Machine Learning
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Sunday 6 September
Kernel methods separate data representation from decision-making, but typically require the kernel to be chosen in advance. We show that this kernel can instead be learned by alignment, and develop the resulting framework through the recently introduced Collaborative Learning and Inference (CLaI). We show that Collaborative Learning can be viewed as a kernel alignment process, in which an embedding is trained so that its induced similarity matches a label-derived target kernel. We also prove that Collaborative Inference is equivalent to kernel Bayes classification with Parzen-window density estimation. Motivated by these perspectives, we generalise CLaI by replacing cosine similarity with a learned Mahalanobis distance and extend it to multiclass classification. On CIFAR-10, PathMNIST, and SleepEDF, the Mahalanobis formulation improves accuracy, converges faster, and yields lower calibration error than the cosine-based variant. Auxiliary experiments further support these connections, showing that CLaI produces latent signals of the same form as a Gaussian process, while achieving competitive calibration on sepsis prediction. Together, these results establish a principled learned-kernel framework that unifies representation learning, kernel alignment, and Bayesian classification, and extends naturally to the multiclass setting.
Machine Learning
Steering Under Compression: Dose-Response, Capability Cost, and Failure Asymmetry in Quantized LLMs
Inference-time activation steering enables behavioral control of large language models without parameter modification, while post-training quantization reduces memory and compute costs for deployment. Despite their growing convergence in practice, the interaction between these two techniques remains uncharacterized. We systematically study activation steering under weight-only quantization (INT8 and NF4) across four open-weight 7-9B models and two behavioral targets: judged sentiment and judge-free reasoning length. Using an iso-effect framework that compares capability costs at matched behavioral effect, we find that sentiment steering survives quantization intact. After correcting a GSM8K parser artifact with a uniform v2.3.1 rescore, the pooled INT8 contrast is -0.010 (90% CI [-0.026, +0.007]), descriptively Equivalent under the preregistered three-label rule, while NF4 remains Inconclusive at -0.017 ([-0.067, +0.033]). In contrast, reasoning length exhibits a surprising asymmetric dose-response: lengthening is graded but terminates in cap-runaway and collapse, while shortening is a step function with only 12-30% shortening (model-dependent) before discontinuous failure. We expose a methodological pitfall: the naive iso-effect ladder anchors on the collapse floor for floor-bounded targets, and we introduce a censored construction that restores interpretable crossings. We also quantify a substantial baseline capability shift for Mistral-NF4 (0.545 to 0.365 GSM8K at alpha=0), demonstrating that compression can dominate the steering intervention. Despite this, steering vectors remain highly collinear with their FP16 siblings (cosine similarity 0.989-0.998 for INT8, 0.945-0.990 for NF4), confirming that the behavioral direction survives quantization even when the cost structure does not. All code and data are released.
Read the full abstract
Strongly Correlated Electrons
NMR evidence of pressure-induced structural transition and enhanced spin fluctuations up to 14~GPa in SrCu$_2$(BO$_3$)$_2$
The Shastry-Sutherland compound SrCu$_2$(BO$_3$)$_2$ has attracted considerable interest as a platform for exploring quantum phases and quantum phase transitions driven by magnetic frustration. The pressure-induced structural and magnetic phase transitions in SrCu$_2$(BO$_3$)$_2$, however, remain controversial. To address this issue, we performed high-pressure $^{11}$B nuclear magnetic resonance (NMR) measurements on SrCu$_2$(BO$_3$)$_2$ up to 14~GPa. The NMR spectra reveal two pressure-induced monoclinic phases. With pressure above 4~GPa and with temperature below 10~K, the rapid broadening of the NMR spectrum and the power-law behavior in the spin-lattice relaxation rate $1/T_1$ provide clear evidence for a gapless 3D antiferromagnetic (AFM) phase in the monoclinic phase. At an intermediate temperature range around 20~K, the emergence of the field-dependent NMR line splits resolves a two-dimensional, short-range ordered AFM phase; at temperature above 30~K, the sublinear power-law behavior of $1/T_1$ identifies an extended correlated paramagnetic regime.
Read the full abstract
Quantum Physics
Generation of entanglement statistics with a large-scale integrated photonic-electronic circuit
Continuous-variable quantum processors scale with the modes they can transform, and each transformation adds loss. We demonstrate optoelectronic quantum information processing, moving the mixing into radio-frequency electronics after homodyne detection. Squeezed vacuum falls on a large-scale integrated photonic-electronic circuit, a 32-channel coherent receiver array whose radio-frequency network mixes the analog photocurrents. A local-oscillator ramp reconstructs the two-mode covariance at the aperture, which is physical, and its partial transpose has smallest symplectic eigenvalue 0.99270, three standard deviations below the separability bound. Mode transformation becomes a circuit function whose scale follows integrated electronics. With electro-optic transducers in place of the photodiodes, the same receiver would transfer optical modes into superconducting circuits and back, toward a photonic-electronic quantum computer spanning the optical and microwave bands.
Read the full abstract
Data Structures and Algorithms
A Group-Based Resource Allocation Model for the Fractional Knapsack Problem
To solve the fractional knapsack problem, Dantzig's greedy rule orders items according to their value-to-cost ratio. This ordering introduces priority issues. An arbitrarily small perturbation to the input can change the allocation if the budget is exhausted between two items with very similar ratios. To mitigate that problem, we introduce a two-stage rule. We group items sharing attributes within a radius $\delta$. We then evaluate these groups in descending order of ratio and divide their group's budget share without further ranking. Consider a group featuring an aggregate capacity $U_G$, unit costs contained in $[w^-,w^+]$, and a representative value $\widehat{v}$. The group's loss relative to the exact optimum is bounded by $\widehat{v}\, U_G\frac{w^+-w^-}{w^++w^-}+\varepsilon_v U_G$, where $\varepsilon_v$ limits the group's internal value variation. Moreover, this harmonic factor remains tight for any group size. The overall loss becomes restricted to the single budget-binding group whenever the grouping remains order-compatible; thus, groups containing at most $K$ items suffer a per-item loss of $\mathcal{O}(\frac{K}{n})$. Should group ratio intervals exhibit an overlap of at most $\omega$, an additive term $\omega C$ degrades this bound. Within the separation margin between adjacent groups, the grouped allocation remains Lipschitz continuous with respect to cost data, exhibiting a modulus of $\frac{K}{w_{\min}}$. Computing this allocation takes $\mathcal{O}(n+m\log m+|\Gamma|\log|\Gamma|)$ time given $m$ groups and a boundary group $\Gamma$. Alternatively, the time complexity drops to $\mathcal{O}(n+m\log m)$ if a linear-time selection method identifies the boundary group's allocation.
Read the full abstract
Machine Learning
One Step, One Lead: Mitigating Higher-Order Interference in Multi-Domain Reinforcement Learning via Cross-Step Control
Reinforcement learning (RL) across multiple domains can broaden the reasoning capabilities of large language models (LLMs), yet joint training often degrades individual-domain performance and can destabilize optimization. Existing work typically diagnoses such interference from a single-step view using first-order gradient alignment or curvature-based proxies. We show that this view can miss a critical form of sequential interference: same-point domain gradients may remain nearly orthogonal even when consecutive realized updates partially reverse one another in output space. We further show that consecutive token log-probability footprints recover this interaction directly from adjacent checkpoints as a local second-order interaction in output space, without explicitly reconstructing same-step curvature. Building on this insight, we propose OSOL, which designates a focus domain at each iteration, uses the preceding checkpoint footprint to rank token-level rebound risk, and applies a drift-ranked, adaptively scaled correction within the standard GRPO update. Our analysis shows that this correction suppresses the targeted cross-step output backtracking component. Controlled studies further show that cross-step backtracking is more strongly associated with subsequent task damage than same-point gradient diagnostics, while the preceding footprint ranks future rebound risk more accurately than Hessian-based proxies. On Qwen3-30B-A3B, OSOL reaches a domain-macro average of 0.4822, improving by 5.7% over the strongest compared baseline, without explicit higher-order differentiation.
Read the full abstract
Machine Learning
Sector-Mean: Deterministic Initialization of K-Means Centroids via Angular Sector Partitioning
K-Means is one of the most widely used clustering algorithms, but its susceptibility to initial centroid selection remains a primary bottleneck for its convergence speed and clustering accuracy. This paper proposes Sector-Mean Initialization, a deterministic initialization strategy with O(N) time complexity that partitions the two-dimensional data space into angular sectors around the global centroid and initializes centroids using sector-wise means. We evaluate the method on established two-dimensional benchmarks (SIPU, Birch) and multiple real-world datasets, comparing against random, K-Means++, and Max-Min initialization under identical Lloyd iterations. The statistical analysis of Friedman's test (p<0.05) and Nemenyi post-hoc comparison indicates that, while delivering equivalent clustering quality as K-Means++ and Max-Min, Sector-Mean offers significant computational efficiency. Experimental results show that Sector-Mean reduces the initialization time by 74.9% and 59.8% in comparison to K-Means++ and max-min, respectively. And, it yields the lowest average number of iterations, achieving approximately 5% fewer iterations than K-Means++ and 16% fewer than max-min. These results highlight that Sector-Mean initialization offers a deterministic and computationally efficient initialization strategy while preserving cluster quality.
Read the full abstract
Machine Learning
Local and Global Stability in Performative Reinforcement Learning
In performative reinforcement learning the deployed policy shapes the environment that generates the learner's future data, and the natural solution concept is a performatively stable policy that is optimal in the environment it induces. Existing convergence guarantees rely on Lipschitz sensitivity assumptions on the environment map $\pi \mapsto (P_\pi, r_\pi)$, which are hard to verify and fail in settings such as multi-agent best-response dynamics. We instead study stability for mixtures of policies, and show that the resulting picture is fundamentally different from performative prediction, where randomization removes the need for any sensitivity assumption. We distinguish local mixed stability, an occupancy-weighted first-order relaxation that we show is equivalent to stationarity, from global mixed stability, which certifies against arbitrary deviating policies. Our first result is that a weighted per-state Hedge dynamic drives the local stability gap to zero at an $O(1/\sqrt{T})$ rate for an arbitrary, possibly discontinuous, environment map, both with exact and with trajectory feedback. The two notions genuinely differ: we exhibit an instance where local stability is achieved exactly but every mixture has global stability gap bounded away from zero. For global stability we introduce a bounded transition range assumption, strictly weaker than Lipschitz sensitivity, under which unweighted per-state Hedge converges up to a floor of $O(\gamma\epsilon_P/(1-\gamma)^3)$, and we prove a matching-in-$\epsilon_P$ lower bound of $\Omega(\gamma\epsilon_P/(1-\gamma))$ under trajectory feedback, so this floor is unavoidable. Finally, we extend both notions to $n$-player performative Markov games, obtaining local stability with no assumption on the joint environment map or game structure, and global stability for performative Markov potential games.
Read the full abstract
Mathematical Physics
Condensation and Collapse in the Mean-Field Limit of Rotating 2D Bose Gases with Two-Body and Three-Body Interactions
We consider a system of $N$ interacting bosons in a rotating harmonic trap in $\mathbb{R}^2$, where the two-body interaction is attractive and scaled as $N^{2\alpha}U(N^\alpha x)$ with $0<\alpha<1/12$, and the three-body interaction is repulsive and scaled as $N^{4\beta}W(N^\beta x,N^\beta y)$ with $0<\beta<1/24$. In the mean-field limit, the ground state energy is effectively described by a rotating cubic-quintic nonlinear Schrödinger functional. We analyze the collapse regime where the two-body coupling $a$ approaches the critical value $a_*$ and the three-body coupling $b$ tends to zero. The NLS ground states blow up with a universal profile given by the optimizer of the Gagliardo-Nirenberg inequality, and the energy satisfies $E^{\mathrm{NLS}} = (1-\zeta/4+o(1))\mathcal{Q}{\rm{pot}}\ell_n^2$ with $\zeta\geqslant0$ determined by the relative rates of $a_n\to a_*$ and $b_n\searrow0$. From the many-body theory, we rigorously justify this effective description: the quantum ground state energy converges to the NLS energy with the same asymptotic expansion in the collapse regime, and the many-body ground states exhibit complete Bose-Einstein condensation onto the universal blow-up profile.
Read the full abstract
Algebraic Geometry
When is a polynomial in three variables cylindrical?
We prove that a nonconstant polynomial $f\in\mathbb{C}[x_1,x_2,x_3]$ is cylindrical (that is, $f\in\mathbb{C}[x_1,x_2]$ after an invertible linear change of coordinates) if and only if its bordered Hessian determinant vanishes identically. The same conclusion holds over $\mathbb{R}$. We also give explicit counterexamples to this criterion in every dimension $n\ge4$.
Read the full abstract
Data Structures and Algorithms
Minimization Prophet Inequality with Bounded Costs
We study the cost-minimization prophet inequality problem, in which a decision-maker sequentially observes $n$ independent and identically distributed (IID) random variables. After each observation, the decision-maker must either accept the current realization and stop, or reject it and continue with the next variable. The goal is to minimize the selected value. Unlike the classical maximization setting, if none of the first $n-1$ values is accepted, the final realization must be selected. Esfandiari et al. (SIDMA 2017) showed that this problem does not admit constant-competitive online algorithms in general, which motivates subsequent work on restricted distribution classes, such as entire distributions (Livanos and Mehta, SODA 2024) and distributions with bounded extreme values (Livanos and Mehta, EC 2025). In this work, we focus on distributions with bounded support. For distributions supported on $[1,b]$, we characterize the competitive ratio of online algorithms as a function of both $b$ and $n$. We consider both distribution-aware and distribution-oblivious settings. In the distribution-aware case, we provide a nearly tight characterization of the optimal online algorithm, showing that its competitive ratio is at most $b^{(1-1/n)^n}$ for all $n\geq 2$. More significantly, we design a distribution-oblivious algorithm that achieves the same guarantee, with a sequence of predetermined thresholds that depend only on $b$ and $n$. Furthermore, we extend our analysis to the non-IID setting, where we show that a simple single-threshold algorithm attains an asymptotically optimal competitive ratio of $\Theta(\sqrt{b})$.
Read the full abstract
Complex Variables
(Semi-)Models for Slice Regular Functions on Real Division Algebras
We study slice regular functions on the real division algebras $\mathbb{H}$ and $\mathbb{O}$ from the point of view of equivalence under automorphism and conjugation actions. Motivated by recent orbit-theoretic descriptions in terms of the invariants $(Tr,N,cdiv)$, and by the quaternionic theory of $*$-conjugation by semiregular functions, we investigate whether a slice regular function can be replaced by a canonical representative in its equivalence class. The normalization considered in this paper is algebraic: we look for representatives whose isolated nonreal zeros are aligned in a single complex slice. We call such a representative a model when it belongs to the strong-equivalence class of the original function. We prove that purely vectorial functions always admit models, while in general strong equivalence is too rigid and such representatives need not exist. We therefore introduce semi-models, which preserve the symmetrization and the real and spherical part of the zero set, but are allowed to leave the original strong-equivalence class. Our main result proves that every slice regular function $f$ with $N(f)\not\equiv 0$ admits a semi-model. The construction clarifies the different roles of the invariants $(Tr,N,cdiv)$ and the distinction between strong equivalence, weak equivalence, and semiregular conjugacy.
Read the full abstract
Combinatorics
Feedback edge set in bipartite digraph
Let \(\beta(G)\) denote the minimum size of a feedback edge set of a digraph \(G\), and let \(\gamma(G)\) denote the number of unordered pairs of nonadjacent vertices. Motivated by the Chudnovsky–Seymour–Sullivan conjecture for \(3\)-free digraphs, we study the corresponding feedback-edge problem for bipartite digraphs. In the bipartite setting, \(\gamma(G)\) is taken to count only nonadjacent pairs with ends in distinct partite sets. We prove that every \(4\)-free bipartite digraph \(G\) satisfies \(\beta(G)\le \gamma(G)/2\). We also determine the exact Turán number of \(2k\)-free strong bipartite digraphs with partite sets \(X\) and \(Y\): if \(|X|,|Y|\ge k+1\), then the maximum number of edges is $$(|X|-(k-1))(|Y|-(k-1))+2k-2.$$ Finally, for the extremal case \(k=2\), we analyze the structure of \(4\)-free strong bipartite Turán digraphs and prove the sharper bound \(\beta(G)\le \gamma(G)/3\) for all such digraphs. This constant is attained by a natural balanced three-block construction.
Read the full abstract
Complex Variables
From $\log 2$ to $\pi/2$: the sharp asymptotic inradius of polynomial lemniscates
Let $R_n$ be the infimum of the inradii of $\{z:|p(z)|<1\}$ over monic degree-$n$ polynomials whose zeros lie in the closed unit disk. We prove $nR_n\to\pi/2$, matching the asymptotic obstruction supplied by $z^n-1$. We first establish the exact universal radius $2^{1/n}-1$ for disks centered at zeros, which recovers the $(\log2)/n$ bound. Small inradius then forces radial concentration of the zeros and decay of their low reciprocal moments. These estimates give an entire limit with a modulus reflection identity; a second rescaling produces an exponential tangent and strict sublevel disks of every radius below $\pi/2$. The proof is accompanied by a Lean~4 formalization and a step-by-step source index.
Read the full abstract
Machine Learning
A Theoretical Framework for Masked Pretraining (MPT)
Recently, Masked Pretraining (MPT) based on reconstruction pretraining tasks has risen to a promising self-supervised learning paradigm across various domains and achieves remarkable performance in multiple downstream tasks. However, the theoretical understanding of the working mechanism behind MPT is still limited. In this paper, we introduce a new theoretical framework to analyze MPT and understand the crucial role of masking in extracting meaningful representations. We establish theoretical connections between MPT and another popular self-supervised paradigm: contrastive learning. We prove that the masking technique implicitly creates positive pairs that are semantically similar and the reconstruction loss pulls them together in the feature space. Besides, as a result of the implicit alignment, we point out the dimensional collapse issue of MPT and propose a Uniformity-enhanced MPT (U-MPT) loss that can effectively address this issue and bring significant improvements in downstream tasks including linear evaluation, cross-dataset fine-tuning and out-of-distribution generalization on real-world data sets. Furthermore, we establish downstream guarantees of U-MPT and theoretically analyze the influence of masking strategies. Based on the theoretical analysis, we propose a new masking strategy which enhances the downstream performance of MPT and explains current improvements of masking strategies with our theoretical perspective.
Read the full abstract
Numerical Analysis
Rate of convergence of a fully discrete structure-preserving midpoint scheme for the stochastic Landau–Lifshitz–Gilbert equation
The stochastic Landau–Lifshitz–Gilbert (sLLG) equation is a strongly nonlinear stochastic PDE with a non-convex pointwise constraint arising in the theory of micromagnetics. We analyse a fully discrete, structure-preserving finite element approximation of the sLLG equation with coloured multiplicative Stratonovich noise on a bounded interval. The method utilises continuous piecewise affine finite elements, mass lumping, and midpoint time discretisation to preserve the unit-length constraint exactly at the finite element nodes. Under suitable regularity assumptions on the initial data and the noise, we establish uniform higher-moment stability and develop an error analysis for the scheme. The analysis exploits the geometric structure of the equation and the stochastic midpoint discretisation. For every $\gamma\in(0,\frac12)$, we prove first-order spatial convergence and temporal convergence of order $\gamma$ in the natural discrete energy norm, locally in mean square on events of arbitrarily large probability and, consequently, in probability. To the best of our knowledge, this is the first convergence-rate result for a fully discrete structure-preserving finite element scheme solving the stochastic Landau–Lifshitz–Gilbert equation.
Read the full abstract
Representation Theory
Criteria for discrete decomposability and admissibility in the branching problem
In this paper, we study when the restriction of an irreducible representation of a reductive Lie group $G$ to a reductive subgroup $H$ is discretely decomposable. T.\ Kobayashi gave a necessary condition for discrete decomposability in terms of associated varieties. In a previous paper, we proved that the condition is also sufficient if $(G, H)$ is a symmetric pair. The purpose of this paper is to prove the sufficiency for any reductive subgroup $H$ under some hypothesis on $G$ (e.g., $G$ is linear). As an application, we show that discrete decomposability is equivalent to $\mathfrak{h}$-admissibility when the compact factor of $\mathfrak{h}$ is maximal in some sense.
Read the full abstract
Fluid Dynamics
Unification of The Navier-Stokes Equation for Fluid Mechanics and Aeroacoustics
The Navier-Stokes equation for a compressible Newtonian fluid is reassessed, in view of recent investigation results on the foundations of aeroacoustics. In the process, the role and necessity of the so-called Stokes' hypothesis is reconsidered, which in its current practice omits the bulk viscosity. In contrast, the relevance of a new generalized kinematic viscosity is proposed, along with a new methodology that will enable estimation of second coefficient of viscosity completely obviating the necessity of making the Stokes' hypothesis. This opens up the vista of a new formulation for the Navier-Stokes equation applicable for both the subjects of fluid mechanics and aeroacoustics in a unified approach.
Read the full abstract
Commutative Algebra
Ranks of Verdier quotients of derived categories over local rings
Let $R$ be a commutative noetherian local ring with residue field $k$. Denote by $\operatorname{\mathsf{D^b}}(R)$ the bounded derived category of finitely generated $R$-modules. In this paper, we introduce the notion of ranks of triangulated categories and study Verdier quotients of $\operatorname{\mathsf{D^b}}(R)$ from the viewpoint of this invariant. Our main result determines the rank of the Verdier quotient $\operatorname{\mathsf{D^b}}(R)/\operatorname{\mathsf{thick}}(R\oplus k)$ in terms of the ranks of the categories of maximal Cohen-Macaulay modules on the punctured spectrum.
Read the full abstract
Metric Geometry
A finite set with noncompact closed convex hull in a Hadamard space
We show that there is a Hadamard space in which the closed convex hull of a set containing merely three points is noncompact, by providing an AI-generated construction. The seemingly innocuous problem of finding a compact or finite set with this property, or showing that none exists, was recorded by Gromov and periodically revisited by researchers in metric geometry, functional analysis, fixed-point theory, and geometric group theory. Nevertheless, it remained unanswered for over three decades. The construction itself is based on building an inductive sequence of CAT(1) metric graphs and taking the Euclidean cone over the limiting space. We discuss the history of this problem, the main ideas and details of the construction, and its relationship with existing techniques.
Read the full abstract
Optics
Energy-scalable single-burst attosecond emission at kiloelectronvolt photon energies
Attosecond soft-X-ray pulses at kiloelectronvolt photon energies would provide direct access to the L-edges of transition metals, enabling element-specific studies of charge, spin, and orbital dynamics on their intrinsic timescales. However, attosecond emission in this spectral region has remained elusive because the only phase-matched keV high-harmonic source demonstrated to date employed multicycle driving pulses without carrier-envelope-phase (CEP) control. A long-standing gap has therefore persisted between keV photon energies and CEP-controlled attosecond emission. Here, we bridge this gap using a CEP-stabilized mid-infrared laser system with pulse durations tunable to the sub-cycle regime. Through phase-matched highharmonic generation in low-pressure neutral helium under meter-scale loose-focusing conditions, we generate coherent soft-X-ray continua reaching 1.2 keV. Crucially, we observe CEP-dependent spectral modulation that reaches into the transition-metal L-edge region, providing experimental evidence consistent with the generation of isolated attosecond pulses. To demonstrate the capabilities of our source, we perform broadband soft-X-ray absorption spectroscopy spanning the Ti, Fe, Co, and Ni L-edges, as well as the O K-edge, and resolve near-edge fine structures. By combining meter-scale loose focusing with low-pressure phase matching, our approach addresses key energy-scaling limitations of previous keV high-harmonic sources and establishes a phase-matched, CEP-controlled keV soft-X-ray platform for attosecond spectroscopy of magnetic, quantum, and strongly correlated materials.
Read the full abstract
Superconductivity
$^{139}$La nuclear quadrupole resonance studies of pressurized La$_4$Ni$_3$O$_{10}$
Density-wave (DW) orders are considered as competing orders to unconventional superconductivity and are commonly seen in a variety of superconductors including but not limited to the recently discovered Ruddlesden-Popper-phase nickelates. By utilizing $^{139}$La nuclear quadrupole resonance, we systematically investigate into the nature of DW orders and their evolution under pressure in La$_4$Ni$_3$O$_{10}$. Spin and charge DW orders are found to be intertwined in this material, which is in stark contrast to those in La$_3$Ni$_2$O$_7$. Short-range DW orders are observed near 150 K, well above the development of long-range DW orders at around 139 K. Upon applying a hydrostatic pressure of 2.3 GPa, the transition temperatures of the short-range and long-range orders decrease at rates of 1 K/GPa and 10 K/GPa, respectively. Our results thus affirm that both spin density wave and charge density wave as competing orders with the superconducting state in La$_4$Ni$_3$O$_{10}$, and provide new insights into the interplay between DW orders and unconventional superconductivity.
Read the full abstract
Numerical Analysis
High-Order Structure-Preserving SBP Finite Difference Methods for the Vlasov-Maxwell System on Matrix-Free GPUs
In this paper, we present a high-order, stable summation-by-parts (SBP) finite difference method for solving the Vlasov-Maxwell system in a 2D2V phase space. Central SBP operators for the advection terms are not stable when the solution becomes non-smooth and fine-scale filamentary structures develop, as is typical in high-dimensional Vlasov-Maxwell simulations. To address this issue, the method is stabilized using high-order upwind SBP operators. High-order explicit Runge-Kutta methods are employed for time integration. We prove that the fully discrete scheme exactly conserves mass and preserves momentum up to truncation error. Furthermore, we present a matrix-free implementation of the method on modern GPU architectures. A range of challenging benchmark problems is solved to demonstrate the accuracy, robustness, and performance of the proposed scheme.
Read the full abstract
Differential Geometry
On Miyaoka-Yau Inequalities and Weil-Petersson Metrics
We use the Kähler-Ricci flow to give alternate proofs of several known Miyaoka-Yau inequalities and slope semi-stabilities, in particular, for the case of compact Kähler manifolds with semi-ample canonical line bundles and K-semistable Fano manifolds. Moreover, when the canonical line bundle is semi-ample, and the Kodaira dimension is $n-1$, we prove the Miyaoka-Yau quantity is equal to the intersection of the Weil-Petersson metric and the Fubini-Study metric on the canonical model. Consequently, equality holds in the Miyaoka-Yau inequality if and only if the pluricanonical map is a holomorphic fibre bundle.
Read the full abstract
Robotics
PLATO: Preintegration Learning from Accurate Trajectory Observations for Neural Inertial Odometry
Neural inertial odometry has demonstrated strong potential for motion estimation in challenging environments, yet inertial-only preintegration remains sensitive to IMU bias and uncertainty. To this end, this paper introduces PLATO:~Preintegration Learning from Accurate Trajectory Observations, a likelihood-based framework that leverages accurate trajectory observations to jointly learn IMU bias dynamics modeled by a neural ordinary differential equation~(NODE) and gyroscope and accelerometer noise covariances. Optimization exploits the sparse structure of the negative log-likelihood, with IMU noise-parameter gradients computed by forward differentiation. A tailored double-adjoint scheme couples a discrete invariant-error adjoint with a continuous-time adjoint for the bias NODE, enabling memory-efficient likelihood optimization over the nested bias-dynamics and IMU-preintegration rollouts. Validation on EuRoC shows improved performance, and underwater robot experiments demonstrate applicability under intermittent lighting failures and visual degradation.
Read the full abstract
Data Structures and Algorithms
PACO: A Fully Cache-Oblivious Parallel FFT with One Global Redistribution
Fast Fourier transforms on parallel machines incur two forms of data movement: processor-local transfers through the memory hierarchy and global redistribution between processors. Four-step FFT organizations switch the active transform dimension with one global transpose-like exchange, but this alone does not yield cache-efficient local computation. Cache-oblivious FFTs achieve asymptotically optimal local memory traffic through recursive layout transformations, whose direct materialization can require extra data-rearrangement passes and global exchanges. We present PACO, a fully cache-oblivious parallel FFT framework that reconciles these objectives. PACO executes LocalFFT -> OneGlobalPermutation -> LocalFFT. Its local stages recursively partition both transform and batch dimensions without knowledge of the cache parameters. Rather than materializing the transpose-like layouts induced by this recursion, PACO defers them, showing that they compose into a base-b digit-reversal permutation that is fused with the redistribution already required to change the local transform dimension. The resulting middle stage is a perfectly balanced parallel cache-oblivious digit-reversal permutation in which each source-destination processor pair exchanges exactly $N/p^2$ elements. Under an exact base-b slab decomposition in a hybrid ideal-cache/BSP model, PACO computes an N-point DFT exactly with maximum per-processor work $\Theta((N/p) \log N)$ and maximum per-processor cache complexity $\Theta((N/(pB))(1 + \log_M N))$, using exactly one global redistribution round. PACO returns canonical logical DFT coefficients under a factor-swapped target slab ownership. The single redistribution is necessary under the stated no-replication ownership model, while its communication volume is optimal for the prescribed fused permutation and source-target slab distributions.
Read the full abstract
Emerging Technologies
Phase-cycled randomized benchmarking of quantum processors: recovering hidden classical noise correlations
Randomized benchmarking can hide classical temporal correlations because its Clifford-twirled response is even in the noise phase. For a stationary symmetric telegraph fluctuator, we show that continuous evolution and independent stationary resets at slot boundaries yield identical mean responses for arbitrary fixed idle modulations. We construct an eight-setting phase-cycle measurement of the connected sine-phase covariance under ideal Clifford twirling and classical idle dephasing. This observable vanishes for independent slot noise and fixed detuning without a weak-phase or Gaussian approximation. A closed telegraph response, independent circuit calculations and 800 simulation trials validate the construction and quantify the empirical coverage of a paired bootstrap estimator. A separate conservative confidence set states its finite-sample assumptions. Two acquisitions on an IBM processor compare engineered shared-sign and independently reset phases with identical marginals. Their primary contrasts are 0.254 and 0.211, with empirical 95% intervals [0.177, 0.331] and [0.136, 0.285], respectively; all negative-control intervals include zero. At equal shot and sensingwindow budgets, an ideal Ramsey/echo estimator is more precise in every tested class. The result supplies an explicit connection between a benchmarking identifiability limitation and a controlled correlation measurement. No native or quantum-memory detection is claimed.
Read the full abstract
Operator Algebras
KS-Groupoids and KS-Crossed Products associated with Inverse Semigroup Actions
For an action of an inverse semigroup $S$ on a locally compact Hausdorff space $X$, we construct an étale groupoid $S\ltimes X^\mathrm{KS}$, called the KS-groupoid associated with the action. This construction may be viewed as a dynamical analogue of the construction of universal groupoids for inverse semigroups. We provide a dynamical version of Paterson's theorem which states that the full and reduced crossed products (in the sense of Khoshkam and Skandalis) of $C_0(X)$ are canonically isomorphic to the full and reduced groupoid C*-algebras of $S\ltimes X^\mathrm{KS}$, respectively. We further investigate some topological properties, functoriality and universality of KS-groupoids.
Read the full abstract
Chemical Physics
Tensor-based Approximation of Molecular Kinetics: Generator Learning, Reaction Coordinates and Incremental Updating
We present an approach to analyze long-timescale kinetics of molecular dynamics simulations - meta-stable states and transition timescales - by a tensor-based approximation of the infinitesimal generator. We start from the variational approximation of low-lying generator eigenvalues, and discretize the associated eigenvalue problem using a tensor-product basis. We derive analytical expressions for the resulting tensor operators in tensor train (TT) format, and provide efficient algorithms to solve the corresponding linear problems. We also treat the case of a state-dependent diffusion field, which is relevant when working in reaction coordinates. We show how the number of reaction coordinates can be gradually increased without having to solve the variational problem from scratch. Using MD simulations of fast-folding proteins and TICA coordinates as reaction coordinates, we demonstrate the effectiveness of the proposed method at identifying long-timescale transitions and meta-stable states.
Read the full abstract
Artificial Intelligence
Causal Attribution for Agentic Decisions: Estimators, Coupling, and a Traceability Specification
A provider of a high-risk AI system must keep records that make a decision traceable, and for agentic systems it has not been established what those records must contain for post-hoc causal attribution to be possible. We give the estimator framework and then the conditions under which it fails. We separate the marginal total effect that prior work measures from a common-random-numbers total effect that isolates a step's own contribution, add the natural direct effect under a pinned downstream, and check the estimators against hand derivations. Both estimands then fail, in the same direction. Under the marginal estimand a causally inert step has the identical total effect to the decisive one on every run of our planted chain, an algebraic identity and not a coincidence at one draw. Under common random numbers the decisive step returns exactly zero on the runs where the executing step flips, about one in ten, while its direct effect there is 0.25 and it demonstrably acts; an exact zero does not certify that a step did nothing, and we put that here rather than in the limitations. We derive the coupling that keeps the direct effect estimable once contexts diverge, with a closed form for its degradation, and show that the mediated share on which a natural ranking is built is not a share under suppression: where the direct and mediated paths oppose, it exceeds one and ranks a suppressed component above a pure mediator. We publish the discrepancy experiment's pre-registration rather than a result, because the live pipeline it requires was not available in the study window. We contribute the traceability specification such a filing would need, against a gap the Act's calendar opens: Article 86's right to an explanation has applied since 2 August 2026, while the Article 12 logging and Annex IV documentation that could evidence one were deferred to 2 December 2027 by Regulation (EU) 2026/1744.