Computer Vision
INTCORT: Training-Free Spatial Reasoning Enhancement for Vision-Language Models via Input Transformations and Confidence Routing
Haoran Sun, Jingqi Xu, Yanhui Li et al. (+3 more)
Vision-
Language Models (VLMs) have demonstrated remarkable capabilities in
multimodal tasks, yet they still exhibit poor ability in spatial
reasoning. Existing training-dependent and training-free enhancement methods suffer from high computational costs with catastrophic forgetting and internal mechanism interference that compromises general capabilities, respectively. In this work, we first verify two key hypotheses: appropriate
geometric image transformation and query-reversal transformation can recover incorrect spatial predictions, and correct predictions exhibit higher relation-
token confidence than incorrect ones. Based on these findings, we propose INTCORT, a training-free spatial
reasoning enhancement framework that constructs multiple inference views through input transformations and aggregates their predictions via relation-
token confidence routing, without modifying the VLM's internal mechanisms. Experimental results on several commonly-used
benchmarks demonstrate that INTCORT substantially improves spatial
reasoning accuracy across diverse VLMs, achieving an average improvement of 10.01% over all models and
benchmarks. Compared with prior works, INTCORT achieves superior performance with improvements of up to 25.01%.
Read the full abstract ▾
All authors Haoran Sun, Jingqi Xu, Yanhui Li, Enci Liu, Kaidi Xu, Yanwei Liu
Computation and Language
MSI-Bench: Evaluating Multi-Speaker Voice Interaction for Collaborative AI Agents
Chenxu Xiong, Dongming Shen, Yuzhi Tang et al. (+3 more)
Voice provides a natural and immediate interface for AI
agents. Many settings in which voice
agents could be useful, including meetings, households, and collaborative work, are inherently multi-speaker. Supporting these settings introduces challenges that are largely absent from one-on-one interaction. We introduce the Multi-Speaker Interaction
Benchmark (MSI-Bench) for evaluating multi-speaker voice interaction. Each test case is a short multi-party multi-turn audio scene with participant context, expected tool calls, and atomic rubrics. The
benchmark targets three capability families: multi-speaker
memory, multi-speaker instruction following, and multi-speaker
reasoning. It comprises 1,152 test cases, evenly split between Mandarin Chinese and English (576 each). The strongest configuration on each split passes all rubrics on only 66.8% of English and 54.5% of Mandarin cases, and the strongest open-weight configuration on 34.0% and 19.3%. Failure analysis separates perception from
reasoning: open-weight models are bottlenecked by the multi-speaker audio front-end, while frontier systems still fail speaker-scoped decision making on clean transcripts—and models across the board often respond when no one has addressed them. These results identify speaker-grounded perception, speaker-scoped decision making, and conversational restraint as concrete targets for future voice
agents.
Read the full abstract ▾
All authors Chenxu Xiong, Dongming Shen, Yuzhi Tang, Wentao Ma, Mu Li, Alex Smola
Astrophysics of Galaxies
A Newly Discovered VVVX Star Cluster Surfing the Galactic Disk
Elisa R. Garro, José G. Fernández-Trincado, Casmir O. Obasi et al. (+15 more)
We report the discovery of Garro 04, a previously unidentified
stellar cluster toward the
Galactic disc, uncovered as part of the ESO/KMOS VVVX-GalCen Spectroscopic Survey. This object was serendipitously confirmed as a
stellar overdensity in the VISTA Variables in the Vía Láctea eXtended (VVVX) survey footprint and subsequently confirmed through medium-resolution $K$-band spectroscopy obtained with the K-band Multi Object Spectrograph (KMOS) at the ESO Very Large
Telescope. The combined photometric and spectroscopic analysis reveals a coherent
stellar population with common radial velocities and chemical properties, supporting its classification as a genuine
star cluster rather than a statistical fluctuation of the disk field. We derive the fundamental properties of Garro 04, including its structural parameters, distance, kinematics, and compute its orbital properties, finding that its motion remains confined to the
Galactic disk region. Finally, our analysis suggests that Garro~04 is likely an old open cluster, or an object straddling the boundary between the canonical globular and open cluster classifications, with an estimated Age $ = 5 \pm 2$ Gyr and a
metallicity of $\mathrm{[Fe/H]} = -0.5 \pm 0.2$, finding an excellent agreement between the photometric and spectroscopic determinations.
Read the full abstract ▾
All authors Elisa R. Garro, José G. Fernández-Trincado, Casmir O. Obasi, Dante Minniti, Matías Gómez, Roberto K. Saito, Javier Alonso-García, Alonso Luna, Valentin D. Ivanov, Maria Gabriela Navarro, Sara Federle, Bernardo P. L. Pereira, Beatriz Barbuy, Devika Bhadrakumar Sindhu, Bruno Dias, Tali Palma, Ilaria Petralia, Sasi Saroon
Algebraic Geometry
On Free and Nearly Free Conjecture of Rational Unibranched Projective Plane Curves
Xiping Zhang
Let $C\subset \PP^2$ be a reduced irreducible complex plane curve with complement $U$. In this paper we prove that $\chi(U)\geq \nu(C)$, where $\nu(C)$ is the freeness defect of $C$ defined in \cite{Dim24}. When $C$ is a rational curve with unibranched singularities, this inequality proves the
conjecture of Dimca and Sticlaru, that a rational unibranched plane curve is either free or nearly free.
Read the full abstract ▾
Sound
Fast Time-Varying Exponentiated Convolution Methods for Generative Direction Dependent Reverberation
Yuancheng Luo
Spherical harmonic encoded acoustic sound-fields capture directional characteristics of room impulse responses that are useful for accurate spatial audio reproduction. However, high costs of multi-microphone measurements and numerical simulations motivate alternative data-set augmentation and synthetic data generation methods that supplement small collections. This paper introduces time-varying exponentiated convolution methods that transform both Gaussian noise and impulse responses into reverberation and modified spectral-decay fields respectively. We derive two recursive and fast convolution
algorithms that extend into the spherical harmonic domain, model smooth reverberation time distributions with non-stationary Gaussian processes, and realize an optimal filter design. Experiments evaluate computational performance, and validate out-of-distribution generated impulse responses.
Read the full abstract ▾
Applied Physics
A State-Space Framework for trivial and Topological Metamaterial Stochastic Analysis
Luiz Henrique Marra da Silva Ribeiro, Abhilash Sreekumar, Vinícius Fonseca Dal Poggetto et al. (+4 more)
Topological phononic
crystals and elastic metamaterials support edge states defined by global topological invariants, offering a route toward vibration-control and wave-guiding devices that remain functional in the presence of defects. However, manufacturing-induced spatial variability can perturb these invariants and compromise their robustness, making its quantification essential during design. In this work, we first demonstrate that a previously proposed linear time-varying (LTV) formulation is mathematically equivalent to the spectral element method based on transfer
matrices for elementary rod, Saint-Venant shaft, and Euler-Bernoulli beam theories. The deterministic formulation is then extended to
stochastic analyses through a
stochastic linear time-varying (SLTV) framework. The proposed methodology combines Monte Carlo simulations with
stochastic Fourier series and an analytical Karhunen–Loève expansion, providing closed-form
stochastic fields and their derivatives required by the LTV formulation. The SLTV approach enables the computation of
stochastic dispersion diagrams and forced responses of one-dimensional waveguides with arbitrarily varying geometry and mechanical properties. Because the LTV-based transition
matrix isolates individual wavemodes without requiring the mode tracking needed by conventional eigenproblem-based formulations, the framework is particularly suited for evaluating topological invariants, specifically the Zak phase, and assessing the robustness of topological bands under spatial variability. Numerical results for rods, shafts, and Euler-Bernoulli beams demonstrate the applicability of the proposed framework as a unified methodology for deterministic and
stochastic analyses of trivial and topological periodic waveguides under continuous spatial uncertainty.
Read the full abstract ▾
All authors Luiz Henrique Marra da Silva Ribeiro, Abhilash Sreekumar, Vinícius Fonseca Dal Poggetto, Vicent Romero-García, Dimitrios Chronopoulos, Carlos De Marqui Jr., José Roberto de França Arruda
Classical Analysis and ODEs
Some Results on Lame-Heun Spectral Geometry
Ubong Sam Idiong
We develop a unified spectral and complex analytic theory of the Lame equation through its canonical Heun realization. Starting with ellipsoidal separation, we derive essential elements, including the Lame-preserving anharmonic parameter action and finite accessory spectral
polynomials. We construct the global Heun monodromy representation on the four-punctured sphere, connected to the (2,2,2,2) pillowcase orbifold and the SL(2,C) character surface. The accessory parameter is framed within a Riemann-
Hilbert map that links spectral values to monodromy conditions. Utilizing Seifert van Kampen theory reveals the fundamental group of the punctured sphere, and the Nielsen Schreier rank formula applies to monodromy coverings. The vanishing of higher homotopy groups for unbranched covers and the emergence of a Riemann-Hurwitz genus formula post-compactification are noted. Colored Schreier graphs are employed for a visual calculus of the covering data, connecting
elliptic geometry and
topology.
Read the full abstract ▾
High Energy Physics (Theory)
Finite-Time Doppler Covariance Spectroscopy in the AdS/CFT correspondence
Feiyi Liu, Shiyang Chen and Yang Wang
From a
holographic perspective, we formulate an exact finite-duration covariance framework that resolves chirality in the AdS/CFT correspondence. In a rotating BTZ black hole spacetime, relative motion between two boundary controls converts sum-frequency matching on the compact boundary cylinder into discrete velocity resonances indexed by the angular momentum of the scalar field. Because the complete real-control covariance also contains difference-frequency sectors, a two-setting phase cycle with a common switching window cancels them exactly at finite duration and reconstructs the resonant pair contribution from complete positive-semidefinite covariance
matrices. For an isolated angular mode, reconstructed pair responses at opposite resonance velocities sample the boundary spectrum at opposite angular momentum and a common control-selected frequency. Their normalized contrast cancels reflection-symmetric control factors and the state-independent operator normalization, yielding a finite-linewidth
estimator of the reflected-mode BTZ contrast. Under the stated smoothness and moment conditions, common centered even narrowing windows give a bias beginning at quadratic order in linewidth. A real-projection leakage bound controls contamination from neighboring modes. Long-pulse calculations with the exact BTZ spectrum verify phase-cycle reconstruction and quadratic linewidth scaling. Bath-weighted calculations with exact truncated transfer amplitudes establish convergence to this baseline. Under the default controls, complete-sum contrasts for the first two nonzero angular modes reconstruct the reference-state rotation parameter at fixed mean inverse temperature with subpercent deterministic bias in the untruncated Gaussian limit. The inferred rotation parameter fixes the horizon parameters of the rotating BTZ black hole once the mean inverse temperature and AdS radius are known.
Read the full abstract ▾
Quantum Physics
Stabilizing temporal quantum-enhanced sensitivity via sub-optimal measurements
Wangsheng Zheng, Yaoling Yang and Abolfazl Bayat
In non-
equilibrium probes,
quantum-enhanced sensitivity is quantified through super-linear scaling of precision with respect to time, as a central metrology resource. However, this requires optimal measurements, which are often complex and time-dependent, making it challenging in practice. Sub-optimal measurements typically fail to retain robust super-linear scaling and show oscillations. Here, we begin by establishing a universal temporal scaling law in a general multi-parameter non-
equilibrium quantum sensing framework. Within this setting, we identify the universal behavior of the sensing precision with sub-optimal measurements which shows quadratic scaling modulated by an additional bounded oscillatory function. To remove the oscillatory part, we propose a protocol in which measurements are partitioned into different groups, each performed at different times. The collective precision obtained from this protocol stabilizes a robust quadratic scaling even for sub-optimal measurements in the multi-parameter regime. We validate our protocol through three distinct examples as well as
Bayesian estimation. Our protocol is applicable to every informative measurement and takes a key step towards practical realization of
quantum-enhanced sensitivity.
Read the full abstract ▾
Data Structures and Algorithms
FPT Isomorphism Test for $F$-Free Tournaments
Daniel Neuen
We show that isomorphism of $F$-free tournaments can be solved in FPT time $f(k) \cdot n^{O(1)}$, where $k$ denotes the size of $F$, and $n$ denotes the size of the input tournaments. Our result extends on a previous FPT isomorphism test for tournaments of bounded twin-width [Grohe, Neuen 2024], as well as XP isomorphism tests parameterized by the VC dimension or the chromatic number [Raßmann, Schweitzer 2026]. It also implies that every non-trivial hereditary class of tournaments admits a
polynomial-time isomorphism test. Our
algorithm builds on a novel combination of spectral,
geometric, combinatorial and group-theoretic tools.
Read the full abstract ▾
Superconductivity
Interface engineering of spin triplet Cooper pairs for spin-valve implementation
Debashree Nayak, Abhisek Sahoo, Pratap Kumar Sahoo et al. (+1 more)
Spin singlet Cooper pairs can be converted into equal-
spin triplet pairs at a
superconductor/heavy-metal interface under suitable conditions. Under current carrying conditions, these triplet correlations can give rise to a non
equilibrium spin moment in the heavy metal through the predicted supercurrent
spin-
Hall effect. Here we demonstrate that this current induced triplet
spin moment, in conjunction with the
magnetic moment of a
ferromagnetic layer, can enable a
magnetic spin-valve response in Nb/Pt/Ni/Pt/Nb vertical nano-devices. The magnitude of this
spin-valve effect depends on the efficiency of singlet-triplet Cooper pair conversion at the Nb/Pt interface. In order to facilitate efficient triplet generation, we deliberately introduced interfacial roughness to introduce finite Rashba
spin-orbit coupling at the Nb/Pt interface and an out-of-plane component of
magnetic moment at the Ni interface. In contrast, no discernible
spin-valve response was observed in devices with smooth interfaces within measurement resolutions. Since the
magnetic moment of the Ni layer and the current-induced triplet
spin moment in the Pt layer are independently switchable using a
magnetic field and bias current, respectively, the
spin valve can be controlled via either parameters. These results demonstrates a novel approach to directly utilizing the
spin polarized triplet Copper pairs in superconducting spintronic applications.
Read the full abstract ▾
All authors Debashree Nayak, Abhisek Sahoo, Pratap Kumar Sahoo, Kartik Senapati
Distributed Computing
Tiga: Compiling Graph Message Passing at Scale
Mingyuan Chi
Graph message passing offers a common way to express learning
algorithms, physical simulations, and numerical solvers. Efficient execution depends on interaction structure and data movement, which can be obscured when a program is expressed as a sequence of tensor operations. On
memory-constrained systems such as laptops, materializing connectivity and intermediate messages can also exhaust device
memory. We present Tiga, a just-in-time compiler that separates the definition of a message-passing program from how its interactions are traversed, computed, and stored. Tiga preserves graph relations and reducer
algebra in a multi-level intermediate representation, enabling traversal specialization, fusion of relation generation with aggregation, and reverse-mode automatic differentiation. Backend-specific lowering targets CPUs and GPUs. Its runtime extends execution beyond device-
memory capacity by streaming graph partitions from disk through host
memory with page-sized device staging buffers; partition ownership and
halo exchange extend the same programming model to distributed execution. A Python interface interoperates with ordinary PyTorch tensors and autograd for forward and backward computation. Numerical checks validate outputs and gradients for differentiable workloads. A differentiated
geometric workload characterizes the
memory–time tradeoff of fused forward and backward execution. Evaluation against matched Torch and PyTorch
Geometric baselines demonstrates reduced runtime and device-
memory use for generated-relation workloads, while offloaded forward execution processes billion-edge graphs on a single
memory-limited GPU. Measurements on heterogeneous devices further characterize the communication and load-balance costs of distributed execution.
Read the full abstract ▾
Cryptography and Security
Decoding Guardrails: XAI-Guided Perturbation Analysis of Prompt Injection Detection
Fernando Outeda, Gustavo Betarte, Juan Diego Campo et al. (+1 more)
Large
language models (LLMs) are increasingly deployed in production systems, raising concerns about their exposure to
adversarial manipulation through
prompt injection and jailbreak attacks. Classifier-based guardrails, such as
Prompt Guard 2, are widely used as a first line of defense against such attacks, but their internal decision logic is largely opaque to both defenders and attackers. This paper presents an exploratory case study that applies explainable artificial intelligence (XAI) techniques to analyze how
Prompt Guard 2 distinguishes malicious from benign
prompts. We conduct four experiments to probe this question empirically. Guided by Vanilla Gradient and SHAP attributions, we find that
Prompt Guard 2's decisions rely on the cumulative contribution of many
tokens rather than a few dominant ones, yet saliency-guided synonym substitution and sentence-level paraphrasing can flip its predictions while altering only a moderate fraction of the text, in some cases yielding a successful jailbreak against the underlying LLM. A
dataset-scale saliency analysis further shows that undetected injection
prompts systematically lack the lexical markers the classifier relies on. We discuss the implications of these findings for the design and evaluation of classifier-based guardrails, and argue that explanation methods intended to support transparency can simultaneously lower the cost of constructing successful
adversarial bypasses.
Read the full abstract ▾
All authors Fernando Outeda, Gustavo Betarte, Juan Diego Campo, Fiorella Cravero
Optics
Collective cavity quantum electrodynamics in solid-state optical clocks
Karen Mamian, Georgy A. Kazakov, Thorsten Schumm et al. (+1 more)
Solid-state frequency standards are generally limited by strong decoherence, rendering conventional interrogation schemes inefficient. The $^{229}$Th
nuclear clock provides a unique and timely platform for solid-state optical metrology and
nuclear cavity
quantum electrodynamics (QED), featuring a coherence time many orders of magnitude shorter than the radiative lifetime in current experiments. Here, we propose and analyze three cavity QED-enhanced clock interrogation schemes that turn this timescale mismatch into an advantage, leveraging collective coupling of thorium
nuclei to nanophotonic modes to enable fast interrogation and detection despite the long population lifetime. We reveal the central role of collective cooperativity in determining the clock frequency instability, and derive the optimal conditions (power, working-point detuning, thorium density) for clock operation.
Read the full abstract ▾
All authors Karen Mamian, Georgy A. Kazakov, Thorsten Schumm, Charles Roques-Carmes
Computation and Language
When Quantization Preserves Accuracy but Not Evidence: Explanation-Aware Post-Training Quantization for Medical LLMs
Yeji Kim, Mi-Young Kim and Randy Goebel
Post-training quantization (PTQ) enables efficient deployment of large
language models, and PTQ methods are usually optimized and evaluated with generic reconstruction, perplexity, or answer accuracy. But in explanation-critical domains, preserving only the final answer may be insufficient, since users may also inspect generated rationales to judge whether a prediction is trustworthy. We study this issue in medical multiple-choice question answering, where rationales should provide evidence that supports the selected answer. We propose an explanation-aware objective for transformation-based PTQ. Our method builds an offline faithfulness cache from full-precision teacher rationales and uses it during optimization to preserve answer-supporting evidence
tokens and evidence-conditioned answer behavior. We instantiate it on OSTQuant under W4A4KV4 quantization and evaluate four 7B–8B medical and instruction-tuned LLMs on MedExQA, MedExpQA, and ChallengeClinicalQA. While a same-calibration OSTQuant baseline preserves task accuracy, it can substantially weaken answer-supporting rationales. Our objective is to preserve the full-precision model's answer-supporting behavior rather than improve gold-label accuracy, and our method better preserves the full-precision model's answer behavior and rationale-to-answer support. These results suggest that PTQ for explanation-critical settings should evaluate preservation of answer-supporting evidence, not only answer accuracy. Code and evaluation scripts are available at https://github.com/dut0817/EAQuant.
Read the full abstract ▾
Optics
Non-Hermitian mode locking in an L-band monolithic InP dual-ring laser
Xiao Sun, Mohanad Al Rubaiee, Jue Wang et al. (+4 more)
Coupling between
laser cavities provides a means of controlling optical modes through their gain, loss, and hybridization. Here we demonstrate electrically controlled mode locking in a monolithic InP dual ring
laser operating in the L band without a dedicated saturable absorber. Independent biasing of the two rings accesses single mode emission, mode locking, and hybridized supermode operation, with a measured intermode beat frequency of 542 MHz in the latter regime. The
laser generates 7.9 ps pulses at 18.11 GHz. A coupled cavity model describes how differential nonlinear detuning modifies the modal gain loss contrast, and time domain simulations yield pulse durations and spectral widths close to those measured experimentally. In a device incorporating saturable absorbers, reverse bias enables 3.1 ps pulses and reduces the integrated timing jitter from 9.06 to 2.46 ps relative to the unbiased condition. These results connect electrical control of coupled-cavity states with picosecond pulse generation in a monolithic InP
semiconductor platform for high repetition rate optical communications and microwave photonics.
Read the full abstract ▾
All authors Xiao Sun, Mohanad Al Rubaiee, Jue Wang, John H. Marsh, Stephen J. Sweeney, Anthony E. Kelly, Lianping Hou
Machine Learning
Complex KDA: Understanding and Enhancing the Expressivity of Kimi Delta Attention
Julien Siems, Riccardo Grazzi, Korbinian Pöppel et al. (+8 more)
Linear RNNs based on the delta-rule enable efficient sequence modeling, but their linear updates with a low-rank correction constrain their expressivity. Prior work has shown that composing two delta-rule transitions in a single recurrent update can model a 2D rotation, but this increases the rank and the cost of the updates compared to a single transition. We show that Kimi Delta
Attention (KDA) can realize 2D rotations by combining a single delta-rule transformation with a second reflection supplied by its channel-wise gate. This requires extending the parameter ranges of KDA by combining two existing range extensions: allowing gates in $[-1,1]$ and the delta-rule coefficient $\beta$ in $[0,2]$. We call the resulting model Complex KDA (CKDA). It preserves KDA's stability and efficiency, with transitions that remain diagonal-plus-rank-one and non-expansive, while reaching the state-tracking expressivity of DeltaProduct$_2$. We characterize the expressivity of CKDA and prove that every orthogonal diagonal-plus-rank-one
matrix is exactly a CKDA transition
matrix. A single CKDA layer can track every finite group isomorphic to a subgroup of $\mathrm{SO}(3)$, and many state-tracking results use one fewer layer for CKDA compared to other diagonal-plus-rank-one Linear RNNs. Empirically, combining both extensions yields the strongest length extrapolation among tested KDA range settings on $S_3$, $S_4$, and periodic audio continuation. In
language modeling, CKDA outperforms
Transformers and other linear RNNs, obtains similar results to a KDA baseline, and shows promising scaling behavior. Our code is open source at https://github.com/OpenEuroLLM/ComplexKDA and our models are available at https://huggingface.co/collections/openeurollm/complexkda.
Read the full abstract ▾
All authors Julien Siems, Riccardo Grazzi, Korbinian Pöppel, Jaisidh Singh, Arber Zela, Timur Carstensen, Jenia Jitsev, Frank Hutter, Volkan Cevher, Antonio Orvieto, Aaron Klein
High Energy Physics (Experiment)
Goodness-of-fit for multi-distribution neutrino cross-section measurements with shared events
Francisco Martinez Lopez, London Cooper-Troendle, Steven Gardiner et al. (+2 more)
When multiple differential cross-section measurements are extracted from a common event sample, the same events contribute simultaneously to multiple distributions. This event-sharing structure imposes exact linear constraints among the bin counts, reducing the effective dimensionality of the measurement below the total number of bins. Since the covariance
matrix obeys these inter-distribution constraints, it contains a rank deficiency. While limited numerical precision may inadvertently restore invertibility, the $\chi^{2}$ contributions along constrained dimensions will be arbitrary, yielding unphysically inflated $\chi^{2}$ values in global goodness-of-fit tests. We present the range-projected $\chi^{2}$, a test statistic that restricts the goodness-of-fit test to the subspace carrying independent statistical information, yielding a $\chi^{2}$ with $N_\text{bins} - N_\text{null}$ degrees of freedom, where $N_\text{null}$ is the number of independent constraints. We show that this rank deficiency is a predictable consequence of the event-sharing structure, and that $N_\text{null}$ decomposes into a structural contribution determined a priori from the binning geometry and a kinematic contribution that depends on the phase-space occupancy. The method is validated with an analytical toy model and a simulated
neutrino–argon cross-section measurement including unfolding and multi-source systematic uncertainties.
Read the full abstract ▾
All authors Francisco Martinez Lopez, London Cooper-Troendle, Steven Gardiner, Patrick Green, Tanaz Mohayai
Logic
Infinitary provability logic
Mojtaba Mojtahedi, Fedor Pakhomov and Giovanni Soldà
Gödel-Löb provability logic $\GL$ is a propositional modal system that on one hand enjoys completeness with respect to conversely well-founded Kripke frames and on the other hand captures all modal principles about $\PA$-provability that are provable in $\PA$ itself. In the present paper we carry out an initial investigation into the question of what the infinitary counterpart of $\GL$ is. We develop a non-well-founded deep inference proof system $\dgla$ for the modal
language with at most countably infinite conjunctions and disjunctions. We show that the calculus is sound and complete for well-founded transitive Kripke frames. Using Kripke-Platek set theory we develop an interpretation of the infinitary modal
language in terms of infinitary provability over admissible sets. Then we show that a natural Hilber-style variant of infinitary $\GL$ is sound for this interpretation. We leave open, however, the question if $\dgla$ proves any additional
theorems in comparison with the
Hilbert-style calculus. Nevertheless, under certain conditions we do show that the infinitary provability logic arising from certain admissible sets lies between the set of
theorems of the
Hilbert-style calculus and the non-well-founded deep inference system.
Read the full abstract ▾
Quantum Physics
Observable-targeted variational quantum simulation of Hamiltonian dynamics
Leonardo Zambrano, Luciano Pereira and Antonio Acín
Standard variational
quantum simulation seeks to reproduce the
evolution of the full
quantum state, although many applications require only the expectation values of a few observables. We study a variational method for pure-state
Hamiltonian dynamics that updates
circuit parameters to reproduce the
evolution of selected expectation values. An exact error identity guides the choice of observables, motivating a construction based on repeated commutators of the target with the
Hamiltonian. For Pauli observables and Pauli-rotation
circuits, the update can be estimated without ancillary
qubits or controlled operations for overlap estimation. Across six- and ten-
qubit spin, fermionic, and molecular
benchmarks, the observable-targeted update extends the median time within the target-error tolerance compared with standard variational
quantum simulation with equal shot budgets. These results show that directing the variational update toward the target observable can extend accurate simulation without increasing the measurement cost per time step.
Read the full abstract ▾
History and Philosophy of Physics
The wavelength of light as Thomas Young invented it
Olivier Morizot
This paper documents Thomas Young's invention of a concept which he alternately refers to as the ”magnitude”, ”breadth”, ”interval” or ”length of an undulation” of light. First, the paper highlights that Young was first to link some concept of a wavelength to a theory of optics and to assign a precise value to this length for each component of the colour spectrum. Then, it sets out an explanation of how Young made up these values. Finally, it examines the reasons why Young introduced an optical wavelength into his theory of light, insofar as these reasons might in turn shed light on why he was first to do so.
Read the full abstract ▾
Materials Science
III-V antiphase boundaries are not generated by Si or Ge substrate step edges
Charles Cornet, Sreejith Pallikkara Chandrasekharan, Audrey Gilbert et al. (+7 more)
We critically review recent experimental and theoretical advances on the valence- and structure-mismatched heteroepitaxy of III-V
semiconductors on group-IV substrates. We examine key aspects including wetting behavior, atomic configurations at the hetero-interface, substrate passivation, and the stability of antiphase boundaries (APBs). By synthesizing these findings with pioneering studies, we propose a refined description of antiphase boundaries formation. Our analysis shows that hetero-interface formation is dictated by substrate terrace reconstruction, demonstrating that APBs are not generated at monoatomic step edges. This generalized III-V/IV growth framework provides new insights for the integration of III-V
semiconductors on group-IV platforms and, more broadly, for valence-mismatched heteroepitaxy.
Read the full abstract ▾
All authors Charles Cornet, Sreejith Pallikkara Chandrasekharan, Audrey Gilbert, Milan Silvestre, Rozenn Bernard, Pascal Turban, Gilles Patriarche, Eric Tournie, Laurent Pedesseau, Jean-Baptiste Rodriguez
Human-Computer Interaction
Detecting Agitation Before Behavioral Escalation in Autistic Youth Through Multimodal Wearable Sensing
Nibraas Khan, Abigale Plunk, John Staubitz et al. (+8 more)
Challenging behaviors including aggression, self-injury, and property destruction are observed in 68% of autistic youth and pose risks to youth and caregivers. These episodes are preceded by agitation, a rising state of distress expressed through movement, vocalization, and autonomic arousal. Its signs are subtle and individualized, and its autonomic components are invisible without instrumentation. We collected upper-body movement from inertial measurement units, physiology from a wrist-worn device, and vocalizations from lapel microphones across 30 clinician-led sessions with 15 autistic youth, paired with expert behavioral annotations. We adapt four pretrained foundation models, one per modality, project each to a shared 128-dimensional space, and fuse them into a single group model. The model detected agitation with an area under the ROC curve of 0.724 at the clinician-annotated onset (within-participant permutation p=0.0005), declining to 0.608 at 30,s before onset. Thirteen of fifteen participants were above chance. A from-scratch configuration reached only 0.58, while frozen and fine-tuned features performed comparably (0.71 and 0.72). Audio contributed most of the signal, and a watch-only configuration stayed near chance. Individualized agitation is therefore detectable, including in unannotated windows preceding the annotated onset, using foundation-model transfer with one shared model rather than one per child.
Read the full abstract ▾
All authors Nibraas Khan, Abigale Plunk, John Staubitz, Ingrid Shragge, Jordan Brooks, Suzanne Wright, Alec Brewer, James Dieffenderfer, Alper Bozkurt, Amy Weitlauf, Nilanjan Sarkar
Combinatorics
The $g$-theorem in smooth approximation
Joel Hakavuori
We show that a sequence of simplicial polytopes approximating a smooth convex body has primitive Betti numbers diverging from the upper bound given by the $g$-
theorem. This complements a result of Adiprasito-Nevo-Samper, who proved the corresponding divergence from the lower bound. The proof uses a new lower bound for the size of the shadow of a set of monomials.
Read the full abstract ▾
Probability
The stochastic Landau-Lifshitz-Baryakhtar equation in critical spaces
Federico Butori, Foivos Evangelopoulos-Ntemiris, Lorenzo Marino et al. (+1 more)
We study the
stochastic Landau–Lifshitz–Baryakhtar equation with multiplicative noise and homogeneous Neumann boundary conditions on smooth bounded domains in dimensions $d\le 3$. Relying on
stochastic maximal regularity results, we establish global well-posedness in a variational strong, intermediate, and weak setting, as well as for initial data in the scaling-critical Besov spaces $B^{d/q-1}_{q,p}$ for every $q\in[2,\infty)$ and $p\in(2,\infty)$.
Read the full abstract ▾
All authors Federico Butori, Foivos Evangelopoulos-Ntemiris, Lorenzo Marino, Max Orteu-Capdevila
Computer Vision
Streaming Video Editing with Easy Adaptation
Yujia Hu, Jiajun Li, Zihao He et al. (+1 more)
In this paper, we propose SVEET, a framework that requires merely training on a pretrained bidirectional
video diffusion model but supports high-quality streaming
video editing in an auto-regressive fashion. To tackle this problem, we first systematically revisit existing
video-to-
video diffusion approaches and identify two key principles for such streaming adaptation: backbone feature disentanglement and conditional frame independence. Building on these insights, we develop a novel paradigm for controllable
video generation. At its core, an auxiliary model branch encodes source
video inputs with temporally independent self-
attention, and the intermediate features are injected into the corresponding backbone blocks for streaming-compatible control. Moreover, to bridge the discrepancy between the feature spaces of bidirectional and streaming models, we propose a decoupled training scheme that explicitly enforces the orthogonality between the optimization directions of
video controllability and model causality. Such disentanglement ensures compatibility between the two objectives at inference and facilitates smooth zero-shot knowledge transfer across heterogeneous backbone architectures. Extensive experiments demonstrate that SVEET achieves superior editing quality while maintaining real-time performance, attaining 15 FPS on a single H100 GPU 17 without any auxiliary acceleration techniques. Codes are available at https://github.com/YujiaHu1109/SVEET.
Read the full abstract ▾
All authors Yujia Hu, Jiajun Li, Zihao He, Songhua Liu
Computer Vision
Toward a foundation model for forest point clouds
Yuanwen Yue, Stefano Puliti, Damien Robert et al. (+7 more)
Forest inventories increasingly rely on artificial intelligence (AI) models to derive forest attributes from large-scale 3D point clouds. Current models are typically specialized to a single task, sensor, and forest type, making adaptation expensive in terms of annotations, computation, and expertise. We ask whether a single pretrained model can instead learn transferable representations across diverse forest inventory settings. Inspired by recent developments in
language modelling and computer
vision, we take a step toward a foundation model (FM) for 3D forestry. Using LitePT as backbone, we first establish a strong supervised baseline that sets a new state of the art on forest semantic and instance
segmentation, tree species classification, and age regression
benchmarks. We then curate a large-scale unlabelled corpus spanning airborne,
UAV, and mobile
laser scanning across diverse forest ecosystems, and pretrain the same backbone using self-supervised learning. We systematically evaluate representation learning strategies by comparing training from scratch, supervised pretraining, and self-supervised pretraining across four representative forestry tasks, under varying annotation budgets. Compared with training from scratch, self-supervised pretraining accelerates model convergence and consistently improves performance when annotations are scarce. Compared with task-specific supervised pretraining, self-supervised pretraining yields more transferable representations across downstream forestry tasks. These findings identify the practical regime in which pretrained representations are most valuable and suggest that instance discrimination, rather than forest semantics, is the main remaining obstacle to a general-purpose 3D forest foundation model. Code and models are available at: https://github.com/prs-eth/ForPT.
Read the full abstract ▾
All authors Yuanwen Yue, Stefano Puliti, Damien Robert, Atakan Topaloğlu, Binbin Xiang, Maciej Wielgosz, Jan Dirk Wegner, Rasmus Astrup, Christian Rupprecht, Konrad Schindler
Differential Geometry
Positive Scalar Curvature on Foliations: Leafwise Band Width
Hengyu Chen
We establish an optimal leafwise width estimate for compact,
spin, foliated bands with positive leafwise scalar
curvature and infinite vertical $\widehat{A}$-cowaist. This result addresses a special case of the foliated version of Gromov's band width
conjecture and generalizes the scalar and mean
curvature comparison
theorem for
spin bands due to Cecchini and Zeidler. Our approach, which also applies to
spin foliations, relies on a local boundary value problem for deformed sub-Dirac operators on the Connes fibration. A key technical ingredient is a Lichnerowicz-type formula in the adiabatic limit for almost isometric foliations, which captures the sharp coefficient $\operatorname{rank} F/(\operatorname{rank} F - 1)$ associated with the integrable subbundle $F$.
Read the full abstract ▾
Fluid Dynamics
A high-order ghost-point immersed boundary method coupled with auxiliary differential equations for time-domain impedance walls
Abdoulaye Ouattara
A ghost-point immersed boundary method is coupled with auxiliary differential equations to impose a locally-reacting impedance condition on a wall that does not conform to a Cartesian grid, in the framework of the linearized Euler equations. The ghost-point reconstruction follows the normal-stencil approach: an elliptical least squares cloud of fluid nodes is used to build a high-order
polynomial along the local wall normal, which is then extrapolated onto the ghost points by one-dimensional La grange interpolation. The impedance, modeled as a massspringdamper oscillator, is advanced in time through two auxiliary state variables per ghost point, coupled to the reconstruction at every RungeKutta stage. The method is validated against exact and semi-analytical references on a flat immersed wall, a cylindrical wall reflecting an acoustic pulse, and a cylinder scattering a harmonic monopole. Convergence orders ranging from about three on the curved wall to about four and a half on the flat wall are measured, and a cross-check against an exact harmonic reference shows that a persistent offset observed against a semi-analytical reference is a property of that ref erence rather than of the coupled solver. A spectral stability analysis of the linearized semi-discrete operator further shows that the normal-stencil reconstruction reduces, without eliminating, a known instability that occurs when the immersed wall becomes tangent to the grid.
Read the full abstract ▾
Artificial Intelligence
Convex AI Compositionality and the Governance of AI System Populations
Andrea Ferrario
AI governance increasingly requires providers and public authorities to reason about multiple AI instantiations, alternative versions, and deployment configurations of multiple AI systems. Yet current regulation remains predominantly single-system-centric, acknowledging such multiplicity only sparsely without treating collections of related AI systems as governance objects. This creates an AI population governance problem: determining which instantiations can be meaningfully considered together and how their changing configurations can be represented and monitored. The first requirement has recently been addressed through trustworthiness-based accounts of AI identity. We address the second by introducing convex AI compositionality: a formal representation of the configurations generated by finite AI system populations that uses convex spaces. The core idea is that convex compositions of the operational states that a population of AI system instantiations may occupy over time are compatible with lifecycle reachability across the population and can preserve the formal identity relations between these systems. Well-known statistical and
geometric constructions, such as weighted state distributions and convex hulls, become AI governance tools for distinguishing operational states, population weights, heterogeneity, and AI configuration change across different governance modes while remaining compatible, under stated conditions, with lifecycle reachability and AI identity. We illustrate our AI population governance framework through distributed healthcare deployments and controlled deployment of recruitment AI variants.
Read the full abstract ▾