observed when the driving parameters cross critical thresholds. To interpret these transport behaviors, we develop a simplified model based on the competition between the driving force and the ratchet confinement. The model reveals that directional transport is governed by two requirements: the driving force must exceed the depinning threshold, and the duration of a driving semicycle must be longer than the uphill escape time from a ratchet well. The resulting dynamic phase diagram quantitatively reproduces the experimentally observed transport regimes and current reversals. These results demonstrate dusty plasma as a versatile platform for investigating nonequilibrium transport phenomena of underdamped particles in rocking ratchets.
Read the full abstract

All authors Shun-xin Zhang, Shuo Wang, Ting-yu Yao, Yong-liang Zhang, Bao-quan Ai, Ya-feng He

Human-Computer Interaction

FrankenReport: Early Exiting in Long-Form Generation Using Expected Value of Computation

While deep research systems address interactive information-seeking needs impressively, their real-world deployments face latency and resource-consumption challenges. We present FrankenReport, an interface for long-form knowledge-seeking report generation that supports adaptive early exiting per section: it evaluates intermediate outputs during generation and predicts whether further targeted computation will yield significant quality gains. In a simulation study, FrankenReport outperforms random allocation baselines by a large margin (up to 4x) under low budgets and smoothly recovers full-pipeline quality as the budget grows, showing that future quality gains are predictable from intermediate drafts. Through experiments and user studies, we further show that despite varying preferences across users and topics, FrankenReport adapts to simple, natural user feedback as efficiently as methods requiring much costlier supervision such as generated drafts and explicit rationales.
Read the full abstract

All authors Zhengping Jiang, Gonzalo Ramos, Jina Suh, Shiqian Rachel Ng, Elias Stengel-Eskin, Justin Svegliato, Benjamin Van Durme, Andy Huntington, Sam Thomson

Combinatorics

Resolution of a problem of Mohar on non-positive inertia

For a graph $G$ of order $n$, its positive, negative and non-positive inertia is the number of positive, negative and non-positive eigenvalues of its adjacency matrix $A(G)$, respectively. Mohar asked whether every graph with $k$ non-positive eigenvalues has order $O(k^2)$ as $k\to \infty$. Using NEPS, we construct a sequence of non-singular graphs with negative inertia $k$ and order $\Omega(k^{\frac{7}{3}})$ as $k\to \infty$, thus resolving Mohar's problem. Our result also strongly refutes a recent conjecture of Akbari, Elphick, Kumar, Pragada, and Tang involving positive and negative inertia.
Read the full abstract

All authors Clive Elphick, Hitesh Kumar, Shivaramakrishna Pragada, Thomás Jung Spier

Optimization and Control

A 5/3 Guarantee for Echelon Stock (R,nQ) Policies in Two-Stage Stochastic Serial Systems

Fixed shipment costs encourage large orders, while inventory and customer backlog costs favor frequent replenishment. In a serial system, the stages must also coordinate their lot sizes and material availability. We study the classical echelon stock $(R,nQ)$ policy: the downstream stage requests a fixed lot, the upstream stage orders an integer multiple of that lot, and transfers contain complete lots. For unit Poisson demand, deterministic lead times, setup and holding costs at both stages, and full backlogging, we prove that the optimized cost in this class is at most $5/3\approx1.667$ of the optimal admissible system cost, uniformly over all nonnegative cost rates, demand rates, and lead times. The guarantee holds even when the selected policy pays a setup charge for every lot while the optimal benchmark pays only once per aggregate shipment. To obtain the guarantee, we use a common inventory calculation that allows independently chosen replenishment frequencies while respecting upstream material availability. We compare its optimized value, an upper bound on the best integer-ratio policy cost, with a novel lower bound on the optimal system cost that combines allocated single-stage costs with the minimum inventory cost when setup costs are zero. We also show that no uniform factor below $(1+\sqrt{3})/2\approx1.366$ is possible for the classical class. Existing numerical experiments place best-found classical policies within $22.7\%$ of an evaluated lower bound across three broad grids. The results quantify both the reliability and the limits of a simple operating rule and explain how shipment consolidation, replenishment timing, and flexibility affect its value.
Read the full abstract

Analysis of PDEs

An Inverse Problem for Determining the Piston Speed from a Given Lipschitz Leading Shock

We analyze an inverse problem for determining the piston speed and the associated flow field from a prescribed leading shock and the initial data in a shock tube. The gas flow is described by the isentropic Euler equations (i.e., the $p$-system), while the trajectory of the leading shock is prescribed as a given Lipschitz curve. Under an Oleĭnik-type entropy condition on the leading shock, we develop a modified wavefront tracking scheme to construct the flow field behind the shock. This construction enables us to determine the corresponding piston speed and the associated flow field.
Read the full abstract

All authors Gui-Qiang G. Chen, Qianfeng Li, Yun Pu, Yongqian Zhang

Astrophysics of Galaxies

Deep learning from the crowd Fundamentals of morphological galaxy classification

Aims. The objective of this work is to adapt a deep neural network model to perform galaxy morphological classification trained from crowd annotations, considering the training scheme, the agreement between the annotators, and the hierarchy. Methods. We use Galaxy Zoo 1 as our experimental testbed and trained a convolutional neural network (CNN) for the automatic classification of galaxies' morphologies. We analyze the impact of the following aspects on the classification accuracy and training efficiency: (i) Training only the last layer vs. training all the network; (ii) Classification with only the CNN vs. considering the hierarchy; (iii) Comparing the models trained with different amounts of data and levels of agreement between the annotators; (iv) Training by stages, transferring knowledge from one model to another; and (v) Combining several models as an ensemble. Results From the experiments, we derive the following results: (i) Training all the layers in the network significantly improves the accuracy (10% increase in exact match), compared to training only the last layer; (ii) There is a tradeoff between the amount of data and the level of agreement between the annotators used for training; (iii) Using the hierarchy can improve accuracy when the amount of training data is reduced; (iv) Training by stages through transfer learning (curriculum learning) produces higher accuracy for limited data; (v) Ensembles can improve accuracy; (vi) Models achieve a low accuracy for the most difficult cases, but, if we consider hierarchical measures, we can derive useful results for upper levels in the hierarchy. An accuracy above 99% is achieved when training all layers of the network and considering a high agreement between the annotators. Conclusions. Training deep learning models from crowd annotations involves additional challenges than learning from hard annotations.
Read the full abstract

Computation and Language

Reliability, validity, and diagnostic evidence for multi-model LLM short-answer scoring

Large language models (LLMs) are increasingly used or proposed for educational scoring, but single-model and single-run evaluations provide limited evidence for assessment use. Short-answer scoring requires evidence about reliability, validity, severity, diagnostic value, and failure cases. This study evaluated repeated multi-model OCG-PRES guided LLM scoring for short-answer assessment. The analysis used 996 SciEntsBank responses. GPT, DeepSeek, and Qianwen each scored every response across three independent runs using five OCG-PRES dimensions: concept coverage, relation accuracy, reasoning completeness, contradiction control, and domain relevance. Scores were evaluated against official binary and five-category labels and compared with non-LLM baselines based on answer length, Jaccard keyword overlap, TF-IDF cosine similarity, and a combined traditional logistic model. Repeated-run reliability was high for all models, with ICC(3,k) = .977 for GPT, .992 for DeepSeek, and .981 for Qianwen. DeepSeek was the most stable across runs. GPT showed the strongest official-label alignment by AUC (.909), while Qianwen was stricter, with higher precision but lower recall under the fixed threshold = 3.0 rule. OCG-PRES scores followed expected diagnostic patterns across five official categories and outperformed all non-LLM baselines in AUC and F1. Repeated multi-model OCG-PRES scoring provides reliability, validity, and diagnostic evidence for LLM-assisted short-answer scoring. The findings support cautious, evidence-based use as a scoring support tool rather than a replacement for human judgement.
Read the full abstract

Numerical Analysis

A review of weakly enforced Dirichlet boundary conditions in computational flow analysis

Strongly enforced Dirichlet boundary conditions require highly refined near-wall meshes to resolve steep velocity and thermal gradients. This introduces high computational costs, especially for practical flow simulations. Weakly enforced boundary conditions alleviate this burden by acting as a variationally consistent near-wall model. By allowing a controlled slip at the solid wall, weak enforcement recovers accurate flow quantities on coarse boundary-layer meshes across both incompressible and compressible regimes. Furthermore, weak boundary conditions serve as the fundamental enabling technology for immersogeometric analysis. Because the weak operator evaluates boundary integrals independently of the background mesh, high-fidelity flow analysis can be performed directly on complex geometries without fitting a mesh to the surface. This capability has facilitated direct geometry-to-analysis workflows for boundary-representation CAD models, raw point clouds, photogrammetric reconstructions, and segmented medical images. This review examines the unified mathematical development of the weak boundary condition framework from scalar advection-diffusion equations to the full Navier-Stokes equations, illustrating its versatility and robustness across incompressible and compressible flows, whether in traditional boundary-fitted, sliding-interface, or advanced immersogeometric applications.
Read the full abstract

Analysis of PDEs

Ill-Posedness of the Euler Equations Linearized around Homogeneous Steady States

Let $L^2_m(\mathbb{R}^2)$ be the space of square-integrable functions on $\mathbb{R}^2$ with $m$-fold rotational symmetry. Let $\overline{\omega}(r,\theta) = r^{-\alpha}f(\theta)$ be a homogeneous steady state of the two-dimensional incompressible Euler equations with $(m\cdot l)$-fold rotational symmetry. If $f(\theta)$ is a constant function we say that $\overline{\omega}$ is a radial power-law vortex. We prove that the incompressible Euler equations in vorticity form, linearized around any homogeneous steady state $\overline{\omega}$ that is not a radial power-law vortex, are ill-posed on $L^2_m(\mathbb{R}^2)$ for any $m\geq 2$, any $l\geq 1$ and $\alpha\in (0,1)$.
Read the full abstract

Combinatorics

On some k-fold generalizations of Lovász theta and their sandwich theorems

We study several $k$-fold generalizations of the Lovász theta function associated with the maximum $k$-colorable induced subgraph problem. The first is the Narasimhan–Manber parameter $\vartheta_k$. We prove that, for graphs whose adjacency matrix belongs to a homogeneous partially coherent algebra, this parameter is recovered by the theta number of the Cartesian product with the complete graph on $k$ vertices. This class includes distance-regular and $1$-walk-regular graphs, and thus our result generalizes a theorem by Sinjorgo and Sotirov (2022) for graphs that are vertex- and edge-transitive. We introduce a new parameter $\varphi_k$ obtained from orthonormal representations of graphs and show the inequality $\varphi_k \leq \vartheta_k$. For both parameters, we study the smallest $k$ for which the parameter is equal to the number of vertices; these saturation parameters yield lower bounds on the chromatic number. We determine which vertex-weighted versions of these parameters are gauges, and discuss a natural definition for the $k$-fold theta body of a graph. We conclude with open questions comparing $\vartheta_k$, $\varphi_k$, $\vartheta(G\square K_k)$, and related convexifications.
Read the full abstract

All authors Marcel K Carli Silva, Gabriel Coutinho, Thiago Oliveira, Levent Tunçel

Saturday 5 September

Statistics Applications

Geometry-Aware Smoothing Selection for Noisy Scattered Surface Reconstruction: Evidence from Synthetic Benchmarks and Water Quality Data

Sampling geometry may influence both reconstruction stability and the smoothing needed to control noise. This study distinguishes geometry-dependent failure of exact multiquadric radial basis function (RBF) interpolation from variation in validation-selected smoothing. A synthetic benchmark crossed four response families, four 16-point geometries, three slice levels, five observation regimes, and 24 repeated holdouts, yielding 17,280 method-level evaluations. Clough–Tocher cubic interpolation, exact RBF, and smoothed RBF were compared, followed by a Water Quality Portal pilot with 1,950 complete activities. On the synthetic shared-valid domain, exact RBF achieved the lowest noise-free mean RMSE (0.0206), whereas smoothed RBF performed best under noise (0.1807 versus 0.2414 for cubic and 0.6106 for exact RBF). Clustered-gap support produced the largest noisy exact-RBF RMSE (1.2706). Geometry added little explanatory power for log-transformed smoothing: the combined model had $R^2=0.4259$, compared with 0.4096 for noise alone. Reported failure-regression results suggest a stronger geometry signal, but remain unverified without their original model specification. In the environmental pilot, smoothed RBF achieved mean RMSE 0.9877 with no negative-$R^2$ failures. Conflicting responses at identical predictor coordinates made exact interpolation unsuitable. All environmental smoothed-RBF fits selected the largest candidate, preventing geometry–smoothing correlation analysis. The findings motivate geometry-aware failure screening but do not establish a transferable rule for smoothing magnitude. Independent multisite validation, appropriate coordinate scaling, joint scale–smoothing tuning, and conditioning diagnostics are needed before an adaptive policy can be claimed.
Read the full abstract

Statistics Applications

Quantile Regression Postprocessing with Negative Electricity Price Forecasts

Negative electricity prices create market-relevant risks for trading, bidding, renewable revenues, and flexibility valuation. Quantile regression-based postprocessing is a widely used approach for converting point forecasts into predictive distributions. We show that the standard specification, which imposes a single linear relationship between the point forecast and predictive quantiles, can become structurally incoherent when the conditioning point forecast is negative. In such cases, the resulting predictive distribution may exhibit not only ordinary quantile crossing, where adjacent quantiles are in the wrong order, but also a severe version of crossing, which we refer to as tail-quantile inversion. This can reverse downside and upside risk signals. To address this problem, we propose sign-dependent QR-based postprocessing specifications that allow the quantile function to differ between positive and negative point forecasts. Using German day-ahead electricity prices as an empirical illustration, we show that these specifications improve predictive accuracy, particularly for negative point forecasts and for extreme probability levels. Moreover, when combined with joint estimation across quantiles, the sign-dependent specifications largely eliminate tail-quantile inversion and all quantile crossing. The results demonstrate that the sign of the point forecast should be explicitly accounted for when constructing probabilistic forecasts in markets where negative prices can occur.
Read the full abstract

Computational Physics

A Boson-informed neural network for Ground States of Many-Body Schrödinger equations

We propose a symmetry-preserving Boson-informed neural network (BINN) for the variational Monte Carlo computation of interacting-boson ground states in continuous space. The framework explicitly separates the two dominant physical scales of the system. At the microscopic interaction scale, a cell Jastrow factor absorbs the leading near-contact stiffness, while at the macroscopic scale, a potential-aware envelope encodes the external confinement, leaving the neural network to capture a smoother many-body residual. Across both two- and three-dimensional tests, the BINN approaches the Hartree and Gross-Pitaevskii limits with an empirical $\mathcal{O}(N^{-1})$ finite-size trend in both energy and density. The energies agree with diffusion Monte Carlo to within $0.06\%$ for the particle number tested up to $350$ in the three-dimensional shrinking-range case. Applications are made under various trapping scenarios, justifying robust performance of BINN.
Read the full abstract

All authors Zhizhong Kong, Jerry Zhijian Yang, Cheng Yuan, Xiaofei Zhao

Differential Geometry

A Quantitative Reeb-Form Obstruction for Real Hypersurfaces of Strict Nearly Kähler Six-Manifolds

Let $M^5$ be an oriented real hypersurface of a six-dimensional strict nearly Kähler manifold $(\widetilde M^6,\widetilde g,J)$, and let $\xi=-JN$ and $\eta=g(\,\cdot\,,\xi)$ denote the induced Reeb field and Reeb one-form. We prove the pointwise obstruction \[ (d\eta)_p\neq0\qquad\text{for every }p\in M. \] The proof is local and uses only the hypersurface identity \[ \nabla_X\xi=\phi AX-G(X,N) \] together with the six-dimensional strict nearly Kähler algebra. If $(d\eta)_p=0$, then \[ \phi A+A\phi=2G_N,\qquad G_NX=G(X,N), \] which forces the Hopf condition. On the contact hyperplane $\mathcal D_p=\ker\eta_p$, the endomorphism $\phi A+A\phi$ then commutes with $\phi$, whereas $G_N$ anti-commutes with $\phi$ and satisfies \[ G_N^2=-\alpha I_{\mathcal D_p},\qquad \alpha>0. \] This contradiction proves the obstruction. As a connection-theoretic consequence, for the standard Yano semi-symmetric metric connection associated with $\xi$, \[ \widehat\nabla_XY=\nabla_XY+\eta(Y)X-g(X,Y)\xi, \] one has in dimension five \[ \operatorname{Skew}\operatorname{Ric}^{\widehat\nabla}=-\frac32\,d\eta, \] and hence $\operatorname{Ric}^{\widehat\nabla}$ is nowhere symmetric. The obstruction also rules out any local induced almost contact class whose defining conditions include $d\eta=0$. We illustrate the result on the standard homogeneous strict nearly Kähler spaces.
Read the full abstract

Differential Geometry

Gradient Ricci Solitons on PTR-Submanifolds of Complex Space Forms

We study gradient Ricci solitons on partially totally real (PTR) submanifolds of complex space forms. In nonzero ambient holomorphic sectional curvature, the Gauss–soliton equation reconstructs the square $P^2$ of the tangential Kähler operator from the Hessian endomorphism and an extrinsic Ricci operator. Consequently, the PTR tangency spectrum is encoded by intrinsic soliton data together with the second fundamental form. For compact minimal PTR solitons with positive ambient curvature we obtain $\lambda\le c(n-1)/4$, with a strict version under a relative-nullity condition. We also prove compact triviality criteria, an exact trace-free anisotropy identity, a parallel-data rigidity result, and a nonexistence theorem for compact real hypersurfaces. A flat multi-level Gaussian model and a nonflat Segre $CR$-product provide explicit consistency checks and identify the limiting cases of the reconstruction formulas.
Read the full abstract

Commutative Algebra

Alternating Divisoriality of Prime Ideals along Infinite Chains of Bézout Domains

A question posed in the literature asks whether the divisoriality of a prime can alternate indefinitely along an ascending chain of overrings of a Prüfer domain or a descending chain of same-quotient-field underrings, and what happens at the limit. We construct such chains with all finite-stage domains Bézout and show, in both directions, that the limit prime can be either divisorial or nondivisorial.
Read the full abstract

Computational Complexity

Exact Kernel Transfer to Clique Complexes and the Hardness of Normalized Persistence

For clique complexes $X_1\subseteq X_2$, normalized persistence in degree $d$ is $\operatorname{rank}[H_d(X_1)\to H_d(X_2)]/\dim H_d(X_1)$. Estimating it requires exact endpoint homology and the inclusion-induced map, even with inverse-polynomial endpoint Laplacian gaps. We prove that additive-error $1/24$ estimation is hard for $\mathsf{BQP}_{1}^{G_2}$, the perfect-completeness class over the exact gate set $G_2=\{X,\mathsf{CX},\mathsf{CCX},H\otimes H\}$, and hence for $\mathsf{BQP}_{1}$ over every finite gate set with entries in a cyclotomic field $\mathbb{Q}(\zeta_{2^k})$, even for unweighted clique complexes. The main tool is a finite-certificate kernel-transfer theorem. For a fixed palette of weighted clique gadgets satisfying finitely many exactly checkable local conditions, every unit chain $x$ of the full geometric complex satisfies $\operatorname{dist}(x,K)^2\le C(t\lambda^2+\langle x,\Delta x\rangle/(g\lambda^{26}))$, where $K$ is the embedded kernel of the simulated projector Hamiltonian, $g$ its gap, $t$ the number of gadgets, and $\lambda$ the private vertex weight. Since $\lambda$ is chosen independently of $g$, the geometric Laplacian has exactly $\dim K$ zero modes and a gap linear in $g$ above them. Exact fillings identify the endpoint homology with a quotient $V/W_A$ of the register cycle space, and nested term sets induce the natural quotient epimorphisms, so the persistent rank equals the later kernel dimension without any choice of compatible harmonic representatives. A fixed eight-dimensional label register turns a $\mathsf{BQP}_{1}^{G_2}$ verifier into instances with $\beta_d(X_1)=8$ and normalized persistence $3/4$ or $1/8$, and an established common-copy blow-up transfers everything to unweighted graphs.
Read the full abstract

Machine Learning

Travel Time Prediction in Supply Chain Management Using Machine Learning

The purpose of this research is to find data and methods using machine learning and deep learning to correctly predict the estimated travel time for transportation and logistics in a supply chain system. The supply chain ecosystem is very complex and heavily relies on the transportation and logistics of raw materials and finished goods. Accurate travel time estimation is critical because it helps supply chain members to improve logistics consistency and performance. This helps in planning, demand forecasting, lead time management and assembly planning. The logistics on the delivery side of the customer also plays a crucial role in customer satisfaction and voice of customer. With the collection of huge historical data and using novel techniques, the research builds an accurate model to predict travel time of inventory.
Read the full abstract

Data Structures and Algorithms

Dense Interprocedural Dominance in Acyclic Graphs: Context Bounds and Compact Queries

Interprocedural dominance asks which program points every matched call-and-return execution must pass on its way to a target. An analysis that writes out immediate dominance pays for the transitive reduction of this order, the minimal dominator graph. Measured instances of that graph look near-linear, with edge-to-node ratios around 1.1–1.26~\cite{DeSutter2007}, but no program property is known to force this: it is unclear which structural restrictions imply sparse output, and whether dense output also rules out cheap queries. This paper answers both through calling contexts. A static node stands for the group of its reachable configurations; write $\kappa$ for the largest group size and $\alpha$ for how many nodes have more than one configuration. One context per node keeps the tree with $n-1$ edges. Two already allow $\Theta(n^2)$ cover edges in a sparse, loop-free, recursion-free ICFG with total degree three, constant-size procedures, and logarithmic stack depth, while the reachable configuration graph itself stays linear, so the blowup comes from grouping, not from many configurations. Three keep $\Theta(n^2)$ covers even when the explicit call/return graph is acyclic: all target-reaching runs visit the mandatory procedures in the same order, and extra runs that stop short of any target remove the comparabilities that would otherwise kill the covers. Complementing the lower bounds, $\alpha$ ambiguous nodes admit at most $\min\{\floor{n^2/4},(\alpha+1)(n-1)\}$ cover edges, with $\Omega(\alpha n)$ attainable at stack depth one. And for every ICFG with at most two contexts per node, deleting a context group reduces dominance to dual-failure reachability, giving $O(n)$-word, $O(1)$-query representation after polynomial preprocessing. Explicit cover size, context ambiguity, and query-space cost are three separate quantities.
Read the full abstract

2609.27818 · 5 Sep · We identified a substantive technical issue in the proof of the main theorem that affects several downstream results. Addressing the issue requires substantial reworking of the current technical development, so we are withdrawing the paper in its current form

Metric Geometry

Compact Screens and Pyramidal Compactification of Geometric Data Sets

We introduce an observable distance that compares realvalued features through a fixed bounded coordinate. The coordinate retains the distinction between finite feature values while compressing their independent escape to infinity, the source of nonseparability for the classical observable distance on all geometric data sets. The new distance makes the full class separable and geodesic: every pair is joined by a constant-speed path, and on metric measure spaces the induced topology agrees with the concentration topology. From the same coordinate we construct compact screens, whose features take values in one fixed compact interval. The screened class is Polish and geodesic for the Box distance. Organizing its finite-feature quotients by the feature order yields a compact pyramid space. Pyramids generated by single compact screens form a dense subspace, so this pyramid space compactifies the original class after passage to compact screens. Convergence is detected through the Box-Hausdorff behavior of every finite measurement layer. Finally, taking sum-metric products with a common metric measure factor is nonexpansive for the new distance. More precisely, the comparison determined by a prescribed coupling of the original spaces and the diagonal coupling of the common factor is preserved, whereas optimization over all product couplings yields the nonexpansive inequality.
Read the full abstract

Computation and Language

A Tutorial on Prompt Engineering: From Messy Thoughts to AI Workflows

This paper treats prompt engineering as a discipline for turning informal human intent into structured AI work specifications. It develops the practice as a sequence of reusable design moves: define the work, construct only the context the answer depends on, choose a role, or a moderated panel of roles, as an attention lens, and state affirmative quality targets, reserving prohibitions for hard boundaries. To keep prompts lean, it adapts two classical principles, Occam's razor and Chekhov's gun, so that every instruction earns its place. For consequential tasks, it adds structured critique through steelmanning and premortems, followed by verification and, where tools or multi-step actions are involved, agentic operating loops with explicit boundaries and escalation. Aimed at a general readership, this tutorial is not a benchmarking study; it offers a practical, technically grounded path from casual prompting to disciplined AI workflow design, illustrated by a worked example that carries one task from a weak prompt to a strong specification. The framework is presented as principles and checklists that remain useful as models and tools change. The strongest prompt is rarely the longest prompt; it is the one that makes desired behaviour, required sources and checks, and success criteria unmistakable.
Read the full abstract

All authors Erfan Loweimi, Hadi Daneshvar, Samira Loveymi, Samir Ouelha, Zhengjun Yue, Hajar Mozaffar, Saturnino Luz

Computation and Language

Checkpoints Are Not Enough: Trust Calibration in CoSLR, a Human-AI System for Systematic Literature Reviews

Systematic Literature Reviews (SLRs) are essential for evidence-based research but remain time-consuming, requiring researchers to manage large volumes of publications across planning, screening, analysis, and reporting. Large language models (LLMs) can now produce fluent, well-structured review text, which makes it difficult to distinguish synthesis that was verified by a researcher from synthesis that merely appears authoritative. This raises the risk that unverified AI-generated synthesis enters the scholarly record carrying the credibility of a systematic review. We present CoSLR, a Human-AI collaborative multi-agent system that supports the SLR workflow through a modular three-phase pipeline using large language models and Retrieval-Augmented Generation (RAG), and that places explicit, mandatory human checkpoints on the path between generated output and its acceptance. In a survey-based study with 63 participants, the system was received positively: 27 of 63 participants (42.9 percent) rated its usability highly, indicating that the mandatory checkpoints did not come at the cost of a workable interface. However, a checkpoint safeguards the review only if researchers use it to verify: 22 of 63 participants (34.9 percent) reported that they would trust AI-generated summaries and reports without additional human checking after only a short interaction with the system. These findings indicate that Human-AI collaboration can support literature review work, but that the effectiveness of human oversight depends on whether users are willing to exercise it. This is a calibration problem that interface design must address directly, not assume.
Read the full abstract

All authors MD Aidul Islam, Malik Abdul Sami, Muhammad Waseem, Zeeshan Rasheed, Kai-kristian Kemell, Zheying Zhang, Pekka Abrahamsson

Machine Learning

CHART: A Harness-Rotation Curriculum for Harness-Robust Search Agents

Search agents are usually trained under a single harness. But once an agent is deployed in a real application, its harness is frequently updated (e.g., a rewritten system prompt) to fit production needs. This exposes a fragility of post-trained agents: because a learned behavior is entangled with its training harness, even a harness update that leaves the task unchanged can fail to elicit the behavior. We train a search agent to perform parallel search, a popular strategy for improving both search efficiency and performance. We find that training under a fixed harness makes the behavior harness-local, overfit to that harness's surface form: when the harness changes, the model falls back to serial search. An intuitive fix is harness augmentation, but simply training on more harnesses does not resolve the problem. GRPO learns from the reward gap between parallel and serial rollouts of the same question: a small harness pool saturates that gap early, while a large pool dilutes the per-harness signal too thinly for any harness to consolidate. We therefore propose Curriculum HArness Rotation Training (CHART), a rotating curriculum that lets a search agent gradually consolidate parallel search across harnesses. At each periodic evaluation, CHART "graduates" the harnesses whose expected behavior is learned and replaces them with still-learnable ones, keeping the reward gap alive throughout training. Starting from the same harness pool, CHART makes the model learn parallel search on all harnesses, whereas static augmentation succeeds on at most half of them. The behavior also carries to held-out harnesses: CHART parallelizes on 89% of held-out turns, against at most 5% for the static pools. It further transfers to a new QA task and search environment, improving pass@1 by 5.6pp over the best static pool. Finally, CHART-trained agents benefit more from meta-harness search than baselines.
Read the full abstract

All authors Xinlu Zhang, Ying-Chun Lin, Zhihan Zhang, Besnik Fetahu, Xi Chen

Computation and Language

The Corroboration Illusion: When More News Makes LLM Forecasts Less True

Large language models (LLMs) are increasingly used to forecast real-world events by retrieving and reasoning over news. We show that this dependence on an open, crawlable news corpus creates a new attack surface: an adversary who can merely publish articles–without access to the retriever, the model, or the user's queries–can systematically move the forecaster's output probabilities. We formalize news-corpus poisoning of probabilistic forecasters, a threat model distinct from prior RAG poisoning, which targets factual answers or opinion polarity rather than calibrated probabilities. We evaluate the attack on 500 resolved ForecastBench questions against a 17.4M-article Common Crawl News corpus with a strict crawl-date cutoff, using three retrieval-augmented forecasters built on open 7-8B models. A single LLM-written article per question flips 56% of forecasts across the 0.5 boundary; five articles flip 69-73% and shift probabilities by +0.13 to +0.22 net of a neutral-article placebo, degrading the Brier score from 0.18 to 0.37. The effect is monotone in the number, retrieval rank, query similarity, and context share of injected articles, transfers across model families, and is unaffected by the claimed publisher. We then evaluate three natural defenses–source allow-lists, isolate-then-aggregate forecasting, and perplexity filtering–and show that each has a cheap bypass: spoofed publishers, majority poisoning, and higher-temperature generation, respectively. Our results indicate that probabilistic LLM judgments inherit the full fragility of the information supply chain they consume.
Read the full abstract

Computation and Language

Do Chess Explanations Reflect Model Decisions? Behavioral and Token-Level Tests of LLM Reasoning Faithfulness

Large language models can produce fluent explanations for chess moves, but plausible language does not necessarily reflect the reasoning behind a decision. We study this question in chess, where the board state is fully observable, legal actions can be enumerated, and move quality can be evaluated independently. Across 200 Lichess endgame puzzles, we test explanations using move recoverability, decoder-side controls, and token-level scoring of legal candidate moves. Unmasked explanations make generated moves easy to recover, but this advantage drops sharply after explicit move hints are removed. Under strict masking, explanations provide only small and decoder-dependent gains over the board state alone. Token-level scoring shows that explanations can nevertheless alter move preferences: random but plausible explanations from other puzzles reduce the probability of the correct move, indicating that irrelevant reasoning text is not simply ignored. We also find that recognizable endgame motifs can make generated moves easier to recover without reliably improving move correctness. Together, these results show that linguistic plausibility, consistency with a generated action, and solution correctness are distinct properties. Fluent chess explanations can influence action preferences and support a coherent move narrative while providing only limited evidence of faithful reasoning.
Read the full abstract

Machine Learning

Universal Observatory Graphs for Distributed Sky Coverage and Artificial Intelligence Based Interplanetary Routing

This research proposes the Universal Observatory Graph (UOG), an AI-driven framework for distributed astronomical observation across the Solar System. The proposed architecture models autonomous observatories located at the Sun planet L2 Lagrange points as nodes in a weighted graph, while communication links are represented as graph edges characterized by multi-objective physical and operational metrics, including interplanetary distance, communication latency, transmission power, and link reliability. The resulting graph provides a unified mathematical representation of a cooperative interplanetary observatory network. This proposal examines a six-observatory Solar System configuration comprising Earth, Mars, Jupiter, Saturn, Uranus and Neptune. Instantaneous sky coverage is evaluated independently using a 200,000 direction Fibonacci sphere, a 2,000,000 direction fixed seed Monte Carlo calculation and deterministic spherical integration. All three methods yield complete network union coverage, approximately 0.43% complete six observatory intersection and approximately 24.96% mean pairwise Jaccard similarity under the adopted pointing model. Communication routing is subsequently formulated as a finite horizon Markov decision process and solved using tabular Q-learning. The reward balances node participation and a distance dependent reliability proxy against distance, light time latency and a distance squared transmission power proxy. The learned Earth-Saturn-Uranus-Neptune route is also the highest discounted return route among all 41 feasible simple paths under the four hop constraint. The framework provides a reproducible baseline for sequential coverage assessment and multi objective routing; time dependent ephemerides, mission specific visibility, calibrated link budgets and scalable graph policies remain future work.
Read the full abstract

Computation and Language

Replay-Gated Neural Execution: Decoupling Persistent Behavioral Specifications from Neural Realizations in Frozen Language Models

Input-conditioned neural interventions raise a runtime question: what persists when one behavioral specification admits multiple actions whose validity depends on execution state? We introduce replay-gated neural execution, separating five objects: a persistent behavioral predicate, its state-indexed certified realization set, a transient action witness, a budget-limited finder, and execution authorization. Candidates undergo isolated FP32/BF16 replay of the frozen model; commitment additionally requires a valid run audit. Experiments on Qwen3-0.6B and SmolLM2-360M-Instruct establish distinct failure modes for these objects. Independent initializations yield distinct certified actions in all 24 tested fixed-state cells. Unchanged SmolLM2 witnesses remain certified in all 128 native states but only 66 of 384 off-diagonal transfers. All 767 archived Qwen witnesses replay successfully, yet a budget-limited finder misses one known-realizable cell in all three prespecified runs. Of 1,141 replay-submitted candidates, 174 fail item certification. A frozen three-tier cascade uses these boundaries to reject uncertified proposals and escalate audit-valid search misses. On 256 previously sealed Qwen Fresh requests, 221 first certify at the lowest-cost tier and all 256 receive audited authorization, with no observed bypass. Relative to frozen full search, the median singleton search-and-certification cost ratio is 0.1055 and P95 is 1.3485, including failed tiers. Within the studied behavioral family on two small models, these results support state-indexed, set-valued execution semantics: specifications persist, search proposes witnesses, and replay certification plus run audit grants execution authority.
Read the full abstract