Peer Revenue Publish or profit

Papers

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

All of arXiv

Saturday 3 October

Mesoscale and Nanoscale Physics

Electron-phonon control of obstructed and quantum spin Hall topology in an altermagnet

Kuntal Bhattacharyya, Bilal Tanatar and Saurabh Basu

We theoretically demonstrate control over the interplay between obstructed atomic insulating and first-order topological phases via electron-phonon coupling (EPC) in a $d$-wave altermagnet (AM) that hosts nonrelativistic spin-split bands with zero net magnetization. By embedding a $d_{x^{2}-y^{2}}$-wave AM in a quantum spin Hall (QSH) insulator, we show that an altermagnetic order beyond a critical strength reconstructs the parent QSH state of a Bernevig-Hughes-Zhang (BHZ)-type topological insulator (TI) into an obstructed atomic insulator (OAI), characterized by nontrivial bulk polarization with quantized fractional corner charges, accompanied by characteristic in-gap boundary states, while the spin Chern number vanishes. Upon incorporating sufficiently strong EPC, the polaronic dressing reorganizes the Wannier centers and the associated effective edge topology, thereby weakening the altermagnetism-induced OAI phase and restoring the helical QSH edge modes. Further, EPC eventually drives the system into a trivial insulator. By uncovering a controllable switching mechanism between topologically distinct insulating phases in AMs, our study establishes a direct route toward phonon-tuned spintronic functionalities in unconventional magnetic platforms.
Read the full abstract

Computation and Language

Steering Speech-Language Models: Training-Free Task Specialization via Contrastive Activation Addition

Séverin Baroudi, Yanis Labrak, Pierfrancesco Melucci et al. (+5 more)

Activation steering has proven effective for controlling the behavior of Large Language Models (LLMs) at inference time, but its application to SpeechLLMs remains new, and training-free steering approaches for such models are still largely unexplored. We propose a training-free Contrastive Activation Addition (CAA) protocol that derives steering vectors for common speech tasks (e.g. transcription) in SpeechLLMs from a small number of labeled utterances. We showcase that adding these vectors in the representation space, at inference time, enforces better the targeted speech task. We further show that, when combined with prompting, these vectors yield to consistent improvement over prompting alone on most evaluated tasks such as Automatic Speech Recognition (ASR) or Emotion Recognition (ER), and transfer to out-of-domain data. We additionally demonstrate the usefulness of script-normalization directions to enforce the target script of a specific language.
Read the full abstract

All authors Séverin Baroudi, Yanis Labrak, Pierfrancesco Melucci, Sergio Burdisso, Petr Motlicek, Hervé Bredin, Mirco Ravanelli, Ricard Marxer

Accelerator Physics

Design of the Electron-Beam Collimator for the Electron-Ion Collider

Andrii Natochii, Elke-Caroline Aschenauer, Jay Best et al. (+5 more)

The design and optimization of a beam collimator for the Electron Storage Ring (ESR) of the Electron-Ion Collider (EIC) are presented. The collimator must efficiently intercept stray beam particles while maintaining acceptable beam coupling impedance, surviving rapid accidental beam impacts, and withstanding steady-state beam-induced heating. An integrated design approach combining beam-loss tracking, electromagnetic-shower simulations, thermomechanical analysis, and impedance calculations is developed and benchmarked against published simulations and beam-impact experiments. A comparison of candidate absorber materials identifies Al2219-T62 as providing a favorable balance of beam-impact robustness, thermal and electrical conductivity, mechanical strength, and ultra-high-vacuum compatibility. For the ultimate ESR bunch charge of 28 nC at 10 GeV, the Al2219-T62 jaw remains below the adopted practical thermal-shock threshold for approximately 50 fully intercepted bunches and below the localized melting threshold for approximately 100 bunches under the reference impact conditions considered. These results provide an engineering input to the machine-protection requirements, for which a complete protection response within several machine turns is required. Optimization of the absorber-tip length yields a 70 mm tapered jaw that balances collimation performance, beam-impact robustness, and impedance. The resulting design maintains acceptable impedance margins and remains below both the yield and ultimate tensile strengths under a conservative 3.6 kW beam-induced heating scenario. The results establish a baseline ESR collimator design and a benchmarked methodology for evaluating collimator robustness in high-current electron storage rings.
Read the full abstract

All authors Andrii Natochii, Elke-Caroline Aschenauer, Jay Best, Alexei Blednykh, Xiaofeng Gu, Kuppan Gunavathi, Charles Hetzel, Christoph Montag

Robotics

RoboIRS: Inference-Time Internal Representation Steering for Generalist Robot Policies

Jiuzhou Lei, Chang Liu, Dayou Li et al. (+5 more)

Vision-language-action (VLA) and world-action models (WAMs) often degrade under out-of-distribution task variations despite retaining partial task capability. To recover such capability, we propose RoboIRS, an inference-time internal representation steering method that uses successful and failed rollouts to train linear classifiers, select outcome-relevant intervention locations, and derive task-specific steering directions without updating policy parameters. On 15 simulation tasks with a frozen $\pi 0.5$ policy, RoboIRS improves the average success rate from 44.4% to 66.2%, outperforming alternative inference-time intervention baselines while adding little inference time. We further validate RoboIRS on real-robot manipulation using the same $\pi 0.5$ policy and demonstrate its applicability to a world-action model Cosmos Policy, where the average success rate improves from 35.4% to 55.4%. These results show that directly steering internal robot-policy representations can improve the performance of robot policies at inference time. Project website is available at https://rollingoat.github.io/roboirs/.
Read the full abstract

All authors Jiuzhou Lei, Chang Liu, Dayou Li, Zhiyuan Zhang, Xiao Liang, Yu She, Zhiwen Fan, Minghui Zheng

Computer Vision

COMPASS: Comet Object Measurement Pipeline with Automated Selection and Scoring

Jack Roberts, Canya Lu, Alexis Michelle Lawson et al. (+4 more)

Summary: The single-cell gel electrophoresis ('comet') assay is a widely used technique for quantifying DNA damage at the individual cell level. However, image analysis often relies on manual inspection or semi-automated software, which can be labor-intensive, difficult to reproduce, and sensitive to image quality and comet morphology. COMPASS automates comet assay image analysis by combining deep learning-based comet segmentation with damage measurement and automated comet selection. The pipeline produces standardized DNA damage measurements while offering robust detection, reducing manual effort and improving reproducibility through transparent selection and optional manual review. Availability and implementation: COMPASS is implemented in Python and is freely available at https://github.com/ rsinghlab/COMPASS. Installation instructions, pretrained weights, and example usage are provided in the repository. Contact: jack_roberts2@brown.edu, ritsingh@illinois.edu Supplementary information: Available onlime upon publication.
Read the full abstract

All authors Jack Roberts, Canya Lu, Alexis Michelle Lawson, Kaitlyn Holden, Gerald S. Wilkinson, Anne M. Bronikowski, Ritambhara Singh

Differential Geometry

Analyticity and tautness of smooth Dupin submanifolds

Stephan Wiesendorf

We prove that a smooth immersed Euclidean submanifold is locally Nash if it satisfies the Dupin condition on the regular part of its unit normal bundle. This condition is equivalent to the Morse-Bott property of all squared distance functions and, for compact connected embedded submanifolds, to tautness over $\mathbb{Z}_2$. No local finiteness assumption on the regular locus, or regularity assumption on its complement, is required. We also prove that, in each fixed Euclidean space, every compact connected smooth taut submanifold can be joined to one of finitely many representatives by a Nash isotopy through taut embedded submanifolds. In each fixed intrinsic dimension, we also obtain only finitely many diffeomorphism types admitting taut embeddings in Euclidean spaces. We first treat hypersurfaces by combining a dimension-dependent degree bound on proper Dupin pieces with a continuation argument using finite jets and resultants. An analytic Hessian criterion then gives the Morse-Bott property, and small tubes extend the conclusions to arbitrary codimension. The degree bound also provides uniform semialgebraic complexity. Together with the semialgebraicity of tautness in parameter families and Nash triviality, this yields the finiteness and deformation results. Classical classification theorems give further topological restrictions.
Read the full abstract

Number Theory

Few small primes suffice to cover unit groups

Scott D. Hughes

Klurman, Shparlinski and Teräväinen showed that there is a set of at most $(\log Q)^{1+\varepsilon}$ primes of polylogarithmic size whose subset products cover the unit group $(\mathbb{Z}/q\mathbb{Z})^\times$ for almost all moduli $q\le Q$. We show that $(1/\log2+o(1))\log Q$ primes suffice, which is optimal to first order, since $m$ primes have at most $2^m$ subset products. The proof combines a doubling argument of Erdős–Rényi type for random subset products in a finite abelian group, modified to tolerate a small set of exceptional characters, with a one-time “repair” of the few primitive characters whose $L$-functions have zeros near $s=1$. An exact weighted repair gives a variant with at most $(2/\log2+o(1))\log Q$ primes and an exceptional set described explicitly.
Read the full abstract

Statistical Methodology

Estimation and inference in penalized function-on-function regression with dependent residual curves

Fabian Scheipl

Penalized function-on-function regression, as implemented in the R package refund, treats all grid values of all curves as independent. Residual curves are smooth, so values within a curve are dependent. We study what this does to point estimates and pointwise confidence intervals, in a simulation with Gaussian, count and binary responses and in a plasmode study built on nine real datasets. Under within-curve dependence, model-based 95% intervals cover 0.63–0.82. A curve-clustered sandwich covariance with a leverage adjustment for the penalized hat matrix (CL2), computed after the usual REML fit, brings coverage close to nominal in the simulation; in the plasmode study it is a few points short on eight datasets and 0.84–0.91 for the coefficient surface on the ninth. Satterthwaite degrees of freedom recover most of the remaining shortfall except on that dataset, also with only 20 curves, at the cost of wider intervals. REML, however, undersmooths the coefficient surface: its error is two to four times that of a fit whose smoothing parameters are chosen by curve-blocked neighbourhood cross-validation (NCV), and its CL2 intervals for the surface are wide. Intervals centred at the smoother NCV estimate are narrow but undercover, the familiar tension between estimation and pointwise inference in nonparametric regression. We recommend the NCV estimate to describe the shape of the surface and REML + CL2 with Satterthwaite degrees of freedom for inference. For a surface that is zero everywhere, under dependent errors, these intervals exclude zero at about 5% of the grid points, and the NCV estimate is five to seven times closer to zero than REML's. For fitted means, intercepts and scalar effects, REML + CL2 intervals are close to nominal, except for binary intercepts and scalar effects and for misregistered counts. We state which defaults of refund the evidence supports.
Read the full abstract

Computation and Language

Extracting Persona Subspaces Through Iterative Nullspace Projection For Modulation

Ananya Malik and Mai ElSherief

Large Language Models (LLMs) can adopt distinct personas to tune their semantics, expertise, and perspective to different users and tasks. Precise control over these traits is critical to ensure safety and reliability in model behavior. Existing methods like activation steering and prompt-based persona induction reduce a persona to a single dominant direction, missing the finer, nested traits that emerge only once that dominant signal is factored out. We introduce modulation as a setting where the persona context is already embedded in the content being manipulated, requiring control methods to amplify or suppress a trait already present rather than inject it from scratch. PaSS is an inference-time control paradigm that models personas as multi-dimensional subspaces in a model's latent space without supervised contrastive examples. The persona subspaces are extracted via iterative concept erasure and applied to modulate persona-guided generation without retraining. To extract this subspace, we use Iterative Nullspace Projections (INLP) to linearly and iteratively isolate persona-specific directions. We causally evaluate six personas against diverse tasks like MATH-500, TinyAlpaca, GSM8K, and IFEval, showing that discriminative, iterative subspace extraction captures diverse traits underlying a given persona, enabling stronger and larger modulation than single-direction additive methods, while maintaining content fidelity. We further study individual peeled directions within each subspace to uncover the distinct aspects of persona behavior they encode. Overall, we show that persona subspaces offer a controllable, interpretable, and generalizable framework for modulating LLM behavior without sacrificing task performance.
Read the full abstract

Combinatorics

Domination and Total Domination in Johnson graphs

María Gracia Cornet, Tanja Dravec and Pablo Torres

In this paper we study domination and total domination of Johnson graphs $J(n,k)$. We establish monotonicity results proving that both $\gamma(J(n,k))$ and $\gamma_t(J(n,k))$ are non-decreasing with respect to $n$. We compute exact values for $\gamma_t(J(n,k))$ in specific cases, showing that $\gamma_t(J(n,2))=\lceil\frac{2}{3}(n-1)\rceil$ for $n\geq 4$, and determining $\gamma_t(J(n,3))$ for $n\geq 6$, which depends quadratically on $n$.
Read the full abstract

Cryptography and Security

Indistinguishability Lifting for Keyed Oracles, Compressed Ideal Cipher, and More Applications

Ritam Bhaumik and Yu-Hsuan Huang

Cryptographic security proofs often involve an adversary interacting with a larger, keyed oracle that consists of (potentially exponentially) many independent instances of a smaller, base oracle. However, showing quantum indistinguishability between two such keyed oracles can be tricky, since a single query made by an adversary may involve a superposition that covers all instances of the base oracles simultaneously. In this paper, we establish a generic indistinguishability lifting theorem of the following form: if the two base oracles are indistinguishable under quantum queries, then their corresponding keyed oracles are too, up to an O(q^2) multiplicative loss in distinguishing advantage, where q is the number of queries made by the adversary. Our lifting theorem applies to both statistical and computational settings, and to oracles that are stateful as well. It is also optimal in that it matches the obvious Grover search attack for a certain (contrived) choice of oracles. As an immediate application, we extend Carolan's compressed permutation oracle to an efficiently implementable compressed ideal cipher, and use it to prove preimage resistance of the Davies-Meyer compression function in the quantum ideal cipher model. Thanks to our lifting theorem, the soundness of our compressed ideal cipher reduces to that of Carolan's oracle, and any further improvement on the latter would automatically carry over to the former. As our second application, we give a modular construction that doubles the message length of any quantum-secure strong pseudorandom permutation. Along the way, we show that an existing two-round tweakable Feistel construction is indistinguishable from a random permutation under quantum bidirectional queries. This is done via a dedicated polynomial-method argument, which may be of independent interest.
Read the full abstract

Classical Analysis and ODEs

On pointwise convergence of singular integrals of Christ-Journé type

Ankit Bhojak, Sangita Kundu and Saurabh Shrivastava

The primary objective of this paper is to establish the pointwise almost everywhere convergence of the singular integrals of Christ-Journé type. To this end, we prove the weak type $(1,1)$ estimate for the corresponding maximal operator defined as \[ T^*[a] f(x):= \sup_{r>0}\left|\int_{|x-y| > r} K(x-y) \int_0^1 a(tx + (1-t)y)\, dt f(y)\, dy\right|, \] where $K$ is a Calderón–Zygmund kernel and $a\in L^\infty$. This result resolves a problem posed by Seeger (Rev. Mat. Iberoam., 2014).
Read the full abstract

Artificial Intelligence

MASBench: Benchmarking LLM-based Multi-Agent Collaboration under Partial Observability

Qizhi Chu, Zekai Yu, Sijie Wen et al. (+5 more)

Large language models (LLMs) have progressively evolved into the core of autonomous agents. Building on this progress, LLM-based multi-agent systems (MAS) coordinate multiple agents into a synergistic team to accomplish complex tasks that exceed the capabilities of individual agents. The effectiveness of such systems depends not only on the agents themselves, but also on how collaboration mechanisms are designed and organized. Note that real-world collaboration is typically partially observable, where each agent can only access partial information about the environment due to physical or privacy-related constraints. However, many existing multi-agent benchmarks assume global observability, and leave limited support for systematically evaluating collaboration mechanisms. To bridge this gap, we introduce MASBench, a multi-agent collaboration benchmark designed under partially observable constraints. It is organized into three progressive task categories: Reasoning, Scheduling, and Game. Through this structure, we progressively evaluate three representative collaboration mechanisms: Protocol, Memory, and Routing. MASBench further provides deterministic evaluation metrics, including performance score, communication cost, and cost effectiveness, to characterize both collaboration outcomes and communication overhead. Experiments across diverse LLM backbones and mechanism configurations offer empirical guidance for effective MAS design. Code is available at: https://github.com/BUPT-GAMMA/MASBench
Read the full abstract

All authors Qizhi Chu, Zekai Yu, Sijie Wen, Yang Liu, Chen Qian, Cheng Yang, Chuan Shi, Zhiyuan Liu

Optimization and Control

Asking the Crowd the Right Question: Bias-Cancelling Weights for Federated Learning

Ilya Kuruzov, Dmitrii Vishovan, Kirill Novoselov et al. (+3 more)

A federated objective is a weighted sum of client risks, and the weights are almost always fixed in advance. We treat them instead as the only instrument of a wisdom-of-crowds mechanism: clients are noisy views of one truth, each seeing it through an independent distortion that is unbiased across the crowd. That the optimal weights are inversely proportional to the clients' error energies is classical; we begin at the question that answer presupposes, which energies belong there and whether a crowd can recover them from itself. Excess risk on the truth is of the exact order of the aggregate bias energy, so no optimizer can repair a bad weight vector; the truth itself is identifiable only up to a linear tilt, so subtracting estimated client biases provably reproduces uniform weighting. The expected per-client second moments, however, are exactly identified from the law of the crowd's disagreement by a well-conditioned linear inversion, a step random-effects meta-analysis cannot take because a source reports once; their realized counterparts are estimable up to an incoherence floor the algorithm can measure. This yields CROWD, which reads the disagreement off the optimization trajectory at no extra cost and matches a Bayesian minimax lower bound in the same constant: per instance as the horizon grows, and unconditionally as the prior becomes diffuse. For arbitrary distortions it stays competitive with the optimal weights, at a ratio governed by a geometric incoherence the algorithm can measure. On real scans split into sites with their own miscalibrated detectors it attains the oracle excess risk; on a companion federation that pulls bias and noise apart, weighting by noise variance is worse than not weighting at all, and CROWD is not.
Read the full abstract

All authors Ilya Kuruzov, Dmitrii Vishovan, Kirill Novoselov, Yuriy Dorn, Darina Dvinskikh, Alexander Gasnikov

Artificial Intelligence

A Tropical Geometry View of Forgetting: A Per-Unit Projector for Knowledge-Preserving Fine-Tuning

Yuyang Zhang, Xiaoyin Chen, Chunlin Ren et al. (+1 more)

Fine-tuning a language model on new text degrades what it already does. Replay-free projectors such as Adam-NSCL and GPM forbid one shared subspace of a layer's inputs in every row of the update. The tropical geometry of a ReLU layer shows why this is too coarse. In data space, the units' walls are tropical hypersurfaces whose cells are dual to the upper vertices of a zonotope; in weight space, each old token is a hyperplane, and the tokens cut out a polyhedron, the closure of the weights that keep every token on its side. An exact identity joins the two pictures: the squared change of the layer's output under any weight change splits into in-cell, open-to-closed and closed-to-open terms, and the first two live on the tokens each unit fires on. The identity names a gate-aware per-unit projector, and a budget-separation theorem prices exact protection: it costs a unit the rank of its own open tokens, while a shared subspace pays at least the rank of their union in every row. On OPT-1.3b, where 96% of (token, unit) pairs are closed, the projector forgets less than Adam-NSCL at all six matched budgets from 9 to 60 constrained directions per row, the gap widening from $1.1\times$ to $4.3\times$; with 1/5.5 of the directions it halves the forgetting of Adam-NSCL at GPM's energy threshold. On OPT-6.7b, it matches Adam-NSCL's forgetting at matched budget while learning more. As the theory predicts, the open/closed partition is the operative variable: open tokens beat random, sign-blind and anti-gate token sets on 18 of 18 seed-pairs. In pruning repair, every derivative-based local model of the output error at the dense weights is blind to pairs that open: the minimisers of the gate-weighted objective can leave the polyhedron, the objective's closed-form solution is 1.94 nats worse than no repair on OPT-1.3b, and a convex one-sided penalty bounds the escape.
Read the full abstract

All authors Yuyang Zhang, Xiaoyin Chen, Chunlin Ren, Qihuang Zhang

Solar and Stellar Astrophysics

Measured energy exchange in coronal hole solar wind from its solar origins to the heliosphere

Yeimy J. Rivera, Alin R. Paraschiv, Samuel T. Badman et al. (+33 more)

Coordinated ground- and space-based remote observations are used to provide a comprehensive, multi-perspective view of the solar atmosphere that we connect with in situ measurements in the heliosphere centered around the April 8 2024 Total Solar Eclipse observed across North America. These near-contemporaneous, multi-wavelength datasets are used to directly derive the plasma parameters and compute an energy budget of a solar wind stream at its low-coronal birthplace (at an altitude of 1.05Rsun). The energy budget is also independently computed at and beyond the Alfvén surface from the in situ observations of the same stream. The total wave energy flux dominates the low-coronal budget (11 W m-2 Alfvén, 7 to 23 W m-2 in compressive MHD modes), while enthalpy (13 W m-2) and work done against gravity from 1.0Rsun (2.5 W m-2) make up the remainder. The total coronal energy budget is consistent with the total measured in situ by Parker Solar Probe if and only if compressive energy flux is included and attributed to the fast magnetosonic mode. Nearly all this thermal and wave energy is converted to ion heating, acceleration and gravitational potential energy within the Alfvén surface. The stream evolution is well-described as a thermal-pressure driven wind with additional energy deposited by MHD fluctuations. The observed evolution of the Alfvénic component of these fluctuations follows a near-dissipation-free curve below the Alfvén surface while deviating from it more strongly above this critical surface.
Read the full abstract

All authors Yeimy J. Rivera, Alin R. Paraschiv, Samuel T. Badman, Alberto M. Vasquez, Federico A. Nuevo, Ryan J. French, Momchil Molnar, Roberto Susino, Zihao Yang, Marco Romoli, Richard A. Frazin, Michael L. Stevens, Stuart D. Bale, Deborah Baker, Jasper Halekas, Thomas A. Schad, Christopher J. Owen, Philippe Louarn, Jenna Samra, Giuliana de Toma, Daniela A. Lacatus, Giulio Del Zanna, Chad A. Madsen, Enrico Landi, Manolis K. Georgoulis, Amir Caspi, Éric Buchlin, Nour E. Raouafi, Luca Franci, Miho Janvier, Cooper Downs, Roberto Livi, Phyllis Whittlesey, Ali Rahmati, Davin Larson, Nikos Sioulas

Social and Information Networks

Exact computation of the network modularity with branch-and-bound

Emily Weng and Georg Hahn

We consider the computation of the network modularity, a common measure of the structure of networks or graphs that quantifies the division of a network into modules or clusters. The computation of the network modularity is an NP-hard problem, thus making its exact computation often infeasible in practice. This has resulted in the development of several heuristics in the literature. In this work, we consider the exact computation of the network modularity with the help of a branch-and-bound algorithm. Our algorithm is guaranteed to find the optimal clustering that maximizes the network modularity, however it achieves this without a full exploration of the search space. This is accomplished by bounding partial solutions and discarding them if it can be foreseen that a partial solution will not yield an improvement over an existing solution. We assess our algorithm with respect to accuracy and runtime and compare it to a brute-force approach as well as two benchmark algorithms, simulated annealing and the Louvain method.
Read the full abstract

Quantum Physics

Quantum-State Oblivious Transfer over Quantum Erasure Channels

Remi A. Chou

We study quantum-state oblivious transfer over the quantum erasure channel under information-theoretic security guarantees. Specifically, Alice holds two quantum systems, Bob wishes to recover one of them, and the parties have access to independent channel uses and authenticated public classical communication, with no initial shared entanglement. We consider both honest-but-curious and malicious behavior. We establish finite-blocklength converse and achievability bounds, which match in the honest-but-curious setting. The malicious setting construction combines interactive hashing with tests of whether Bob received selected channel outputs, and distributes each input across several qubits so that erasures conceal an entire input from a dishonest receiver, even when the inputs are entangled with an external reference. We also derive fixed-error asymptotic expansions and show that both malicious and honest-but-curious settings have the same capacity.
Read the full abstract

Quantum Physics

Correlations decide a shallow-circuit advantage

Zijian Gong, Zhaobin Lyu, Jingjing Hu et al. (+2 more)

Quantum computers are claimed to produce samples ordinary classical computers cannot reproduce. The sharpest such claim known pits constant-depth quantum circuits, given one entangled state per run, against shallow classical circuits of few-input gates and limited randomness. No efficient test was known that certifies such a claim from classical samples alone. Here we show that what such a test must read is decided by the classical class's pattern of single-bit and pair correlations, not by a distance. The test rests on one inequality: any machine producing the samples sits no further from the target than the fraction it mislabels plus the deviation of its string half, the bits beyond the label, from uniform. A collapse theorem, machine-checked in the Lean~4 proof assistant, confines that second term, one setting of the sampler's random input bits at a time. Four natural checks provably fail; a fifth, reading those same correlations, catches a far-from-target construction they cannot reach, and is exact on the class of few-input gates. The label test is proven and sample-optimal in its own tolerance up to a logarithmic factor; the fifth check is sound, not only effective, on the bounded pinned-residue class, pairwise-uniform samplers whose residue, the string's weight modulo the prime, is pinned by few seeds and whose seed sharing is bounded. What binds an experiment is the $43$-qubit entangled resource state at fidelity near $0.99$, not the readout. What remains are two named open problems; a positive answer to the first would extend the guarantee to the full class.
Read the full abstract

All authors Zijian Gong, Zhaobin Lyu, Jingjing Hu, Dengfeng Li, Shuoming An

Artificial Intelligence

Efficient Neural Surrogates for Linear Radiation Transport on the Lattice and Hohlraum benchmarks

Carmelo Gonzales, Steffen Schotthöfer and Cory D. Hauck

Linear radiation transport equations (RTEs) form the simulation foundations underpinning design and analysis tasks in nuclear engineering, inertial confinement fusion, medical imaging, and astrophysics, but resolving the high-dimensional phase space at engineering fidelity remains expensive enough that outer-loop workflows, such as design optimization, uncertainty quantification, and parameter sweeps, are routinely budget-bound on traditional solvers. Neural surrogates promise to relax this bottleneck by amortizing simulation cost across thousands of downstream queries, but the architectural choices and engineered inductive biases that make a surrogate accurate on one transport problem do not transfer straightforwardly across model families. We benchmark two parameter-matched neural surrogate architectures, the physics-attention Transolver and the multi-scale graph network Bi-Stride Multi-Scale MeshGraphNet (BSMS-MGN), as end-to-end approximations of the final-time particle concentration for the two-dimensional linear RTE on the canonical Lattice and Hohlraum benchmarks. An ablation across Fourier features and region-weighted training loss exposes strongly architecture-dependent inductive-bias preferences, indicating that design choices common to physics-informed surrogate workflows must be revisited per architecture rather than imported across model families, and that downstream utility depends on per-QoI sensitivity rather than a single field-level score. The model training recipe, training data, and evaluation pipeline are released alongside this paper to support reproduction, transfer to related transport problems, and evaluation as amortized forward-model components in larger outer-loop simulation workflows.
Read the full abstract

Computer Vision

FLASHSWIN: Unlocking Large Windows and Dense Tokens in Swin Vision Transformers with Memory Efficient Attention

Tushar Kataria, Gerald Sabin, Ponnuswamy Sadayappan et al. (+1 more)

High-resolution vision backbones have long been forced to trade away local token density to afford larger receptive fields. Hierarchical Swin transformers impose this compromise because standard windowed attention materializes an $M^2\times M^2$ score matrix per window, incurring $O(M^4)$ memory as windows or token grids grow. Furthermore, Swin adds a learned relative-position bias elementwise to attention scores, requiring full materialization of the score matrix and its gradient. This keeps Swin and SwinV2 trapped in a small-window($M=8,16$), coarse-token regime with patch size $4\times4$ ($p=4$), limiting performance for fine-grained tasks. We introduce FLASHSWIN, which replaces standard windowed attention with a FlashAttention implementation that computes exact softmax attention without materializing the score matrix, reducing per-window memory from $O(M^4)$ to $O(M^2)$. This enables higher token density and larger receptive fields without inflating memory overhead. Training memory is flat across window sizes: at a $32\times32$ window, FLASHSWIN-T requires only $12.4$\,GB, unchanged from $8\times8$, compared to $70/90$\,GB for SwinV2/V1-T. However, applying FlashAttention directly to Swin creates a trade-off: bypassing the score matrix precludes Swin's additive relative-position bias, forfeiting spatial information in exchange for memory efficiency. FLASHSWIN restores position information as window-local learnable 2D RoPE, making large windows and dense token grids both affordable and accurate. At matched scale, FLASHSWIN-T outperforms Swin variants. With dense tokens and wide windows ($p=2,M=32$), the same Tiny model reaches $84.1\%$ ImageNet-1K, $44.1$ COCO box AP, and $47.28$ ADE20K mIoU—gains of $+1.3$, $+5.1$, and $+1.82$ over SwinV2-T at $M=16$, respectively. At fixed $M=32$, halving the patch size yields roughly $3\times$ larger gains in boundary quality than in mIoU.
Read the full abstract

All authors Tushar Kataria, Gerald Sabin, Ponnuswamy Sadayappan, Shireen Y. Elhabian

Strongly Correlated Electrons

Does a pole in the self-energy imply a violation of the Luttinger theorem?

Artem Abanov and Andrey Chubukov

Luttinger theorem relates Fermi surface volume to electron density per spin $n$. It holds in an arbitrary Fermi liquid but is widely believed to break when a fermionic self-energy has a pole. We show that this is not necessarily the case. As an example, we consider a fermionic system consisting of two gapped bands, originating from a self-energy $\Sigma (\omega, \mathbf{k})$ with a pole, and a narrow residual quasiparticle band crossing $\omega =0$, with a small residue $Z \ll 1$ (a residual Fermi liquid). We consider Model I, in which $Z$ also renormalizes quasiparticle bandwidth $W$ to $ZW$ and Model II, in which $W$ remains unrenormalized. The Green's function for such a system $G(\omega, \mathbf{k})$ contains two zeros and three poles — one from the quasiparticle band and two from the gapped bands. Yet, we show that in Model I, the Luttinger-Ward analysis is fully applicable for any $Z >0$, and the results are exactly the same as if the gapped bands did not exist. For Model II, the Luttinger-Ward analysis recovers once $Z$ exceeds some critical value, which depends on fermionic density.
Read the full abstract

Systems and Control

Differentiable Ternary Temporal Logic Semantics with Polynomial Surrogate Networks

Ryan Matheu, Sai Sandeep Damera, John S. Baras et al. (+1 more)

Temporal logic provides designers a formal reasoning tool for specifying complex spatio-temporal behaviors and tasks for autonomous systems. Many techniques exist for control synthesis according to temporal logic specifications, spanning a large variety of systems, including robotics. However, the vast majority of implementations are limited to Boolean temporal logics. Boolean temporal logic does not have an innate mechanism for quantifying abstention, rather verdicts are either definitively True or False. Recent work demonstrates that ternary logic is a capable formalism for reasoning about robotic behaviors, with the natural ability to classify uncertainty with the Unknown literal. Ternary temporal logic is still in its infancy, with work limited to synthesis for linear systems and monitoring. This work proposes a probabilistic relaxation of ternary temporal logic that allows for value and gradient computation through the same network while retaining exact verdicts on actual data. The relaxed network can be used as an objective in gradient-based control synthesis for both offline and online robotic control applications. We demonstrate the utility of our framework with two robotic manipulator tasks, one offline and one online, involving intermediate checking and sequencing of subtasks to demonstrate the benefit of our approach in terms of correct and timely execution.
Read the full abstract

All authors Ryan Matheu, Sai Sandeep Damera, John S. Baras, Calin Belta

Cryptography and Security

Post-Quantum Authentication Protocol for Internet of Medical Things

Kartick Sutradhar and Ranjitha Venkatesh

IoMT communication systems should provide origin authentication and ensure integrity protection for the messages produced by patients' devices for transmission to gateways and further processing by other hospital services. With regard to IoMT, this paper discusses tampering, impersonation, and replay attacks on messages provided as medical reports. Long-term security of the communication system is also addressed considering potential quantum attacks. For this purpose, we design a protocol that meets requirements of practical implementation and employs an ISIS like lattice-based construction combined with concepts of the blockchain technology. Patients' devices generate a canonical JSON format medical report containing a timestamp and critical information. Using a nonce, the patients create a cryptographic signature of the message with a challenge generated from hash of a message. The gateway verifies signatures utilizing the public key of a trusted ledger associated with a particular patient and ensures freshness of the transmitted message. To optimize the performance in case when many devices interact with a gateway at once, the gateway performs batch signature validation using vectorized modular linear equations. The trusted ledger uses tamper-resistant REGISTER and TRUST UPDATE entries, maintaining a trust score of each registered patient and updating the value after each ACCEPT/REJECT operation. Our prototype application demonstrates registration procedure of a new patient, signing and verifying medical reports, auditing of the ledger, and performance assessment. Based on our experiments in the prototype, we argue that, when applied to large-scale systems, batch verification is superior to simple signature validation approach.
Read the full abstract

Combinatorics

The asymptotic maximum oriented diameter of graphs

Jiangdong Ai, Yaokun Feng, Hui Lei et al. (+1 more)

The oriented diameter of a connected bridgeless graph is the minimum diameter of a strong orientation. Let $f(d)$ be the maximum oriented diameter among all such graphs of diameter $d$. In 1978, Chvátal and Thomassen proved $f(d)\le2d^2+2d$ and constructed graphs showing that any quadratic upper bound on $f(d)$ must have leading coefficient at least $1/2$. We prove that $f(d)\le \tfrac12d^2+7d$ for every integer $d\ge1$. This matches the leading coefficient of their lower bound and establishes $f(d)=\tfrac12d^2+O(d)$, determining the optimal quadratic coefficient. Our proof gives a polynomial-time algorithm that constructs a strong orientation satisfying the stated bound.
Read the full abstract

All authors Jiangdong Ai, Yaokun Feng, Hui Lei, Zijian Ren

Artificial Intelligence

PermVLA: Factorization Order as a Regularizer for VLA Learning

Yanqiao Chen, Yuhan Rui, Dongsheng Hou et al. (+3 more)

Vision-language-action (VLA) policies commonly learn action chunks through a fixed left-to-right (LTR) factorization, although the same expert trajectory distribution admits many valid chain-rule factorizations. We identify factorization order as an overlooked regularization choice and introduce causally anchored permutation (CAP), which samples action reveal orders with a tunable chronological prefix. Its auxiliary objective trains one shared policy to predict actions from different known subsets of the same expert chunk, while deployment retains deterministic LTR control. We call this conditional-set augmentation: it creates multiple conditional prediction problems from one expert chunk without adding demonstrations. This discourages reliance on the single chronological prefix used by ordinary teacher forcing. Controlled experiments show that CAP consistently outperforms standard LTR training on LIBERO and LIBERO-Plus, with the same advantage appearing in cross-dataset CALVIN evaluation. A diagnostic that measures the expected squared difference between a chunk's joint log likelihood under two reveal orders verifies that CAP training internalizes agreement across reveal orders. These findings position sampled subset-conditioned auxiliary objectives as a general recipe for constructing VLA regularizers, illustrated by an extension to diffusion action generators.
Read the full abstract

All authors Yanqiao Chen, Yuhan Rui, Dongsheng Hou, Zijie Nie, Yutong Wan, Qi Hao

Software Engineering

RETRACE: From Entangled Repair Histories to Reusable Experience for CI Repair

Rabeya Khatun Muna, Muhammad Ahasanuzzaman, Nakhla Rafi et al. (+3 more)

Large language model (LLM) agents increasingly reuse prior experience, but most approaches assume that problems and solutions are already aligned. Software histories rarely provide this alignment: a pull request (PR) may contain multiple continuous integration (CI) problems, failed attempts, reverted edits, and unrelated changes, obscuring which changes resolve each problem. We present RETRACE, a framework for reconstructing problem-level repair experience from such histories. RETRACE combines an endpoint view that reasons backward from changes retained in the passing revision with a development view that traces repair evolution forward through commit history. CI execution evidence reconciles the two views, and the recovered experience is represented at three abstraction levels, from concrete fixes to transferable repair patterns. For new failures, RETRACE retrieves relevant problem-level experience to guide repair. On CI-REPAIR-BENCH, comprising 565 PR-level repairs from 101 repositories across 12 failure categories, RETRACE improves mini-SWE-agent Pass@1 from 19.6% to 31.9% with MiniMax-M2.5 and from 23.3% to 32.8% with DeepSeek-V4-Flash. On a matched subset, Codex improves from 15.5% to 27.5%. Combining both views consistently outperforms either alone, showing that recovering problem-change alignment enables historical CI repairs to serve as reusable repair experience.
Read the full abstract

All authors Rabeya Khatun Muna, Muhammad Ahasanuzzaman, Nakhla Rafi, Yisen Xu, Jinqiu Yang, Tse-Hsun Chen

Machine Learning

Path Laplacian Encodings for Directed Graphs

Lydia Mezrag, Semih Cantürk, Michael Perlmutter et al. (+2 more)

Directed graphs naturally model many real-world systems in which interactions are asymmetric, such as citation networks, web graphs, and information-flow networks. However, graph learning methods commonly rely on message passing with symmetrized graph representations or positional encodings that only partially exploit edge directionality. We introduce PathLapPE, a novel spectral positional encoding (PE) derived from the path Laplacian on directed graphs. PathLapPE provides node- and edge-level features that encode directional higher-order structure and can be incorporated into standard graph learning architectures. Empirical results on node- and graph-level benchmark tasks show that PathLapPE yields consistent improvements across several architectures, especially when combined with direction-aware message passing. Compared with magnetic Laplacian positional encodings, a widely studied spectral positional encoding for directed graphs, PathLapPE does not require additional fine-tuning of directionality hyperparameters while offering competitive runtime and performance.
Read the full abstract

All authors Lydia Mezrag, Semih Cantürk, Michael Perlmutter, Bastian Rieck, Guy Wolf

Neural and Evolutionary Computing

Robust Optimization of Spring Design under Variable Manufacturing Tolerances

Grzegorz Sroka and Slawomir T. Wierzchon

A reliable assessment of an optimizer robustness requires tracking controlled changes in parameters to identify sources of algorithmic weakness. Such an approach was applied to the problem of designing tension/compression springs with three continuous variables describing the spring geometry. Three additional independent binary parameters reflect material variability, manufacturing-related geometric deviations, and an additional safety margin. Without expanding the decision space, these parameters can be combined in eight ways to generate variants that allow for direct comparison. In this study we distinguish between robustness effects and the dimension of the problem, which allows tracking changes in feasibility while keeping the nominal mass objective constant. Seven optimization methods and eight hybridized their variants are used for the analysis. The hybridization mechanism applies the operator selected by the prediction procedure presented in Section 3. All methods are evaluated under the same experimental conditions. Each method is run thirty times for each variant, with a maximum of 30000 function evaluations permitted per run. The final quality assessment takes into account the geometry of the feasibility region, search trajectories, performance variations across landscapes and the computational cost of fixed objectives. Together, these factors allow us to determine the quality of the solution, reproducibility, feasibility control, and the effectiveness of optimization. An interesting observation was that hybridization can mitigate cost or feasibility losses associated with a specific configuration. While it compensates for a specific weakness in the underlying mechanism, it does not provide universal resilience.
Read the full abstract

Instrumentation and Methods for Astrophysics

Xenia: A Table-Driven Code for Astrophysical Multi-Fluid Hydrodynamics

Lei Xing

Models of escaping planetary atmospheres require neutral and ion species with distinct velocities and thermal pressures, coupled by photoionization, recombination, radiation, and collisions. Implementing and extending such coupled models is often hindered when species definitions and stiff interactions are embedded in solver-specific kernels. We present Xenia, a table-driven code for astrophysical multi-fluid hydrodynamics designed to lower this barrier. Species, coupling channels, chemistry, heating and cooling, and radiation channels are declared in Python, compiled to symbolic tapes, and passed through a versioned interface to shared C/Kokkos kernels; changing a problem therefore changes data rather than kernel code. For ionized configurations, electrons provide a pressure-only quasi-neutral closure: electron density follows from ion populations, electron mass and momentum are not advanced, and only electron pressure is transported. Stiff coupling is treated by operator splitting into a radial radiation scan, hyperbolic transport, and a cell-local implicit source solve. The same symbolic tapes yield an analytic finite-difference-free Jacobian for a direct dense LU solve. The public v0.1 release is validated with acoustic convergence, dustybox drag relaxation, conservation and closure contracts, two-fluid cloud-shock coupling, curvilinear geometry bookkeeping, three escape chains, and 394 passing tests. Conservation and positivity are not claimed to match mature production codes; radiation uses a radial Beer-Lambert prefix product rather than a general solver; escape chains are reproducibility checks rather than new science predictions; and backend evidence is limited to functional deployment and equivalence fingerprints. Xenia is released under GPLv2 at https://github.com/YFG-tec/xenia.
Read the full abstract