Peer Revenue Fully funded

Papers

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

All of arXiv

Tuesday 13 October

Instrumentation and Detectors

Cryogenic characterisation of the FBK NUV-HD-cryo Silicon Photomultipliers for the DarkSide-20k dark matter experiment

F. Acerbi, P. Adhikari, P. Agnes et al. (+299 more)

Liquid Argon Time Projection Chambers (TPCs) are highly promising detectors for dark matter searches due to their uniform response, scalability to large target masses, and compatibility with ultra-low background conditions. The DarkSide-20k (DS-20k) experiment, currently under construction at Istituto Nazionale di Fisica Nucleare (INFN) Laboratori Nazionali del Gran Sasso (LNGS), aims to achieve unprecedented sensitivity to Weakly Interacting Massive Particles (WIMPs), approaching the neutrino fog. At the heart of the detector is a dual-phase TPC filled and surrounded by low-radioactivity argon from underground sources (i.e. Underground Argon (UAr)), functioning both as the WIMP target and detection medium. The experiment employs Fondazione Bruno Kessler (FBK) NUV-HD-cryo Silicon Photomultipliers (SiPMs) as the chosen photon sensors for its optical planes. Key SiPM properties such as dark noise rate, gain and direct crosstalk were measured using specialized test setups at 77 K. The photon detection efficiency (PDE) was measured under vacuum at 77 K over the wavelength range 350-820 nm. Secondary photon emission, which can give rise to external optical crosstalk, was measured under vacuum at 86 K over the wavelength range 550-1020 nm.
Read the full abstract

All authors F. Acerbi, P. Adhikari, P. Agnes, I. Ahmad, S. Albergo, I. F. Albuquerque, T. Alexander, A. K. Alton, P. Amaudruz, G. A. Anastasi, M. Angiolilli, E. Aprile, D. J. Auty, M. Ave Pernas, O. Azzolini, H. O. Back, Z. Balmforth, A. Barrado Olmedo, P. Barrillon, G. Batignani, S. Bharat, P. Bhowmick, S. Blua, V. Bocci, W. Bonivento, B. Bottino, M. G. Boulay, T. Braun, A. Buchowicz, J. Busto, M. Cadeddu, M. Cadoni, R. Calabrese, V. I. Calì, A. Caminata, N. Canci, M. Caravati, M. Cardenas-Montes, N. Cargioli, M. Carlini, L. Castellani, P. Castello, D. F. F. S. Cavalcante, P. Cavalcante, S. Cebrian, A. Chepurnov, L. Cifarelli, Y. Coadou, I. Coarasa, V. Cocco, E. Conde Vilda, L. Consiglio, H. Coombes, A. F. V. Cortez, B. S. Costa, P. A. Costa e Silva, M. Czubak, S. D'Auria, M. D. Da Rocha Rolo, A. Dainty, G. Darbo, S. Davini, R. de Asmundis, S. De Cecco, M. De Napoli, G. Dellacasa, A. Derbin, L. Di Noto, P. Di Stefano, D. Díaz Mairena, C. Dionisi, G. Dolganov, F. Dordei, A. Elersich, E. Ellingwood, T. Erjavec, N. Fearon, M. Fernandez Diaz, L. Ferro, A. Ficorella, G. Fiorillo, D. Fleming, L. Flores, P. Franchini, D. Franco, H. Frandini Gatti, F. Gabriele, D. Gahan, C. Galbiati, G. Galiński, G. Gallina, M. Garbini, P. Garcia Abia, A. Gawdzik, G. K. Giovanetti, A. Gola, L. Grandi, G. Grauso, G. Grilli di Cortona, A. Grobov, M. Gromov, J. Guerrero Cánovas, M. Gulino, S. B. Habtemariam, B. R. Hackett, A. Hallin, M. Haranczyk, T. Hessel, C. Hidalgo, J. Hollingham, S. Horikawa, J. Hu, F. Hubaut, D. Huff, T. Hugues, A. Ianni, A. Ianni, V. Ippolito, A. Jamil, C. Jillings, P. Kachru, R. Keloth, N. Kemmerich, A. Kemp, K. Kondo, G. Korga, L. Kotsiopoulou, S. Koulosousas, P. Kunze, M. Kuss, M. Kuźniak, M. Kuzwa, M. I. Kwon, M. La Commara, M. Lai, E. Le Guirriec, E. Leason, A. Leoni, L. Lidey, J. D. Lipp, L. Lippolis, M. Lissia, L. Luzzi, O. Lychagina, O. Macfadyen, J. Machts, I. Machulin, S. Mameli, S. Manecki, I. Manthos, A. Marasciulli, S. M. Mari, C. Mariani, J. Maricic, M. Martinez, G. Matteucci, K. Mavrokoridis, A. Mazzi, A. B. McDonald, L. Meng, A. Messina, R. Milincic, G. Miller, S. Minutoli, A. Mitra, J. Monroe, M. Morrocchi, A. Morsy, V. Muratova, M. Murra, C. Muscas, P. Musico, R. Nania, M. Nessi, G. Nieradka, K. Nikolopoulos, E. Nikoloudaki, J. Nowak, K. Olchanski, A. Oleinik, P. Organtini, A. Ortiz de Solórzano, A. Padmanabhan, M. Pallavicini, L. Pandola, E. Pantic, E. Paoloni, D. Papi, L. Parellada Monreal, B. Park, G. Pastuszak, G. Paternoster, R. Pavarani, A. Peck, P. A. Pegoraro, K. Pelczar, R. Perez, V. Pesudo, S. Piacentini, N. Pino, G. Plante, A. Pocar, S. Pordes, P. Pralavorio, E. Preosti, D. Price, G. Prior, M. Pronesti, S. Puglia, M. Queiroga Bazetto, F. Raffaelli, F. Ragusa, Y. Ramachers, A. Ramirez, S. Ravinthiran, K. Raymond, M. Razeti, A. L. Renshaw, A. Repond, M. Rescigno, S. Resconi, F. Retiere, A. Ritchie-Yates, A. Rivetti, A. Roberts, C. Roberts, D. Rodríguez Rodas, G. Rogers, L. Romero, M. Rossi, D. Rudik, J. Runge, M. Sabia, C. Salerno, P. Salomone, O. Samoylov, S. Sanfilippo, D. Santone, R. Santorelli, E. M. Santos, I. Sargeant, M. L. Sarsa, C. Savarese, E. Scapparone, D. Semenov, C. Seoane, A. Serrecchia, M. Sestu, V. Shalamova, S. Sharma Poudel, A. Sheshukov, M. Simeone, D. Sitzia, P. Skensved, M. Skorokhvatov, O. Smirnov, T. Smirnova, B. Smith, F. Spadoni, M. Spangenberg, A. Steri, M. R. Stock, V. Stornelli, S. Stracka, A. Sung, C. Sunny, Y. Suvorov, A. M. Szelc, O. Taborda, B. Tam, R. Tartaglia, A. Taylor, J. Taylor, G. Testera, A. Thompson, S. Torres-Lara, A. Tricomi, S. Tullio, E. Unzhakov, M. Van Uffelen, P. Ventura, G. Vera Díaz, S. Viel, A. Vishneva, R. B. Vogelaar, J. Vossebeld, B. Vyas, M. Wada, M. Walczak, Y. Wang, S. Westerdale, L. Williams, M. M. Wojcik, M. Wojcik, Z. Xu, J. Yin, A. Zabihi, P. Zakhary, A. Zani, H. Zhan, Y. Zhang, G. Zuzel

Functional Analysis

Isometric dilations and Beurling quotient modules

B. Krishna Das and Shubham Rastogi

In \cite{Dan-Guo}, the authors provided a characterization of all sequences of doubly commuting pure contractions that have the sequence $\textbf{M}_z,$ the sequence of the multiplication by the coordinate functions on the vector-valued Hardy space $H^2_\mathcal E(\mathbb {D}_2^{(\infty)})$ over the Hilbert's multidisk $\mathbb D_2^{(\infty)},$ as their minimal isometric dilation for some Hilbert space $\mathcal E$. In this article, we remove the hypothesis of doubly commutativity and characterize all sequences of commuting contractions that are unitarily equivalent to $(P_{\mathcal Q}M_{z_i}|_{\mathcal Q})_{i\geq 1}$ for some quotient module $\mathcal Q$ of $H^2_\mathcal E(\mathbb D_2^{(\infty)})$. We further characterize those sequences for which $\mathcal Q$ is a Beurling quotient module. This generalizes Theorem 3.2 from \cite{BKD} to a sequence of commuting contractions.
Read the full abstract

Algebraic Geometry

Frobenius liftability of isolated complete intersections

Tatsuro Kawakami and Teppei Takamatsu

We classify Frobenius liftable normal isolated complete intersection singularities of dimension at least three over an algebraically closed field of positive characteristic. In dimension three, they are precisely the regular points and the ordinary double points; in higher dimensions, they are regular. We also characterize Frobenius liftability of the complete local hypersurfaces defined by $xy-g(z,t)$ without assuming that the singularity is isolated. Such a hypersurface is Frobenius liftable if and only if, after a formal change of coordinates, $g$ is a unit times $H^pz^at^b$ with $0\le a,b<p$.
Read the full abstract

Materials Science

MINFF: A Family of Classical Force Fields for Minerals with Tunable Angle Terms

Michael Holmboe, Nathan Lavauzelle and Rogers E. Swai

Atomistic simulations of minerals are limited by the accuracy and transferability of the force field. We introduce MINFF, a force field family for layered and non-layered oxides and hydroxides of M being Si, Al, Fe, Mg, Ca, Ti, Li and F. MINFF has no M-O bond terms. Instead M-O distances follow from electrostatics and dispersion, while a single harmonic angle term on all M-O triplets controls coordination geometry. Metal charges are DDEC6 charges from reference mineral structures, and oxygen charges are assigned by nearest-neighbour charge smearing, covering basal, apical, hydroxyl and edge sites and sites adjacent to isomorphous substitutions. The oxygen Lennard-Jones parameters are based on the OPC3 water model and change by less than 5% when re-optimised per mineral, so compatibility with OPC3 and related ion-pair potentials is measured rather than assumed. Metal Lennard-Jones parameters were fitted to unit-cell metrics, bond distances and angles, both synchronously across more than 30 reference minerals (general, GMINFF) and for individual minerals (tailored, TMINFF), each at four angle force constants (0, 250, 500 and 1500 kJ/mol/rad^2). With angle terms, MINFF reproduces experimental unit-cells metrics to within 0.5% and atom-type distances and angles to within 0.7-1.4% and 0.1-0.4%, with Global Instability Index values as low as 0.11, below those of CLAYFF (0.34) and INTERFACE (0.21). A moderate angle constant reproduces the stiff in-plane and soft shear elastic constants of phyllosilicates and improves the Kaolinite librational bands, and in hydrated Na-Montmorillonite MINFF gives the lowest residual factor against the 2W 00l pattern of the models tested. The general and tailored sets at four angle stiffnesses give direct control over the transferability-fidelity trade-off. Topologies can be generated with the atom (MATLAB) and atomipy (Python) libraries and at www.atomipy.io.
Read the full abstract

2610.18220 · 13 Oct · 63 pages, 7 figures and 3 tables in the main text. Supporting information (included in the same file) has 19 pages with 9 figures and 5 tables. Parameters, reference structures, misc. scripts and topology generators available via https://github.com/mholmboe/minff and https://www.atomipy.io

Abstract PDF $CELL ×2$TOPO ×1

Differential Geometry

T^2-invariant Einstein metrics on closed 4-manifolds

Kyle Broder and Ramiro A. Lafuente

We construct $T^2$-invariant Einstein metrics with positive Ricci curvature on #_m(S^2 x S^2) for every integer $m>1$. These provide infinite family of Einstein metrics on closed $4$-manifolds with $\lambda>0$, which are in addition not Hermitian. We further establish the uniqueness of the $T^2$-invariant Einstein metrics on $S^4$, CP^2, S^2 x S^2, and CP^2#\bar{CP}^2.
Read the full abstract

Quantum Physics

Methods for Automated Equivalence Checking of Quantum Error-Correction Codes

Maxie Bichmann, Ludwig Schmid and Robert Wille

Quantum error-correction code and circuit construction is an active field of research, motivating automated methods for identifying equivalent designs. We develop and evaluate automated strategies for deciding whether two descriptions of quantum error-correction codes are equivalent under specified transformations. We formalize permutation equivalence for stabilizer and CSS codes and local-Clifford equivalence for stabilizer codes within a unified framework that separates physical from representation freedom. We develop three decision strategies (group-action search, isomorphism-based methods, and constraint solving), together with invariants and signatures for preprocessing. The empirical evaluation shows that no paradigm dominates across equivalence notions, code families, and parameter ranges. We therefore combine their complementary strengths in input-dependent hybrid checkers following a staged Refute-Refine-Decide architecture. Mean hybrid runtimes remain below 10 s for most evaluated structured-code cases with $n \leq 37$; for inequivalent code instances with $n=144$, they are about 310 s for permutation and 16 s for local-Clifford equivalence.
Read the full abstract

Performance

Fast and Flexible Spiking Network Simulations on Massively Distributed On-Chip Memory Accelerators

Catherine Mia Schöfmann, Jan Finkbeiner, Jan Vogelsang et al. (+1 more)

Studying brain function and dysfunction in spiking networks with realistic synaptic densities pushes simulation technology beyond its current limits. It requires simulations substantially faster than biological real-time to cover long model times and enable efficient iterative exploration, particularly when connectivity is plastic. The modelling community aspires to a $100\times$ speedup. Yet on conventional many-core systems this remains elusive, bounded by inter-node communication and memory access. So far, significant speedup has only been shown on a custom-built FPGA system. We demonstrate the potential of commercial systems with large distributed on-chip memory to accelerate brain research. To exploit the local SRAM, we introduce a group-based mapping and communication strategy tackling spike-routing challenges on massively distributed systems. On Graphcore IPUs, the cortical microcircuit runs at a record $22.7\times$ faster than biological real-time with negligible build times and a reusable compiled graph. By design, the algorithm makes no model- or hardware-specific assumptions, providing a general strategy for advancing neuronal simulation technology.
Read the full abstract

All authors Catherine Mia Schöfmann, Jan Finkbeiner, Jan Vogelsang, Susanne Kunkel

Computation and Language

TextualRL: Textual Reinforcement Learning for Context Management

Tianyu Huai, Jie Zhou, Bo Mao et al. (+3 more)

LLM agents can adapt to task feedback by revising the textual context that guides their behavior, without changing their model parameters. However, turning execution rewards into reusable context updates is challenging: individual trajectories offer limited comparative evidence about what to change, while task-specific fixes may overgeneralize when incorporated into a shared context. We introduce TextualRL, a gradient-free reinforcement learning framework that treats a group of rollouts from the same task under the same context, rather than an isolated trajectory, as the unit of policy critique. Groupwise Policy Critique contrasts successful and failed rollouts within a task; when outcomes are uniform, it draws on cross-task evidence to identify behaviors worth preserving or formulate repair hypotheses. Cross-Group Policy Update reviews the resulting proposals against evidence pooled across groups, including observations that yield no proposed edit, to refine their content and applicability. Reviewed proposals form a bounded textual patch, with adoption of the candidate context determined by held-out validation rewards. Together, these components establish a reward-driven policy optimization loop in which multiple attempts jointly inform reusable context updates. We evaluate TextualRL with five target agents on six benchmarks. Averaged over the six benchmarks, TextualRL improves test accuracy over the base model by 24.4 and 15.3 percentage points for GPT-5.5 and Qwen3.8-27B, respectively.
Read the full abstract

All authors Tianyu Huai, Jie Zhou, Bo Mao, Qin Chen, Liang He, Xipeng Qiu

Data Structures and Algorithms

No Truly Sublogarithmic-Round Algorithm for Distributed MIS in the LOCAL Model

Seri Khoury and Aaron Schild

We prove that $o(\frac{\log n}{\log\log n})$-round randomized LOCAL algorithms find a maximal independent set (MIS) with subconstant probability on some input. This shows that the 40-year-old $O(\log n)$-round randomized CONGEST algorithm of Luby and Alon, Babai, and Itai is nearly optimal. Our lower bound improves upon the $\Omega(\sqrt{\log n})$-round lower bound of Khoury and Schild and nearly closes the 20+ year-old gap between the $O(\log n)$-round algorithm and the $\Omega(\sqrt{\frac{\log n}{\log\log n}})$-round lower bound of Kuhn, Moscibroda, and Wattenhofer. To close this gap, one must show a lower bound for a graph that is not locally treelike. We do this by starting with a $D$-regular, $\Theta(\log_D n)$-girth graph, with $D = \Theta(\log n)$. Crucially, this is not our final graph. We obtain our hard instance by replacing each vertex of this treelike graph with a clique on $n^c$ vertices for some constant $c \in (0,1)$. We then replace each edge with a bipartite Erdős–Rényi graph between the corresponding cliques, with each edge present independently with probability $1/2$. We analyze arbitrary LOCAL algorithms on our non-treelike hard instance with a novel round elimination approach. We describe algorithms that output a set of vertices in this graph using two metrics: their occupancy probability and their conflict probability. Our round elimination framework takes an $r$-round algorithm and produces an $(r-1)$-round algorithm where occupancy and conflict probabilities collectively slightly degrade, but individually can significantly improve or degrade. This round elimination result is obtained by exploiting the lack of large biclique complements in dense bipartite Erdős–Rényi graphs, which in turn allows us to bound the entropy of the output distribution for any independent set algorithm.
Read the full abstract

Number Theory

A certified classical-strip comparison for the de Bruijn-Newman constant

Michel Planat

We give a directed comparison beginning with the classical zero strip \(\lvert\Ima z\rvert\le1\) for the de Bruijn–Newman heat deformation. Under explicitly stated complete model-to-function source, localisation and global contact premises, translated-probe estimates imply \[ \Lambda\le\frac{148611789321711}{10^{15}} =0.148611789321711<0.148612. \] The construction uses Raymond's finite source data and 57 fixed translated kernels, independently checked by rational Bernstein certificates. An explicit finite Weyl estimate improves the positive first-head tail and its centred logarithmic moment. A centred cone transfer retains the original physical error allowances, exceptional debit and additive cone reserve, and charges the change in existing phase-surplus credits. Directed tail enclosures are regenerated at 192 and 256 bits; a separate exact-rational auditor accepts all 19,023 comparison steps. The controlled decrease from the complete two-source comparison is \(0.000021760452379\). We also prove a compatible matched logarithmic tail transfer and identify the finite requirements for stronger exponent pairs. A supplementary section records Raymond's ceiling below \(0.116398\) under the distinct \(7/8\) zero-free hypothesis.
Read the full abstract

Commutative Algebra

Combinatorial bounds for the v-number of binomial edge ideals

Emiliano Liwski

We study the local $\mathrm v$-numbers of binomial edge ideals from a combinatorial perspective. For each minimal prime $P_{\mathcal S}$ of $J_G$, indexed by a cutset $\mathcal S$ of $G$, we consider two combinatorial invariants, $\mathrm v^-_{\mathcal S}(G)$ and $\mathrm v^+_{\mathcal S}(G)$, defined in terms of transversals of suitable hypergraphs, and show that the local $\mathrm v$-number $\mathrm v_{P_{\mathcal S}}(J_G)$ satisfies $\mathrm v^-_{\mathcal S}(G)\leq \mathrm v_{P_{\mathcal S}}(J_G)\leq \mathrm v^+_{\mathcal S}(G)$. Under a suitable hypothesis on the connected components, we provide explicit graph-theoretic descriptions of these invariants and analyze when they coincide, giving a combinatorial way to determine the local $\mathrm v$-numbers. As an application, we derive a sufficient condition for this coincidence, which applies to a broad class of graphs. Furthermore, we identify a constant that controls the gap between $\mathrm v^-_{\mathcal S}(G)$ and $\mathrm v^+_{\mathcal S}(G)$, thereby showing that the local $\mathrm{v}$-number lies within a bounded interval. Applications to cycles and trees illustrate both the equality criterion and the resulting bounds.
Read the full abstract

Nuclear Theory

Exact solution of adiabatic time dependent Hartree-Fock-Bogoliubov inertia for fission

C.V. Nithish Kumar, Gabriel Martínez-Pinedo, Samuel A. Giuliani et al. (+1 more)

An efficient method to exactly compute the adiabatic time-dependent Hartree-Fock-Bogoliubov collective vibrational inertias typical of fission is proposed. The method uses the quasiparticle finite amplitude method to invert the stability matrix of linear response. A comparison with other approaches, such as the exact generator coordinate method inertia and the associated cranking non-perturbative approximation, is conducted. It is found that the behaviour as a function of the collective variable (mass quadrupole moment) is essentially the same in the four cases considered, and substantially differs from the commonly used perturbative approach. It is also found that the magnitudes of the inertias are to a good extent related by single factors that seem to be connected with the effective mass of the energy density functional used.
Read the full abstract

All authors C.V. Nithish Kumar, Gabriel Martínez-Pinedo, Samuel A. Giuliani, Luis M. Robledo

Strongly Correlated Electrons

Unexpected stabilization of a single-$q$ charge density wave in pristine 1T-TiSe$_2$

Kimoon Higashihira Han, Dongju Hwang, Dongsin Kim et al. (+16 more)

Over the past decade, studies revisiting the classical charge-density-wave system 1T-TiSe$_2$, long understood to host the $2 \times 2$ triple-$q$ order with equivalent subcomponents, have suggested that these subcomponents may not be identical, challenging a 50-year-old paradigm. Here, we demonstrate not only that these subcomponents are non-identical but also that they develop sequentially rather than simultaneously. By tracking folded-band features, angle-resolved photoemission spectroscopy reveals that a unidirectional $2 \times 1$ single-$q$ order stabilizes first, while the other two components emerge sequentially at lower temperatures, completing the triple-$q$ phase. Our free-energy analysis explains this hierarchy through repulsive coupling among the three subcomponents, as supported by first-principles calculations incorporating phonon anharmonicity. Consequently, the three subcomponents are inequivalent in both phase and amplitude. These results reconcile reported symmetry breaking and differing characteristic temperatures and establish a framework in which ostensibly simple, high-symmetry charge-density order emerges from temperature-dependent competition and cooperation among inequivalent subcomponents.
Read the full abstract

All authors Kimoon Higashihira Han, Dongju Hwang, Dongsin Kim, Yeojin Ahn, Jaehun Cha, Seonggeon Gim, Gyubin Lee, Mingi Jho, Chan-young Lim, Sae Hee Ryu, Chris Jozwiak, Aaron Bostwick, Eli Rotenberg, Jonathan D. Denlinger, Makoto Hashimoto, Donghui Lu, Young Woo Choi, Gil Young Cho, Yeongkwan Kim

Machine Learning

Operator Surrogates

Christoph Schwab, Josephine Westermann and Jakob Zech

This chapter provides a mathematical introduction to operator surrogates, which are finite-parametric maps approximating complex (typically nonlinear) operators between infinite-dimensional (function) spaces. A typical approach to constructing operator surrogates is the encoder-approximator-decoder paradigm, whereby infinite-dimensional inputs are mapped to a finite-dimensional latent space, processed, and then mapped back to the infinite-dimensional output space. We discuss various methods for creating efficient representation systems, including stable linear frameworks like Banach frames and data-driven techniques such as Principal Component Analysis (PCA), Active Subspaces, and nonlinear autoencoders. Subsequently, we survey key surrogate architectures, contrasting spectral methods, based on polynomial chaos expansions and related technologies, with more recent neural network approaches. We further discuss architectures that do not rely on representation systems, such as Fourier Neural Operators and transformers. We elaborate on their relationship and how they relate to other neural network operator surrogates, particularly those obtained by so-called "algorithm-unrolling". Additionally, we provide an introduction to the theoretical analysis of approximation and generalization errors. We explain how underlying low-dimensional structures, provided for example by holomorphy, are essential to overcome the so-called "curse of dimensionality": this refers to the fact that the complexity of numerical algorithms tends to increase exponentially in the problem dimensionality. We present convergence rate bounds for holomorphic operators that depend on the number of trainable parameters and the regularity of the input and output data. Throughout, we prioritize intuition over detailed technical proofs and provide precise pointers to the literature for detailed arguments.
Read the full abstract

Computer Vision

Closing the Estimation Gap: Training-Free Transductive Inference for Cross-Domain Few-Shot Action Recognition

Changyi Li, Zeyu Yang and Yu Xiao

Cross-domain few-shot action recognition (CD-FSAR) aims to recognise unseen action classes in a target domain from a handful of labelled videos. Despite advances in representation learning, CD-FSAR inference still relies on prototypes estimated from extremely limited support examples. To assess this bottleneck, we perform an oracle analysis with the representation fixed. Constructing oracle prototypes and decision boundaries from ground-truth query labels improves 5-way 1-shot accuracy by 7–20 percentage points across five target datasets. This reveals a substantial estimation gap: the representation contains useful discriminative information, but prototypes and boundaries estimated from scarce supports fail to capture each episode's true class structure. Motivated by this, we propose TRACE, a training-free transductive inference framework that improves episode-level estimation without modifying the backbone. TRACE comprises two complementary stages. Confidence-Thresholded Transductive Prototype Refinement (CTPR) refines support-based prototypes using confidently predicted queries. The refined prototypes initialise a lightweight per-episode classifier, which Transductive Classifier Refinement (TCR) further adapts by encouraging confident, class-balanced predictions over all queries. Together, the two stages improve prototype and boundary estimation, enabling better use of information in the unlabelled queries. Across five target datasets (HMDB51, UCF101, SSv2, Diving48, and RareAct) in both 1-shot and 5-shot settings, TRACE consistently improves a strong CD-FSAR baseline, including gains of more than six percentage points on HMDB51 1-shot, while recovering much of the oracle-estimated gap. These results suggest that improving test-time estimation is a powerful, broadly applicable complement to representation learning for CD-FSAR under extreme support scarcity.
Read the full abstract

High Energy Astrophysical Phenomena

Fast Radio Bursts from Monster Shocks and Flares of Inspiraling Binary Neutron Stars: A Kinetic Model for Double Pulsar

Jan Benáček, Benoît Cerutti, Martin Pohl et al. (+2 more)

The coalescence of binary neutron star systems offers unprecedented possibilities for multimessenger observations, including the detection of gravitational waves, gamma-ray bursts, kilonovae, and potentially fast radio bursts. For a detailed investigation of the latter, one requires a kinetic plasma model to describe the interacting magnetospheres of the stars. We present a three-dimensional particle-in-cell kinetic simulation of the combined magnetosphere of a binary neutron star in which the two stars physically orbit one another so that the orbital motion itself twists the magnetic loops. We find that the stellar magnetic-field intensity ratio controls the emergence of flares, while the stellar spins set the direction and modulation of the flare. We identify four spatially and dynamically distinct regions of strong electromagnetic energy release: (i) an expanding fast magnetosonic wave that steepens into a monster shock, (ii) a plasma flare being pushed by (iii) an expanding fireball and containing (iv) a reconnecting current sheet. The fast magnetosonic wave is a promising source of coherent radio emission. It is released when orbital motion cuts a closed plasma loop. The electric and magnetic wave amplitudes are similar and reach that of the background magnetic field. Our kinetic treatment yields particle spectra and plasma structures that can be used to assess coherent emission in kinetic-scale simulations. In conclusion, the stellar orbital motion constitutes a driving channel distinct from models with spin-driven magnetospheric twisting.
Read the full abstract

All authors Jan Benáček, Benoît Cerutti, Martin Pohl, Tim Dietrich, Anna Neuweiler

Numerical Analysis

Hyperelastic Modeling of Cardiac Tissue Deformation for Data Registration

Joyce Ghantous, Yves Coudiere and Michael Leguèbe

This work proposes a biomechanical registration framework designed to combine functional optical-mapping data with structural high-resolution MRI of cardiac tissue. The two acquisitions are performed in different mechanical configurations: optical mapping is obtained after stretching the sample, whereas MRI is acquired in a fixed resting state. To estimate the deformation between these configurations, the tissue is modeled as a nearly incompressible transversely isotropic hyperelastic material, and the traction amplitude is identified from optical-map marker positions through an inverse problem. Two-and threedimensional simulations show physically admissible deformations and provide a first proof of concept for the proposed strategy.
Read the full abstract

Software Engineering

SWE-MISBEHAVIOR: Benchmarking and Mitigating Boundary Violations in Software Coding Agents

Do Xuan Trong, Duc Anh Nguyen, Do Xuan Long et al. (+12 more)

We present SWE-MISBEHAVIOR, the first taxonomy and evaluation framework for diagnosing boundary-violating misbehaviors in coding agents. It contains 7.4K human-in-the-loop synthesized samples spanning three layers of agent action space: (i) Code Generation, (ii) File Management, and (iii) System Interaction. Across these categories, SWE-MISBEHAVIOR covers 11 distinct misbehaviors with a fully automated evaluation protocol, including coding-action violations, Create/Read/Update/Delete (CRUD) file misbehaviors, system integrity and abuse violations, and task-goal misalignment. Experiments on state-of-the-art open-source coding agents (e.g., OpenHands, smolagents) reveal substantial misaligned behaviors, including dependency misinterpretation (up to 98.5% of cases), unintended modification of user inputs (50.5%), and silent compromise of environment integrity (53%). Motivated by these findings, we further propose CODE-KIT, a lightweight adaptive framework that combines misbehavior-aware guidelines with novel tools to help agents avoid such failures. CODE-KIT reduces misbehaviors by 52% for smolagents and 34% for OpenHands, while simultaneously improving task completion performance.
Read the full abstract

All authors Do Xuan Trong, Duc Anh Nguyen, Do Xuan Long, Ngoc-Hai Nguyen, Tongyao Zhu, Duy Dinh, Cayden Chik, Cuong Dang, Jiaying Wu, Xuan-Phi Nguyen, Vinh Van Nguyen, Kenji Kawaguchi, Shafiq Joty, Nancy F. Chen, Min-Yen Kan

Statistics Theory

Interpolation and coherence in multiple testing

Ninh Tran, Dennis Leung and Jesse Hemerik

In the literature on multiple testing methods that control the familywise error rate, a fundamental concept is "coherence". A procedure is coherent if it is "logically consistent" (Hochberg and Tamhane, 1987) in the sense that if a hypothesis is rejected, then all hypotheses that imply it are also rejected. If a method is not coherent, then it can be made coherent without increasing the familywise error rate. Recently, the concept of coherence was generalized to a much broader class of multiple testing procedures, namely all methods that provide simultaneous confidence bounds for false discovery proportions. Such procedures were defined to be coherent if they cannot be further improved through a process called "interpolation". However, we show that such a coherent procedure can sometimes be further improved by exploiting remaining logical inconsistencies. This leads to what we call a "perfectly coherent" method. Thus, interpolation does not necessarily lead to perfect coherence. However, we provide conditions under which it does. Fortunately, many methods based on interpolation turn out to be perfectly coherent.
Read the full abstract

Logic in Computer Science

Reset Early, Reset Often, Eliminate Models

Martin S. Fixman, Fredrik Rømming and Sean B. Holden

Connection-tableau provers typically construct closed tableaux by backtracking search. This gives lightweight proof procedures, but their low-memory search state leaves little derived information to reuse after a failed attempt. SATCoP addresses this by retaining grounded tableau clauses in an incremental SAT solver, allowing the prover to refute the problem once the accumulated ground clauses become propositionally unsatisfiable. We introduce SATResetCoP, which relies more directly on this persistent SAT state by not backtracking on unproductive tableaux and instead starting new ones to generate further ground clauses for the SAT solver. On TPTP, SATResetCoP improves over our leanCoP-style baseline by 54% and our SATCoP-style baseline by 18%. On the bushy MPTP2078 benchmark, it solves 616 problems, a 49% improvement over our leanCoP-style baseline and a 12% improvement over our SATCoP-style baseline. These results suggest that changing tableau guidance to accumulate useful ground clauses faster improves the performance of the prover.
Read the full abstract

Optics

Optical parametric amplification of spatiotemporal optical vortices

Houan Teng, Chenhao Wan and Qiwen Zhan

Spatiotemporal optical vortices (STOVs) are structured light featuring a phase singularity in the spatiotemporal domain, exhibiting unique topology that couples spatial and temporal dimensions. To date, amplification of STOVs remains unexplored, and their wavelength tunability is limited by available light sources. In this work, we present the theoretical analysis and experimental demonstration of optical parametric amplification (OPA) of STOVs. The pulse energy of the STOV signal is enhanced by a factor of 60, while its spatiotemporal structure is robustly preserved. The generated idler inherits the phase-conjugated STOV topology with the opposite topological charge, governed by simultaneous conservation of energy, linear momentum, and spatiotemporal orbital angular momentum. The intrinsic wavelength-conversion capability of OPA further provides a route towards spectrally tunable STOV generation, spanning from visible to near-infrared wavelength regimes. This work paves the way toward further development of multi-stage amplification architectures, facilitating applications of wavelength-tunable spatiotemporally structured light at ultrahigh intensities.
Read the full abstract

Quantum Physics

Theory of Trotter Error with Variable-Weight Schedules

Henrique Ennes

Trotterization is a widely used algorithm for simulating quantum evolutions and consists of approximating a target unitary through a local linear decomposition of its generator that is repeated over $N$ steps. We introduce {scheduled trotterization}, a variable-weight modification in which the coefficients of the local linear decomposition are allowed to change from step to step. Using the geometry of the Lie group of unitary operators, we express the global endpoint error as a product of transported local defects, whose leading terms are obtained by two applications of the Baker-Campbell-Hausdorff formula, one inside each local step and one between different transported defects, separating the internal Trotter error from the external accumulation error. While we recover recent tighter bounds on the global approximation error when the weights are held fixed, the new framework allows us to propose a weight-varying schedule that significantly decreases the global error. For exact infinitesimal synthesis, we prove that this improves the usual fixed-weight $O(N^{-1})$ behavior to $O(N^{-2})$, without changing the basic pulse skeleton. We also extend the framework to imperfect infinitesimal synthesis, as arises in quantum gate synthesis problems, where admissible coefficients may only approximate the desired local corrections. In this case, the error is controlled by a local covering parameter measuring the quality of the available coefficient set. Numerical experiments for random $\SU(2)$ targets further support the predicted theoretical improvements. The results suggest a novel perspective for the analysis of Trotter errors, which can be treated as schedulable quantities to be coherently canceled at the endpoint.
Read the full abstract

Performance

How Reliable Are Large Language Models for Soil Organic Carbon Assessment? An Australian Benchmark

Yingxuan Cheng, Tong Li, Yayong Li et al. (+5 more)

Soil organic carbon (SOC) is an indicator of agricultural soil condition and carbon storage, but numerical outputs from large language models (LLMs) require validation before use in monitoring or management. We evaluated seven general-purpose LLM configurations on three direct-response tasks: estimating 0–10 cm SOC concentrations for 308 Australian paddocks under two information conditions, calculating SOC stock in 120 matched scenarios, and comparing 300 paddock pairs. References were measured SOC records or deterministic calculations; response validity, repeated-call dispersion and estimated API cost were assessed separately. Additional measured soil attributes reduced concentration mean absolute error by 13–62% across configurations, although high-SOC underprediction and paddock-specific errors persisted. Six configurations were highly accurate in stock calculations using equation-native units without coarse fragments, but accuracy declined with physically equivalent alternative units. Pairwise accuracy was 90.5–94.6% overall and 59.1–70.9% for 52 closely separated pairs; some configurations were sensitive to presentation order. Nearly all outputs were parseable, without necessarily being correct. The benchmark identifies three distinct requirements for AI-mediated SOC assessment: agreement with measured concentrations, unit-consistent stock calculations and reliable discrimination between sites. These findings support task-specific validation and auditable calculations, rather than unverified use of LLM outputs as ecological evidence.
Read the full abstract

All authors Yingxuan Cheng, Tong Li, Yayong Li, Yuchuan Fan, You-Gan Wang, Jinran Wu, Scott Chapman, Geoffrey J. McLachlan

Machine Learning

A Few More Tokens Ahead: Reinforcing Parallelism in Diffusion LLMs

Jiahao Kuang, Yangfan Gao, Tao Gui et al. (+4 more)

Parallel generation is widely regarded as an inherent advantage of diffusion large language models. Yet our fixed-quota evaluation of four representative models reveals severe degradation across mathematics and code, with all scores below 15% at 16 tokens per step. Token-level probes identify local predictive potential near the observed prefix, especially in mathematical text. We propose RePara, which reinforces short consecutive token blocks through group-relative policy optimization and reduces rollout and log-probability recomputation costs. On LLaDA and SDAR, \ours improves GSM8K and MATH-500 accuracy over JustGRPO under reduced denoising budgets, with a 1.96x training-step speedup in our LLaDA/MATH setup.
Read the full abstract

All authors Jiahao Kuang, Yangfan Gao, Tao Gui, Yuanbin Wu, Qi Zhang, Xuanjing Huang, Tao Ji

Combinatorics

A counting proof for the smallest critical set in a Latin square

Richard Bean

A critical set in a Latin square is a set of entries that determines the square uniquely and is minimal with this property. We prove that every partial Latin square of order $n$ with fewer than $\lfloor n^2/4\rfloor$ filled cells has an even number of completions. Consequently every critical set in every Latin square of order $n$ has at least $\lfloor n^2/4\rfloor$ entries. Since the cyclic Latin square has a critical set of exactly this size, the minimum size of a critical set over all Latin squares of order $n$ is $\lfloor n^2/4\rfloor$. This confirms a conjecture attributed to Nelder (1979) and made independently by Mahmoodian and by Bate and van Rees. Hatami and Qian had shown that this minimum is at least $n^2/10^4$ for large $n$. The proof is a counting argument modulo two that uses only expansion of products, cancellation in pairs and the pigeonhole principle, and it is illustrated with partial Latin squares of orders three to seven. An appendix discloses the substantial use of AI in developing the proof and writing the exposition.
Read the full abstract

Artificial Intelligence

QORH: Learning Query-Specific Ordered Rule Hypergraphs for Inductive Knowledge Graph Link Prediction

Naixing Xu

Knowledge graph link prediction involving unseen entities and relations requires learning transferable reasoning patterns from observed graph structure. Existing methods based on relation graphs primarily capture pairwise dependencies, limiting their ability to preserve the compositional semantics, relation ordering, and role information of complete path rules. To address this limitation, we propose QORH (Query-specific Ordered Rule Hypergraph), which jointly encodes rule heads and ordered rule bodies as hyperedges conditioned on the query relation, explicitly modeling joint dependencies among multiple relations. QORH performs learnable rule selection using statistical evidence, such as rule confidence and coverage, and enhances relation representations through role-aware hypergraph message passing to guide entity-level reasoning. The model's shared parameters are independent of specific entity or relation identifiers, enabling inductive prediction for unseen entities and relations. Extensive experiments on entity-inductive and fully inductive benchmarks demonstrate that QORH significantly outperforms the state-of-the-art baselines included in our evaluation, validating the effectiveness of rule hypergraph modeling for inductive link prediction.
Read the full abstract

Differential Geometry

Singularities of Mean Curvature Flow with Bounded Total Curvature

Yongheng Han and Zhen Wang

We prove that every tangent flow at the first singular time of a smooth, closed, embedded mean curvature flow has smooth support, provided that the total curvature satisfies $\sup_{t\in[0,T)}\int_{M_t}|A|^n\,d\mu_t<\infty$. This extends Ilmanen's smoothness result for tangent flows to higher dimensions.
Read the full abstract

Machine Learning (Statistics)

Hierarchical Time Series Forecasting with Operator-Valued Kernels

Junjie Yang, Adeline Fermanian, Claire Boyer et al. (+1 more)

Hierarchical time series forecasting requires generating accurate predictions that are coherent across multiple levels of aggregation. Existing methods either rely on two-stage pipelines with post-hoc reconciliation or on end-to-end models that often lack statistical guarantees. We introduce a unified framework that formulates hierarchical forecasting as a multivariate regression problem in a vector-valued reproducing kernel Hilbert space. By designing structured operator-valued kernels, our approach enforces coherence directly within the hypothesis space, ensuring that all predictions satisfy hierarchical constraints by construction. We further propose tailored kernel constructions for hierarchical data and derive excess risk bounds that characterize generalization performance. Empirical results on synthetic and real-world datasets demonstrate the relevance of the proposed method.
Read the full abstract

All authors Junjie Yang, Adeline Fermanian, Claire Boyer, Gérard Biau

Signal Processing

Link Both Views: Cross-Image Routing Attention for Dual-Endpoint V2I Beam-Pair Power Prediction

Shiliang Lu, Lu Bai, Ziwei Huang et al. (+1 more)

For sensing-aided communications, single-endpoint sensing offers only a partial view of propagation, motivating link-centric sensing across both endpoints. We propose cross-image routing attention (CIRA) for sparse association and mutual enhancement of propagation-relevant features from dual-endpoint RGB images. Building on CIRA, CiFormer predicts millimeter-wave (mmWave) beam-pair power matrices for vehicle-to-infrastructure (V2I) communications. Experiments show that CiFormer reduces mean absolute error and improves top-1 beam-pair selection accuracy by 3.15 percentage points over the strongest baseline. Visualizations and ablation studies support dual-endpoint sensing and selective cross-view interaction in capturing propagation-relevant dependencies, suggesting broader applications in wireless tasks beyond beam prediction.
Read the full abstract

All authors Shiliang Lu, Lu Bai, Ziwei Huang, Xiang Cheng

Robotics

Cross-Embodiment Affordance Transfer for Robotic Manipulation

Sawera Yaseen, Serge Thill and Rekha Raja

Robotic manipulation systems commonly learn object affordances from human-annotated datasets, which reflect interaction possibilities for a human hand rather than the physical constraints of a specific robot. For example, a surface region of an object identified as graspable by a human annotator may be geometrically unreachable or mechanically infeasible for a robot, yet this mismatch is rarely addressed explicitly. To tackle this challenge, we proposed a framework that translates human affordance annotations into robot-executable affordance predictions conditioned on the physical properties of the robot. The network predicts which surface regions of an object are physically executable by the robot, given a 3D point cloud of an object and a descriptor encoding the robot's geometry and constraints. The model was trained using human-annotated affordances together with robotic execution outcomes from multiple sampled approach trials in simulation, where execution outcomes determined which regions are executable for the robot. The evaluation was performed on a household objects dataset using two parallel-jaw gripper configurations on a UR5e robot arm. The framework outperformed a model that does not receive information about the robot. The experimental results demonstrated that affordance prediction must be conditioned on the robot and that human annotations alone are insufficient for learning robot-executable affordance representations.
Read the full abstract