Peer Revenue Past citations are no guarantee of future returns

Papers

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

All of arXiv

Friday 9 October

Group Theory

The $R_\infty$-nilpotency degree of some lollipop GBS groups

Wagner Sgobbi, Dalton Couto Silva and Daniel Vendrúscolo

We say that a group $G$ has the $R_{\infty}$ property if the number $R(\varphi)$ of twisted conjugacy classes is infinite for any automorphism $\varphi$ of $G$. The $R_{\infty}$-nilpotency degree is the least integer $c$ such that $G/\gamma_{c+1}(G)$ has property $R_{\infty}$; in case this integer does not exist, we say that $G$ has infinite $R_\infty$-nilpotency degree. In this paper, we determine the $R_{\infty}$-nilpotency degree of some lollipop GBS groups by using different approaches, such as Sylow subgroups, normal forms and presentation of nilpotent quotients, and diagrams of exact sequences. We also compute the $R_\infty$-nilpotency degree of all 2-generated lollipop GBS groups and of all GBS groups whose associated graph is a tree.
Read the full abstract

Computers and Society

Culture in the Code: Anthropological Concepts Decoding AI's Hidden Assumptions

Sonja Schmer-Galunder

Contemporary AI development operates on foundational assumptions about intelligence, rationality, and human values that reflect narrow cultural perspectives, limiting alternative conceptions of how AI can be globally beneficial. This paper introduces seven anthropological concepts – origin story, monoculture, thick description, personhood and community, cargo cult, structuralism, and liminality – to expose hidden assumptions and offer alternative frameworks for understanding AI alignment. The paper examines how AI systems embody Western-centric notions of rationality as utility optimization, personhood as bounded individuals, and values as quantifiable preferences extracted from behavior. The concepts surface three assumptions that current alignment debates rarely examine: that values can be inferred from behavior, quantified, and formalized; that such quantified values capture what actually matters; and that values, once defined with sufficient precision, are appropriate optimization targets. Against "thin alignment" – optimization for behavioral proxies stripped of context – the paper proposes directions toward "thick alignment": recognizing that values emerge through social practice rather than existing as extractable properties; defining more appropriate metrics that capture not only end goals but the conditions that lead to long-term flourishing; and moving from models of fixed identity toward relational ontologies where personhood emerges through social participation. The paper concludes that if AI systems are to create conditions conducive to human flourishing rather than optimize metrics that flatten social experiences, three shifts are needed: redistributing interpretive sovereignty to affected communities, protecting what resists optimization, and extending accountability across generations.
Read the full abstract

Quantum Physics

Below-threshold execution of fault-tolerant ripple-carry addition on a trapped-ion quantum computer

Jacob S. Nelson, Andrew J. Landahl and Andrew D. Baczewski

We deploy a universal fault-tolerant quantum computing (FTQC) architecture based on error-detecting $[\![8,3,2]\!]$ and $[\![4,2,2]\!]$ codes on Quantinuum's Helios-1 trapped-ion quantum computer. We compile and execute ripple-carry adder (RCA) circuits for classical inputs up to five bits on this architecture. Uncomputing a workspace qubit provides an additional “algorithmic check” to detect errors. While available at both the physical and logical levels, this check synergizes with our architecture by increasing the effective fault distance of many circuit locations from two to four. Combining this algorithmic check with the error-detecting codes, we observed only a single logical failure in one-bit addition after burst-error filtering, yielding an estimated logical failure rate of $(6.0 \pm 5.9) \times 10^{-5}$. This is well below the $(3.5 \pm 1.3) \times 10^{-3}$ failure probability observed for one-bit addition on the same hardware in unencoded form. Two- and three-bit adders also showed performance improvements, while the longer four- and five-bit adders executed just above threshold.
Read the full abstract

Functional Analysis

Strong-type endpoint obstructions for bilinear spherical maximal operators

Vicente Vergara

For the bilinear spherical averages in dimension $d\geq2$, we show that the lacunary maximal operator is not bounded from $L^1(\mathbb R^d)\times L^q(\mathbb R^d)$ to $L^r(\mathbb R^d)$ whenever $1\leq q\leq\infty$ and $1/r=1+1/q$. The obstruction is realized by fixed nonnegative inputs; a second construction proves quantitative failure on smooth compactly supported functions and finitely many dyadic scales. The same conclusions hold for the full maximal operator. We give a sharp Lorentz-index criterion in the range $q>d/(d-1)$ and extend the negative result to all multilinear spherical maximal operators with an $L^1$ input. We indicate the precise strong-type boundary statements to which these conclusions apply, without drawing conclusions about separate weak-type or interior estimates.
Read the full abstract

Robotics

ClothCode: Generative Modeling for Garment State Estimation and Dynamics Prediction

Yingdong Ru, Lipeng Zhuang, Shiyu Fan et al. (+3 more)

We introduce ClothCode, a unified generative framework for garment state estimation and action-conditioned dynamics prediction in a shared discrete configuration space. A material-aligned tokenizer encodes garment configuration as residual vector-quantized patch tokens, while task-specific Masked Generative Transformers predict these tokens through parallel confidence-guided refinement. Across three garment templates, ClothCode reduces mean displacement error by 26.6% for state estimation and 37.0% for dynamics prediction relative to the strongest baselines. Dynamics prediction requires only 16.7 ms per window, compared with 556.2 ms for diffusion. Integrated into the same model-predictive controller, ClothCode improves real-world folding and flattening success from 56.7% to 78.9%. Further results and videos are available at: https://anonymous.4open.science/r/ClothCode-DE32
Read the full abstract

All authors Yingdong Ru, Lipeng Zhuang, Shiyu Fan, Zhuo He, Florent P. Audonnet, Gerardo Aragon Camarasa

Statistical Methodology

Dynamic Extended Generalized Many-Facet Rasch Model: A Unified Approach to Rater-Mediated Assessment

Giuseppe Mignemi, Patrícia Martinková and Sally Paganin

Rating procedures are crucial in psychological and medical diagnosis, educational assessment, peer review, and grant evaluation. A major methodological challenge is disentangling multiple sources of variability in ratings, such as subject ability, rater behavior, and item characteristics. Item Response Theory-based models, most notably many-facet Rasch models and hierarchical rater models, have been widely adopted for jointly estimating subject, rater, and item effects, while retaining high interpretability. We introduce the Dynamic Extended Generalized Many-Facet Rasch model, a general class of Bayesian semiparametric models that integrate and extend existing approaches. Its measurement model allows location and discrimination parameters to depend jointly on item and rater characteristics and incorporates subject-item and rater-subject variation. The dynamic extension introduces a rater-specific cross-lagged effect capturing dependence on the previously rated subject, and a Bayesian P-spline capturing systematic changes in discrimination over the rating sequence, accounting for learning, fatigue, or adaptation effects. Subject heterogeneity is addressed through Bayesian nonparametric mixtures, flexibly learning the latent ability distribution without strict parametric assumptions and enabling model-based subject clustering. We also derive a closed-form, model-based inter-rater consensus coefficient and study its characteristics. We discuss theoretical properties, develop posterior computation strategies, and evaluate the model in simulation studies and two real-data applications.
Read the full abstract

Algebraic Geometry

A Functorial Classification of Toric Stacks

John S. Nolan

Toric stacks enrich the study of toric varieties by accounting for quotients of actions of diagonalizable groups. In practice, toric stacks are constructed and understood via "stacky fans" generalizing the fans appearing in toric geometry. Previous attempts to make this construction into an equivalence of categories have either run into crucial errors or have been restricted to special cases (e.g. only considering separated Deligne-Mumford stacks). In this paper, we solve this problem by introducing a new 2-category $\mathsf{StkFan}$ of stacky fans and constructing an equivalence of 2-categories between $\mathsf{StkFan}$ and a 2-category $\mathsf{TStk}$ of toric stacks. Along the way, we develop a theory of "toric categories" which categorifies the classical theory of toric monoids. We also explain how working with toric categories provides a simple axiomatic framework for Gale duality.
Read the full abstract

Information Retrieval

Simple Linear Corrections for Compression in Late Interaction

Mihir Agarwal, Sabyasachi Basu and Viraj Agrawal

Late-interaction (multi-vector) models represent documents using a set of token-level vectors, providing a middle ground between cross-encoders and dense encoders in terms of quality and cost of embedding. However, compared to the latter, the indices are costly to store and score. Pooling and pruning based compression techniques are deployed widely but deteriorate in quality noticeably with aggressive compression. We investigate whether a lightweight change to the scoring function can recover retrieval quality after document-vector compression. For each combination of encoder, compressor, and vector budget, we learn a global linear correction by distilling the uncompressed model's score distributions over candidate documents. The encoder and compressor remain frozen, and the correction can be applied to each query at search time or once, offline, to the stored vectors. We evaluate the method with ColBERTv2 and Jina v4 on MS MARCO and BEIR, and with ColPali2 on the ViDoRe v3 visual document retrieval benchmark. Experiments cover four training-free compression methods, multiple vector budgets, in-domain adaptation, and transfer from MS MARCO to BEIR. In the in-domain setting, the correction improves over baseline compression in 154 of 160 BEIR settings, while transfer from MS MARCO is more variable across models and datasets. We also find that matrices trained with synthetic queries can improve over baseline compression on ViDoRe v3. These results show that a small learned correction can recover some of the quality lost to compression without retraining the encoder or recompressing the documents.
Read the full abstract

Materials Science

Precursor stoichiometry controls dopant structure, defect chemistry, and valence in Mn-doped ZnGa$_2$O$_4$ nanocrystals

Mátyás M. Rudolf, Gábor Bortel, Nikoletta Jegenyés et al. (+5 more)

The luminescence of doped nanocrystals is usually optimised through the activator; here we show that the host cation ratio is an equally decisive, and largely overlooked, design variable. Mn-doped ZnGa$_2$O$_4$ nanocrystals (Mn:Zn = $4\times10^{-4}$) were grown hydrothermally at nominal Zn/Ga ratios between 0.30 and 0.75 (stoichiometric: 0.50) and post-annealed at 400–1100$^\circ$C, and the resulting composition–temperature matrix was mapped by X-ray- and UV-excited luminescence, EPR, Rietveld-refined X-ray diffraction, Raman spectroscopy and SEM. The as-grown crystals are phase-pure spinel at every composition and accommodate their off-stoichiometry internally as a point-defect inventory, reflected in a 0.4% variation of the lattice parameter; on annealing, the excess gallium is exsolved as $\beta$-Ga$_2$O$_3$ in amounts set by the solution composition. The Mn$^{2+}$ emission separates into two optima, near-stoichiometric crystals annealed at 400–600$^\circ$C and Zn-excess crystals annealed at 1100$^\circ$C, the latter being twice as bright under X-ray excitation. Under X-ray and 365~nm excitation the two optima are separated by a minimum at 800$^\circ$C that coincides with a collapse of the Mn$^{2+}$ EPR population and has no counterpart in the Raman linewidths, identifying transient, defect-compensated oxidation of the dopant rather than lattice degradation as its origin. The most Zn-rich composition shows neither Mn$^{2+}$ nor Mn$^{4+}$ emission at any annealing stage, whereas its trace Cr$^{3+}$ emission is strong, indicating that Mn is rejected already during growth. Solution stoichiometry therefore acts on two levels, dopant capture and native-defect inventory, while annealing controls dopant valence and lattice order. The brightest high-temperature phosphor is thus not the nominally stoichiometric one, but the composition whose excess is consumed on its way to stoichiometry.
Read the full abstract

All authors Mátyás M. Rudolf, Gábor Bortel, Nikoletta Jegenyés, Árpád Jakab, László Bencs, Ferenc Simon, Adam Gali, Dávid Beke

Cryptography and Security

Embedding Asymmetry as Attack Signal: Detecting and Defending Active Label Inference in Vertical Federated Learning

Soumya Swagata Biswas and Muhammad Abdullah Adnan

Vertical federated learning lets organizations train a shared model over features they hold separately, while one party keeps the labels private. The active label inference attack of Fu et al. [12] breaks this: a feature-holding party runs a modified optimizer, MaliciousSGD, that forces class structure into its embeddings and then recovers the labels with semi-supervised learning. Defenses that perturb the shared gradient fail, because the attack acts on the party's internal optimizer, not on the message it sends. We show the attack leaves a signal the server can read from the embeddings it already receives. The adversary's Fisher class-separability pulls away from the honest party's, and the gap is visible within a few epochs. A defense triggered by this signal acts on the adversary alone. Scaling its gradient down handles few-class tasks, and a norm non-expansive projection handles many-class tasks, with the Discriminative Coverage Ratio explaining why the projection rank must grow with the class count. The defense breaks the alignment between embedding clusters and labels rather than destroying the embedding. Across four seeds on CIFAR-10, CIFAR-100, and CINIC-10, label inference falls below what honest training already leaks, at under two points of task-accuracy cost.
Read the full abstract

Numerical Analysis

Global Convergence and Reliable Stopping in a Safeguarded Modified Anderson–Björck Method

Latchezar Tomov and Nedelcho Ganchovski

The modified Anderson-Bjorck (modAB) method was first proposed in 2023 by Ganchovski and Traykov for solving nonlinear equations in structural mechanics. It was improved further in 2026 by introducing better switching criteria with account to function symmetry, and bisection fallback based on measuring the bracket reduction rate. Although the algorithm demonstrated superior performance across a great variety of functions, it lacked formal mathematical analysis and proof of convergence. In addition, some discrepancies were noted by L. Tomov in the latest version that need to be addressed: premature application of the Anderson-Bjorck (AB) correction on the first AB step, caused by retaining the endpoint-update record from the preceding bisection step, and using the corrected endpoint ordinates in symmetry factor calculation instead of the true residuals. The purpose of the present paper is to provide rigorous mathematical proof of global convergence in exact arithmetic, remove the above discrepancies, and provide floating-point safeguards for interpolation, switching and stopping, with explicit treatment of infinite and NaN values, underflow, overflow and round-off errors. These safeguards do not certify the accuracy of the supplied function values. Additional performance improvement was introduced by allowing for early bisection fallback when AB fails to make progress on reducing either the bracket width or the residual, while retaining a finite bound on consecutive exceptions to the prescribed width condition.
Read the full abstract

Quantum Physics

Lifting an $[n,k,d]$ Binary Linear Code to an $[[n,k,d]]$ Quantum Stabilizer Code

Thamadathil Aswanth and Navin Kashyap

The codeword stabilized (CWS) framework provides a systematic method for constructing quantum error correcting codes from a simple undirected graph and a classical code. When the underlying classical code is linear, the resulting quantum code is a stabilizer code. In this work, we investigate the conditions under which an arbitrary $[n,k,d]$ classical binary linear code can be lifted to an $[[n,k,d]]$ quantum stabilizer code using the CWS framework. We analyze the random CWS code constructed from a given $[n,k,d]$ binary linear code $\mathcal{C}$ and an Erdős-Rényi random graph $\mathrm{G}(n, \frac12)$. We establish a lower bound on the probability that the resulting quantum code preserves the minimum distance $d$ of the classical code. This probabilistic analysis yields a Gilbert-Varshamov-type condition on the $n,k,d$ parameters of the classical code that guarantees the existence of a graph resulting in a quantum code with the same $n,k,d$ parameters.
Read the full abstract

Robotics

Observation-Referenced Tissue Deformation Forecasting with Tool-Driven and Residual Dynamics

Tom Olesch, Nan Xiao, Xiongye Xiao et al. (+2 more)

Forecasting tissue deformation from future instrument motion could support predictive planning and increasingly autonomous surgical systems. We study this problem using a scene-specific differentiable surface model fitted to an observed tool-tissue interaction window. Rather than forecasting absolute simulated positions, we introduce an observation-referenced formulation that anchors each tissue point to its last valid observation and adds only the deformation induced by the known, preplanned future tool trajectory. This reduces sensitivity to reconstruction, initialization, and persistent simulation offsets while preserving the predicted deformation response. The physical model uses differentiable, mesh-free position-based dynamics with scene-specific rigidity, attachment, region scale, and grasp coupling. We further introduce an optional observation-referenced residual model combining graph-spectral spatial modes with periodic harmonic dynamics for structured motion not explained by the tool input. We evaluate the framework retrospectively on 12 surgical manipulation clips from three datasets. Observation referencing provides the largest gain, with a median within-clip error ratio of $0.598$ relative to the corresponding absolute simulation forecast. The optional periodic residual further improves forecasting when the remaining motion exhibits periodic structure. These results support observation-referenced deformation prediction as a practical building block for future surgical world models, digital twins, and generative simulators that must predict how tissue evolves under instrument actions.
Read the full abstract

All authors Tom Olesch, Nan Xiao, Xiongye Xiao, Jonathan C. DeLong, Fei Liu

Artificial Intelligence

Cross-Modal Latent Influence: Patch-Level Training Sample Attribution for Vision-Language Models

Laura Yao, Shixing Yu, Jie Peng et al. (+3 more)

Explaining a vision–language model (VLM) prediction requires tracing training-example influence to the multimodal evidence that mediates it, yet existing attribution methods largely operate only at the sample level. Recent latent-mediated influence methods provide feature-level attribution in language models, but lack the multimodal and spatial structure needed to ground influence in VLMs. We introduce MOSAIC (Multimodal Orthogonalized Sparse Autoencoders for Influence Characterization). It learns a shared sparse image–text latent space whose concepts are 1) sparse within each modality, 2) supported on contiguous image regions, 3) aligned to the phrases that describe them, and 4) approximately decorrelated. It then applies EK-FAC to decompose training-example influence through these concepts and localize their contributions to specific text phrases and image regions. We evaluate MOSAIC on Qwen and Gemma across captioning, chest X-ray understanding, and visual question answering. The learned features are more localized and coherent without degrading task behavior, and ablating influence-ranked latents changes model outputs more than ablating latents ranked by activation or frequency. We find that visual evidence can substantially reorganize training-data attribution while preserving more of the shared latent structure. MOSAIC thus connects where model behavior comes from with what multimodal evidence carries that influence.
Read the full abstract

All authors Laura Yao, Shixing Yu, Jie Peng, Promit Ghosal, Kyra Gan, Tianlong Chen

Algebraic Geometry

Gauss-Manin Connection in Disguise: Holomorphic Anomaly Equations and Higher Genus Real Gromov-Witten Invariants

Felipe Espreafico

We construct an algebraic framework for the extended holomorphic anomaly equations following ideas of Alim, Movasati, Scheidegger and Yau. We choose the quintic threefold as our main model and use this framework to solve the equations and fix the ambiguity up to genus 5. We obtain predictions for real Gromov-Witten invariants, or, equivalently, real Gopakumar-Vafa invariants that are integers and agree with their complex counterparts modulo 2. In the end, we discuss how the framework could be extended and compute the same predictions for the bicubic up to genus 9.
Read the full abstract

High Energy Physics (Phenomenology)

A Comprehensive Interpretation of the LZ High-Energy Recoil Event: A Minimal $U(1)_B$ Dark Sector and Its Experimental Tests

Mattia Di Mauro

The $248$ keV nuclear-recoil candidate reported by LUX-ZEPLIN (LZ) motivates an inelastic dark matter (DM) interpretation, in which scattering converts the DM particle into a slightly heavier state. We provide a comprehensive study of this interpretation in an anomaly-free model of gauged baryon number, where a new vector boson, $Z_B$, mediates interactions between DM and quarks. Breaking the baryonic gauge symmetry generates the new particle masses and a small splitting between two DM states, while a residual symmetry keeps the lighter state stable. We identify regions compatible with the LZ candidate, the observed DM abundance, and current ATLAS and CMS dijet constraints. Within the scanned domain, these regions span DM masses $m_\chi\simeq0.19$-$4.5$ TeV, mediator masses $M_{Z_B}\simeq0.05$-$5.0$ TeV, and gauge couplings $g_B\simeq0.004$-$0.74$, with correlated values of these parameters. A representative thermal solution has $m_\chi=1$ TeV, $M_{Z_B}=1.44$ TeV, $g_B\simeq0.228$, and a mass splitting of $300$ keV. A small electroweak mixing allows the heavier DM state to decay rapidly, suppressing unwanted signals from its surviving population. This interpretation can be tested through complementary measurements: additional LZ and XENONnT data can establish whether the high-energy recoil signal persists, ATLAS and CMS Run-3 dijet searches can probe the mediator mass and coupling, and a dedicated IceCube analysis of four-quark final states can test DM capture and annihilation in the Sun. Together, these measurements can test complementary parts of the surviving region and assess whether the LZ candidate is consistent with this interpretation.
Read the full abstract

Networking

Semantic Edge-Cloud Communication for Real-Time Urban Traffic Digital Twins

Murat Arda Onsu, Arman Dogru, Poonam Lohan et al. (+2 more)

Urban traffic systems are becoming increasingly complex, demanding real-time traffic perception, proactive collision prediction, and intelligent management to improve safety and reduce congestion. Although AI-driven traffic management has shown strong promise, its success relies on diverse, high-quality data spanning critical scenarios such as collisions, near-collisions, and traffic rule violations. Digital twins offer a powerful solution by enabling synchronized virtual replicas of physical traffic environments for real-time monitoring, collision prediction, traffic-flow optimization, and diverse realistic scenario generation. However, continuously synchronizing a high-fidelity urban traffic digital twin through edge surveillance cameras requires substantial edge-cloud communication, and transmitting raw video streams leads to severe bandwidth and latency constraints. To overcome this challenge, this paper proposes a semantic communication-based edge-cloud framework for real-time urban traffic digital twin applications. By leveraging edge intelligence to extract and transmit only task-relevant semantic traffic information instead of full video feeds, the proposed framework significantly reduces communication overhead while preserving the essential dynamics needed for accurate real-time digital twin reconstruction. Experimental results demonstrate more than 99% transmission savings and an average trajectory reconstruction error of only 0.12 m, confirming the proposed framework as an effective, scalable, and low-latency solution for AI-enabled urban traffic digital twin applications.
Read the full abstract

All authors Murat Arda Onsu, Arman Dogru, Poonam Lohan, Burak Kantarci, Aisha Syed

Artificial Intelligence

Visualizing the World Model in an Agent's Value Function

Shawheen Naderi, Fatemeh Akherati and Yubei Chen

Could model-free reinforcement learning algorithms actually be learning a model of the environment? In this work, we apply counterfactual interventions which hold everything about a state fixed, besides a targeted change, to explore if the value function of a Proximal Policy Optimization agent forms a world model of the Crafter environment. Despite only being trained to optimize reward, we discover that the value function learns concepts such as the rules governing when a diamond is collectible, water and food having an impact on health, heuristics regarding inventory, and more. Finally, we discuss the safety implications of our visualizations and some shortcomings of current reinforcement learning agents that will prevent us from building truly advanced machine intelligence.
Read the full abstract

High Energy Astrophysical Phenomena

A Pipeline for the AI-Assisted Analysis of Imaging X-ray Polarimetry Explorer Observations of Black Hole X-ray Binaries and Limits on Their Polarization Variability

B. Domine, B. Liu, H. Krawczynski et al. (+3 more)

The Imaging X-ray Polarimetry Explorer (IXPE) has opened a new window into the high-energy Universe, with the capability to measure the arrival time, arrival direction, energy, and, for the first time with the required sensitivity, the polarization of X-rays from black holes, neutron stars, and other cosmic sources of X-rays. In this paper, we discuss the use of Artificial Intelligence (AI) agents for the analysis of IXPE data. The pipeline combines an immutable code base with evolving knowledge databases. The generation of analysis projects makes use of a literature data base to judge the novelty of the proposed analysis projects. We emphasize that the analysis requires two types of expert knowledge: detailed instructions about how to reduce the X-ray data, and expert knowledge about how to interpret the results. We employ several levels of auditors that scrutinize the analysis plan, the analysis scripts, and the analysis results for data and method adequacy. The pipeline includes a staggered approach to interpret the analysis results. We present exemplary results from the pipeline applied to IXPE observations of Black Hole X-ray Binaries (BHXRBs) including comparisons of results from the analysis pipeline with published results and a systematic search for polarization degree and polarization angle variability in long IXPE observations. We critically discuss the performance of the pipeline and future avenues for improving it.
Read the full abstract

All authors B. Domine, B. Liu, H. Krawczynski, M. Kundu, Shravan Vengalil Menon, Fang Zhou

Machine Learning

ULPack: Reference-Conditioned Lossless Compression of Checkpoint Trajectories

Aniket Srinivasan Ashok

Studying model training requires checkpoint trajectories, whose storage cost limits temporal resolution. Although adjacent checkpoints are numerically close, existing lossless codecs either operate on XOR bit patterns or ignore trajectory-specific conditional structure. We show that a reference weight's exponent predicts the bit length of its displacement in ordered floating-point space without side information. We introduce ULPack, a lossless codec which exploits this dependence by conditioning entropy models of ordered-integer residuals on the reference exponent. Across bfloat16, float16, and float32 trajectories, ULPack reduces storage by up to 27% against ZipNN, 21% against zstd-19, and 11.5% against FM-Delta. Across two dense trajectories, compression improves as updates shrink and reference conditioning strengthens, while near uniform low-order bits set the remaining compression floor and earlier checkpoints add little information.
Read the full abstract

Strongly Correlated Electrons

Exact SU($N$) symmetry resolution of bosonic Hubbard models: general boson-fermion mappings and flat-band singlet-cluster states

Loïc Herviou

We show how to perform exact diagonalization of SU($N$)-symmetric Bose-Hubbard models separately in each irreducible representation of SU($N$), by exploiting the Howe duality between U($L$) and U($N$). Three structural consequences follow for any lattice and any Hamiltonian built from the U($L$) generators: (i) the spectrum in a given orbital sector is independent of $N$ and of the particle statistics, (ii) the canonical partition functions for bosons and fermions obey the exact identity $Z_{\mathrm{b}}(-N)=(-1)^{N_\mathrm{b}}Z_{\mathrm{f}}(N)$, (iii) with unfrustrated hopping the groundstate is always a color ferromagnet, so that the color-resolved method we introduce is useful to study excited sectors, frustrated lattices and finite temperatures. As applications, we first derive the finite-temperature phase diagram of attractive SU($N$) bosons on few-site rings at large $N$ for arbitrary filling. We then show the emergence of singlet-cluster states from frustrated hopping, the counterpart of fermionic ferromagnetism, with the example of the sawtooth and kagome lattice.
Read the full abstract

High Energy Physics (Phenomenology)

Disappearing Tracks from Electroweak Vector Dark Matter at the LHC

Fabián Hernández-Pinto, Patricio Escalona and Alfonso Zerwekh

We study the disappearing-track signature of a spin-1 minimal dark matter candidate transforming in the real five-dimensional representation of $SU(2)_L$ with vanishing hypercharge. Radiative electroweak corrections induce a compressed mass spectrum among the multiplet components, allowing the singly charged state $V^\pm$ to attain a macroscopic lifetime and decay via $V^\pm \to V^0 \pi^\pm$ inside tracking detectors. We recast the ATLAS search for disappearing tracks with the full Run-2 dataset of $137\text{ fb}^{-1}$, originally targeted at long-lived charginos, computing the tracklet reconstruction probability analytically from the published efficiency so that the full lifetime range is covered. The procedure reproduces the ATLAS limits for pure winos within $6\%$ in mass. The Run-2 data exclude quintuplet masses up to $1080\text{ GeV}$ for proper decay lengths $c\tau \simeq 250\text{ mm}$, and $M_V < 630\text{ GeV}$ for a mass splitting of $166\text{ MeV}$. For the radiative mass splitting predicted by the model with a small Higgs-portal coupling, masses between $120\text{ GeV}$ and $240\text{--}300\text{ GeV}$ are excluded. With $3\text{ ab}^{-1}$ and a Run-2-like analysis, the sensitivity would extend to $1750\text{ GeV}$ for vanishing background and to $1530\text{ GeV}$ for a linear extrapolation of the instrumental background.
Read the full abstract

Astrophysics of Galaxies

Globular cluster Observations for Luminosity-dispersion Distances (GOLD): Project description and velocity catalogue

K. Fahrion, M. A. Beasley, A. Gvozdenko et al. (+10 more)

We introduce and describe the Globular cluster Observations for Luminosity-Dispersion distances (GOLD) project, which has the principal aim of determining independent distances based on the globular cluster (GC) absolute magnitude-velocity dispersion ($M_V - \sigma$) relation calibrated with Milky Way GCs with Gaia distances. The project comprises observations of GCs in 19 early- and late-type galaxies taken with FLAMES/GIRAFFE on ESO's Very Large Telescope. So far, 17 galaxies have been observed. The spectra cover the wavelength range 8483-9000 AA, including the calcium triplet region at a spectral resolution of $R\sim18,000$. The spectra provide line-of-sight velocities with typical uncertainties of $\Delta v = 1-2$ km s$^{-1}$. Using primarily velocities to separate extragalactic sources from foreground stars, we present here our initial sample of 1062 GCs and 101 ultra-compact dwarf galaxies (UCDs). For more than 650 of these objects we present the first velocity measurement. We discuss the available photometry and size measurements, and show that our velocities are in good agreement with previous studies. Our velocity catalogues are made publicly available. Beyond the principal aim of obtaining precision distances, additional science to be explored with these data includes host galaxy halo dynamics and clustering analysis, mass-to light ratios of GCs and UCDs in the sample, the stellar initial mass function and black hole occupation fraction of the GCs, and chemical abundances from high-dispersion spectra.
Read the full abstract

All authors K. Fahrion, M. A. Beasley, A. Gvozdenko, I. Ortega-Casas, A. Ferre-Mateu, F. I. Aros, D. Dold, T. Jerabkova, S. S. Larsen, O. Müller, M. Rejkuba, K. L. Rhode, G. van de Ven

Artificial Intelligence

Unconstrained CTI-REALM with Open-Weights

Sahil Garg, Ruhi Gottumukkala, Vishal Agarwal et al. (+1 more)

Organisations that want language-model agents for security work face a choice between open-weight models, which they can run and audit themselves, and far more expensive closed frontier models. We study that choice on CTI-REALM (arXiv:2603.13517), a benchmark in which an agent turns a threat-intelligence report into a working detection by querying a live telemetry store. We evaluate twenty-six models, nineteen open-weight and seven closed, under a common harness, scorer and panel of LLM judges. Our main finding is that an open-weight model can match the strongest closed models at less than a tenth of their cost, provided it is given enough turns. The benchmark's published limit on agent messages is a confound: a task that exhausts the limit scores zero on every checkpoint, including those it had already passed, so results at that limit partly measure which models finish in time. At the published limit the leading open-weight and closed models cannot be distinguished. With the limit relaxed, the models it was truncating recover, and the highest rewards belong to open-weight models. Among the leading models, price is a poor guide to capability: models with similar rewards differ in cost by more than an order of magnitude. Because the benchmark's checkpoint scores largely record what an agent did or mentioned, we also introduce stage-level metrics that score what the agent submits, by executing its detection queries in the telemetry store and judging its outputs against the answer key. Rather than combining them into one more overall reward, we show that different models lead different stages, which suggests routing each stage of the workflow to the model that performs it best rather than choosing a single model by its overall score. We also report how many tokens each model generates, and find that the share of output spent on thinking does not track the reward.
Read the full abstract

All authors Sahil Garg, Ruhi Gottumukkala, Vishal Agarwal, Shubham Vishwakarma

Astrophysics of Galaxies

Stellar discs and intermediate-mass black holes in galactic nuclei II. Fragmenting the disc and forming the S-stars in a live nuclear star cluster

Taras Panamarev and Bence Kocsis

We present direct $N$-body simulations of a stellar disc around the supermassive black hole Sgr~A*, perturbed by an intermediate-mass black hole (IMBH) and embedded in a live nuclear star cluster. A retrograde IMBH with $m_\bullet=2000\,M_\odot$, relative inclination $\theta_\bullet=150^\circ$, and initial semi-major axis $a_\bullet=0.15$~pc breaks the disc into three disc-like components and an angular momentum filament. The stochastic torques of the cluster, together with the IMBH's inward migration, increase the angular separation between these substructures. By ${\sim}5$~Myr, their separations become comparable to those among the observed Galactic-centre young star structures. The innermost ($a\le0.04$~pc) disc stars evolve into an S-star-like population. Relaxation driven by the cluster disperses their orbital planes, excites their eccentricities and depletes the `zone of avoidance' within ${\sim}10$~Myr even without an IMBH. The IMBH further accelerates their orbital-plane reorientation. However, the most eccentric S-stars may require an additional formation channel, such as the Hills mechanism. The distinctive signatures of the IMBH are the well-separated disc components and the eccentricity excitation confined to the radial range swept by its orbit. Dynamical friction from the cluster reduces the orbital size of a $2000\,M_\odot$ IMBH initially at $0.15$~pc by more than a factor of two within $10$~Myr. Prograde IMBHs tend to align with the stellar disc, whereas a $2000\,M_\odot$ retrograde IMBH can anti-align with the radially overlapping disc stars. We also quantify how the IMBH affects the supply of stars to the loss cone and the rates of stellar tidal disruptions and binary separations.
Read the full abstract

High Energy Physics (Phenomenology)

Collider and direct detection as key probes of asymmetric dark matter: Signals from a compressed Inert Higgs doublet

David Cerdeño, Miguel Escudero, Pietro Ghedini et al. (+4 more)

The cosmic imbalance of particles over antiparticles may be inherited by the dark matter in a certain class of asymmetric dark matter models. In this article, we thoroughly investigate the signatures of the minimal realization of this idea. In this model a new (stable) inert Higgs, $H'$, plays the role of the mediator that facilitates the asymmetry transfer from the visible to the dark sector. It also constitutes a small dark matter sub-component, while the dominant component stems from a singlet scalar asymmetry. A combination of phenomenological and cosmological constraints requires this new inert Higgs doublet to be quite light $46\,{\rm GeV}<m_{H'}< 580\,{\rm GeV}$ and its spectrum rather compressed, $\Delta M = m_{H^\pm}-m_{H_0} \lesssim 1\,{\rm GeV}$ and $\delta m = m_{A_0}-m_{H_0}\lesssim 5\,{\rm MeV}$. Therefore its charged component can be relatively long-lived on collider length-scales. By recasting searches for compressed Higgsinos at the LHC, we derive new limits on this particle and show that, while at present most of the parameter space for this state is still unconstrained, large and key regions of parameter space will be tested at the HL-LHC by a combination of searches for disappearing charged tracks and displaced soft tracks. Quite remarkably, the lifetime of the charged component of the doublet can be correlated with the dark matter interaction rate at direct detection experiments and we show that the next generation of direct detection experiments has a unique sensitivity to the parameter space in the model. The fact that a small neutral mass splitting is a prediction of the asymmetric setup makes it an interesting candidate to account for the high energy nuclear recoil event recently reported by the LZ experiment. More generally, our detailed phenomenological study holds for any dark matter model involving an inert Higgs doublet with mass below 580 GeV.
Read the full abstract

All authors David Cerdeño, Miguel Escudero, Pietro Ghedini, Guilherme Guedes, Thomas Hambye, Abel Muñoz-Navarro, Elena Pressendo

Quantum Physics

Compass in the Mirror: Quantum Backpropagation with the Hopf Ansatz

Ruge Lin and Ranyiliu Chen

A variational quantum state may have many control angles, and its complete gradient contains one derivative for each. The task is to turn access to selected derivatives into a shared finite-shot record for the complete vector. The Hopf ansatz supplies the setting: it constructs any normalized pure state from nested two-way magnitude splits and, for complex states, leaf phases, with one directly programmed angle per split or phase. Earlier Hopf constructions prepare the state, recover coordinates from known amplitudes, and provide indexed derivative access. With a phase-calibrated controlled reflection, the inverse Hopf frame maps each normalized magnitude direction to a computational label, so every all-$X$ outcome contributes to every magnitude derivative. A separate signed leaf record supplies the complex phase block. A checkpoint variant reverses only the suffix below a selected depth and returns that layer's gradient block. For a complete chart with $M$ real-valued parameters, sufficient execution counts are $\mathcal{O}(\log\log M)$ for the global method and $\mathcal{O}(\log M\log\log M)$ for the complete checkpoint schedule at fixed additive coordinate accuracy and confidence. The global saving is record reuse, not a reduction of the $M$-entry output. Under the designated direct-angle schedules, the global method's assigned Hopf CNOT work and input/output-inclusive storage have at most logarithmic overhead relative to a matched scalar evaluation. The same Hopf geometry thus carries a state forward and the objective's complete coordinate response backward.
Read the full abstract

High Energy Physics (Theory)

Defect conformal field theories with free-in-the-bulk fermions

Christopher P. Herzog and Daniel Lloyd-Jones

We study codimension-two conformal defects for massless fermions that are free in a four-dimensional bulk, but interact on a two-dimensional defect. The bulk Dirac equation strongly constrains defect conformal field theory data: it fixes the dimensions of operators in the bulk-defect operator product expansion and, for bulk-defect-defect three-point functions, enforces a double-twist relation when only one member of a shadow pair is allowed by unitarity. The lowest modes of monodromy defects provide an important exception, since both shadow partners can be unitary and their three-point functions are instead related by a shadow transform. We illustrate these structures with defect interactions, including an exactly tractable bilinear flow and Gross-Neveu-type quartic couplings. At a perturbative mixed fixed point, we find an explicit interacting example in which the double-twist condition is violated and the two shadow channels combine to give a regular bulk-defect-defect correlation function.
Read the full abstract

High Energy Physics (Theory)

Semiclassical state on algebra of observables for de Sitter space and time reversal symmetry

Muxin Han and Yichen Hu

In de Sitter static patch holography, coupling bulk quantum fields to an observer's dynamical clock yields a type II$_1$ von Neumann algebra $\mathcal{M}_+$ of gauge-invariant observables. However, the unique tracial state on this algebra fails to reproduce semiclassical bulk correlation functions. We resolve this issue by introducing a family of heavy-clock states, representing an observer reference frame with heavy inertia. We prove that in the heavy-clock limit, $n$-point correlation functions of gauge-invariant dressed observables reproduce the bulk quantum field correlation functions on de Sitter spacetime. Furthermore, we investigate a generalized observer Hamiltonian $\widehat{H}_{\rm obs}=|\widehat{\pi}|$ that incorporates both forward-running ($\pi\geq 0$) and backward-running ($\pi <0$) clocks. The resulting von Neumann algebra $\mathcal{M}$ of gauge-invariant observables remains a type~II$_1$ factor and is isomorphic to $\mathcal{M}_+\otimes M_2(\mathbb{C})$, where $M_2(\mathbb{C})$ is the algebra of $2\times2$ complex matrices transforming between the two clock orientations. Time-reversal symmetry is implemented in this setting as an anti-unitary operator that exchanges the two clock orientations, yielding a type~II$_1$ von Neumann algebra $\mathcal{F}$ of gauge-invariant, time-reversal-invariant observables. We generalize the heavy-clock state construction to both $\mathcal{M}$ and $\mathcal{F}$, showing that $n$-point correlation functions calculated in heavy-clock states match the corresponding semiclassical bulk correlation functions. Finally, by coupling Einstein gravity to the Brown–Kuchař dust field, we derive the observer Hamiltonian $\widehat{H}_{\rm obs}=|\widehat{\pi}|$ from a perturbative reduction along the observer's worldline in the $G \to 0$ limit. This provides a field-theoretic foundation for the observer model developed in this work.
Read the full abstract

Astrophysics of Galaxies

Searching for Nearby Supermassive Black Hole Binaries with GRAVITY

GRAVITY Collaboration: K. Abd El Dayem, N. Aimar, A. Berdeu et al. (+81 more)

We present a systematic interferometric search for close supermassive black hole (SMBH) binaries in 20 nearby type 1 active galactic nuclei (AGN) observed with VLTI/\grav. We investigate parsec-scale asymmetric emission by modeling the squared visibilities and closure phases with a single circular Gaussian component plus a background point source for the central AGN, and searching for additional compact companions through grid-based binary models. The robustness of the search is validated using injection-recovery simulations and point-spread-function diagnostics. Several features are identified in the initial significance maps, but only one candidate out of 20 shows a persistent compact asymmetry, corresponding to an observed incidence of $5\%$. If NGC~3783 is interpreted as a binary AGN, we derive a Poissonian one-sided $1\sigma$ upper limit of $f_{\rm bin}<16.5\%$ for GRAVITY-detectable binaries among bright nearby AGN (K<12.5) at separations of $\sim1$-$100$~pc. Conversely, if the observed asymmetry in NGC~3783 is associated with asymmetric circumnuclear dust emission, the corresponding upper limit on the binary fraction is $f_{\rm bin}<9.2\%$. We compare these results with predictions from the Horizon-AGN simulation, finding comparable binary fractions. We also noted that the expected fraction strongly depends on the assumed distribution of individual SMBH accretion rates within binaries, which is currently unresolved in cosmological models. Our results demonstrate the potential of infrared interferometry to identify parsec-scale binary AGN candidates and motivate future GRAVITY+ observations to expand the sample and follow up known candidates, helping to confirm their nature and distinguish genuine binaries from asymmetric nuclear structures.
Read the full abstract

All authors GRAVITY Collaboration: K. Abd El Dayem, N. Aimar, A. Berdeu, J.-P. Berger, G. Bourdarot, W. Brandner, Y. Cao, C. Correia, Q. D'Amato, R. Davies, D. Defrère, A. Drescher, F. Eisenhauer, M. Fabricius, H. Feuchtgruber, S. Flesch, N. M. Förster Schreiber, A. Foschi, Q. Fournier, P. Garcia, R. Garcia Lopez, R. Genzel, S. Gillessen, F. Gonté, X. Haubois, S. F. Hönig, L. C. Ho, M. Houllé, S. Joharle, J. Kammerer, A. Kaufer, P. Kervella, J. Kolb, L. Kreidberg, L. Labadie, S. Lacour, O. Lai, R. Laugier, J.B. Le Bouquin, J. Leftley, R. Li, K. Li, B. Lopez, D. Lutz, F. Mannucci, F. Mang, A. Mérand, F. Millour, M. Montargès, N. Morujão, H. Nowacki, M. Nowak, S. Oberti, J. Osorno, T. Ott, C. Paladini, S. Pappert, T. Paumard, K. Perraut, G. Perrin, R. Petrov, P. O. Petrucci, N. Pourré, S. Rabien, D. C. Ribeiro, S. Robbe-Dubois, M. Sadun Bordoni, J. Sanchez-Bermudez, J. Sauter, M. Scialpi, J. Scigliuto, J. Shangguan, T. Shimizu, F. Soulez, C. Straubmeier, E. Sturm, M. Subroweit, C. Sykes, L. J. Tacconi, P. Thévenet, I. Urso, F. Vincent, J. Woillez, G. Zins