Quantum Physics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 14 September
The role of entanglement in quantum optimization remains actively debated. To address this question, we focus on the fixed-parameter expanding-depth regime of the quantum approximate optimization algorithm (QAOA), where a compact two-parameter schedule is trained once on small instances and then applied as the problem size and circuit depth increase. To probe this regime beyond full state-vector simulation, we perform approximate matrix product state simulations for up to 50 qubits and 100 layers and quantify entanglement by the bond dimension. We observe an entangle–disentangle profile, with the peak bond dimension decreasing with depth and eventually saturating. This observation motivates an extreme approximation: projecting the state onto the product-state manifold (bond dimension one) after every two-qubit interaction. Based on this approximation, we introduce BOND-1, a quantum-inspired classical solver. Despite the drastic simplification, BOND-1 achieves cut ratios above 0.95 relative to the best known values on standard GSet MaxCut benchmarks with up to 20000 variables, and in some cases it matches those values. It achieves these results without per-instance optimization and has linear memory cost, while per-instance tuning can provide further improvement. These results show that, in this regime, a substantial fraction of the optimization power of QAOA survives even in the complete absence of entanglement. Our conclusions, however, are specific to this setting and do not imply that entanglement is unnecessary for quantum optimization in general.
High Energy Physics (Theory)
Black holes and phase transitions in Conformal Gravity
We explore the thermodynamic properties of static black holes in conformal gravities in arbitrary, even dimensions. In the spherically symmetric case, up to local Weyl rescalings, these spacetimes are static and unique, due to a Birkhoff's theorem that applies for all conformal gravities whose Lagrangians are complete contractions of $k$ Weyl tensors in dimension $D=2k$. The solutions can be chosen to be asymptotically locally (A)dS or flat, and the horizon can be extended to be a surface of arbitrary constant curvature, $\gamma$. We focus on the asymptotically AdS case, for which the cosmological constant at infinity is an integration constant, and compute the mass of the spacetime in terms of the diverse hair parameters which have a purely gravitational origin. We show that the first law of thermodynamics is satisfied, and study phase transitions in the canonical ensemble, showing the existence of zeroth-order black-hole/black-hole phase transitions for arbitrary dimension.
Read the full abstract
Econometrics
Conditional-Moment Estimation and Inference in the BLP Model
The random-coefficient demand model of Berry, Levinsohn, and Pakes (1995) is commonly estimated by the generalized method of moments (GMM), using an unconditional moment restriction with a fixed set of instruments. Identification of the model, however, rests on a conditional moment restriction. The two are not equivalent: the unconditional restriction may admit additional parameter values. We construct a counterexample in which the model is identified by the conditional restriction yet standard GMM is not, even with the optimal instrument. Building directly on the identifying restriction, we propose a two-step estimator, following Ai and Chen (2003), that first estimates the relevant conditional expectations nonparametrically and then selects the structural parameters by a conditional-variance-weighted minimum-distance criterion; standard GMM is recovered as the special case of a linear projection onto finitely many instruments. We establish root-T asymptotic normality for the proposed estimator, and we develop the theory for both kernel and series implementations of the first stage. The two implementations share a common limiting distribution, attaining the semiparametric efficiency bound. Simulation evidence illustrates the consequences of the identification gap and demonstrates that the proposed estimator outperforms standard GMM in finite samples.
Read the full abstract
Signal Processing
Transforming harmonic coefficients for 3D splat compression
We address the problem of color attribute compression for 3D splats. We show that all images generated by 3D splats are linear in the coefficients for each color channel, each spherical harmonic, and each splat, and we identify a basis for the space of all such images. We identify an inner product for the coefficient space that induces the squared error loss on images. We show that orthonormalizing the coefficients with respect to this innner product before coding can yield over 2 dB gain.
Read the full abstract
Materials Science
Twin Domains in Van der Waals Quaternary Oxides
Optical anisotropy is the basis for many intriguing phenomena in van der Waals materials, including hyperbolic polaritons and extreme birefringence. Stacking and twisting van der Waals materials along the out-of-plane direction emerged as a powerful route to tailor this anisotropy, but designing lateral interfaces remains a challenge. Here, twin domains are reported in the van der Waals quaternary oxides MgTeMoO$_6$, MnTeMoO$_6$, ZnTeMoO$_6$, and CoTeMoO$_6$ - materials that possess strong in-plane optical anisotropy and second-order nonlinearity. The domains naturally form in their orthorhombic crystal structure and extend over hundreds of micrometers. It is proposed that this stability is achieved by the domain wall acting as a diagonal mirror plane in the crystal structure, parallel to the (1-10) or (110) crystal planes, resulting in nearly opposite birefringence between domains. This hypothesis is experimentally confirmed by determining the angle between the crystal axes of neighboring domains using polarization-resolved optical microscopy, infrared-visible sum-frequency generation microscopy, and transmission electron microscopy. The latter further allowed an estimate of the domain wall thickness. Overall, the observation of twin domains with orthogonal optical anisotropy opens new routes to use van der Waals quaternary oxides for birefringent waveguiding, polariton steering, and frequency conversion applications.
Read the full abstract
Group Theory
Algebraic and measurable embeddings between square-free right-angled Artin groups
Let $\Gamma,\Delta$ be finite simplicial graphs, and assume that $\Gamma$ has no induced squares. We prove that the right-angled Artin group $A(\Delta)$ measurably embeds into $A(\Gamma)$ if and only if $A(\Delta)$ embeds as a subgroup in a graph product of free abelian groups over $\Gamma$. This in turn has a graph-theoretical characterisation which can be checked algorithmically. If additionally $A(\Gamma)$ and $A(\Delta)$ have cohomological dimension equal to two and $A(\Gamma)$ is not isomorphic to $\mathbb{Z}\times F_n$, we get that $A(\Delta)$ measurably embeds into $A(\Gamma)$ if and only if it embeds as a subgroup in $A(\Gamma)$. Our proof relies on the following algebraic statement. Let $K_1,\dots,K_n$ be a family of pairwise disjoint cliques in the square-free graph $\Gamma$, and let $g_1,\dots,g_n$ be elements of $A(\Gamma)$ with respective parabolic supports $A(K_1),\dots,A(K_n)$. Then the subgroup of $A(\Gamma)$ generated by $g_1,\dots,g_n$ is a right-angled Artin group, where the only relations impose that $g_i$ and $g_j$ commute if $K_i\cup K_j$ is contained in a clique of $\Gamma$. The same holds true more generally for pairwise disjoint cliques $K_1,\dots,K_n$ in the extension graph $\Gamma^{\mathrm{ext}}$, up to possibly replacing the elements $g_i$ by powers.
Read the full abstract
Mesoscale and Nanoscale Physics
A self-compliant selector-free memory with functional bismuth ferrite
In-memory computing requires selector-free operation, low variability, and stable multilevel switching, all of which remain challenging to achieve in a single system. Here, we demonstrate a robust selector-free memory based on a multiferroic BiFeO3 that combines self-compliance with stable bipolar switching, high yield, long endurance, and extended retention. We observe that BiFeO3 memristive crossbar arrays become self-compliant through a naturally formed TiOx interfacial layer, enabling forming-free bipolar switching without external selector devices, addressing major bottlenecks that have limited the practical deployment of resistive memories. The array delivers stable bipolar switching with a 97% device yield, endurance beyond 105 cycles, and data retention exceeding 105 s with low cycle-to-cycle and device-to-device variability. In addition, the devices exhibit 16 well-separated conductance states, enabling 4-bit multilevel operation via precise pulse-controlled programming for digit visualization in edge applications. Noise measurements reveal predominantly 1/f-type behavior with minimal random telegraph noise, indicating robust conductance stability across the programmable window. These results establish BiFeO3 as a practical materials platform for multibit, selector-free crossbar memories offering high reliability, making it well-suited for scalable computing architectures based on functional materials.
Read the full abstract
Probability
Phase transition for the smallest eigenvalue of high-dimensional sample correlation matrices
We study the smallest nonzero eigenvalue of the sample correlation matrix $\mathbf{R}_n$ formed from a $p_n \times n$ data matrix with i.i.d. real entries $\xi$ of mean zero and unit variance, in the high-dimensional regime $p_n / n \to \phi \in (0, \infty) \setminus \{1\}$. In the tall regime $\phi > 1$, we prove almost-sure convergence of $\lambda_n (\mathbf{R}_n)$ to the lower Marčenko–Pastur edge $\lambda_- = (1 - \sqrt{\phi})^2$ without additional moment assumptions. In the wide regime $\phi < 1$, we establish a phase transition at the third-order tail scale. If $t^3 \mathbb{P}\{\lvert \xi \rvert > t\} \to 0$ as $t \to \infty$, the smallest eigenvalue $\lambda_{p_n} (\mathbf{R}_n)$ converges in probability to $\lambda_-$. While if $t^3 \mathbb{P}\{\lvert \xi \rvert > t\} \to \infty$, then $\lambda_{p_n} (\mathbf{R}_n)$ converges in probability to zero. At the critical scale $t^3 \mathbb{P}\{\lvert \xi \rvert > t\} \to \kappa \in (0, \infty)$, the point process of eigenvalues in the lower gap $(0, \lambda_-)$ converges in distribution to a Poisson random measure with explicit intensity. In this critical regime, we also identify the nondegenerate limiting distribution of $\lambda_{p_n} (\mathbf{R}_n)$, which has a continuous density on $(0, \lambda_-)$ and a positive atom at $\lambda_-$.
Read the full abstract
General Relativity and Quantum Cosmology
Perturbations of a Schwarzschild black hole and Newman-Unti gauge
In this paper, we derive the complete transformations of a generic first order perturbative metric in Schwarzschild spacetime to the Newman-Unti (NU) gauge in series expansion near null infinity. This allows us to determine the asymptotic shear, the mass aspect and the angular momentum aspect of the perturbative fields. As a direct application, we derive the corresponding asymptotic NU data for quasinormal modes (QNMs) of a Schwarzschild black hole. The total energy and angular momentum of QNMs obtained from the asymptotic NU data vanish but classical supertranslation charges are non-trivial, which can be applied to fix the BMS frame at first perturbative order.
Read the full abstract
Representation Theory
Virasoro extensions of generalized Kac-Moody algebras
Borcherds-Kac-Moody (BKM) Lie algebras are constructed by combining a number of $\mathfrak{sl}_2$-s and $3$-dimensional Heisenberg algebras in a certain way, generalizing the celebrated Kac-Moody algebras. Such algebras have a wide range of applications, including the examples of the monster and fake monster Lie algebras being BKM algebras. Among their main features are existence of an invariant bilinear form, denominator identity and Weyl-Kac character formulas for irreducible representations. We introduce a construction of closely related algebras that we call Virasoro-Borcherds-Kac-Moody (VBKM) algebras: if it happens that certain $3$-dimensional Heisenberg algebras inside a BKM algebra form the infinite-dimensional Heisenberg algebra, we replace it by the Virasoro algebra in the construction. We construct a family of Lie algebras over $\mathbb{P}^1$, where the generic fiber is a VBKM algebra and a special fiber is a BKM algebra. This allows us to visualize VBKM algebras as deformations of BKM algebras and prove structural properties about VBKM algebras similar to those of BKM algebras, despite the absence of an invariant bilinear form. Those include denominator identity and character formulas of irreducible representations. We give examples of realizations of such $\mathbb{P}^1$-families of VBKM algebras in lattice vertex operator algebras. In particular, we include the fake monster Lie algebra into such a family.
Read the full abstract
Quantum Physics
A Modular, Topology-Aware Software Stack for Entanglement-Based Distributed Quantum Computing
Distributed quantum computing (DQC) seeks to scale beyond the limits of monolithic processors by interconnecting multiple quantum processing units (QPUs) through entanglement-based links. Realizing this vision requires the co-design of hardware and software across the domains of quantum networking, compilation, and scheduling, yet existing tools remain fragmented between monolithic circuit compilers and long-distance quantum network simulators. We present an open-source, topology-informed framework for the compilation and scheduling of distributed quantum programs. Given an input circuit and a description of the target network, the framework partitions and reconstructs the circuit into a distributed program that respects the specified inter- and intra-QPU topology, covers cross-QPU operations through gate and state teleportation, and schedules the result under either deterministic or stochastic entanglement-generation models. By accepting and emitting standard OpenQASM, the framework interoperates with existing monolithic toolchains, and its standardized module interfaces allow partitioning and scheduling strategies to be interchanged and benchmarked. Using this framework, we show that the best-performing compilation strategy varies across the tested circuits and network configurations, and that both network topology and intra-QPU connectivity substantially affect the entanglement cost of execution. These findings underscore the need for a co-design approach to distributed quantum computing, which our framework is designed to support.
Read the full abstract
Artificial Intelligence
Are LLMs Good Financial User Simulators? Multi-view Investor Logic Alignment (MILA)
Large language models (LLMs) are increasingly used as user simulators, yet it remains unclear whether their predictions faithfully reproduce the evolving decisions of individual users. We investigate this question in a controlled longitudinal paper-trading study with 80 participants, where user interactions, simulated transactions, virtual portfolio states, and point-in-time market information are aligned under a rolling next-day prediction protocol. We evaluate behavioral fidelity hierarchically, from trade occurrence to action structure, asset selection, and downstream portfolio consequences. Across 1,239 aligned user-days, no evaluated LLM reliably outperforms a simple recent-activity persistence baseline for predicting whether a user trades. Fidelity further deteriorates at finer levels: models struggle to recover buy–sell structure and traded assets, and similar activity-level predictions can lead to substantially different portfolio trajectories. Controlled evidence ablations show that recent trading history strongly governs activity prediction, whereas asset selection is substantially more sensitive to the available evidence. An observational analysis further finds that intensified ticker-specific research predicts imminent trading, but diagnostic tests do not support a causal interpretation. These findings suggest that current LLMs capture useful short-term behavioral regularities without yet recovering a stable individual decision mechanism.
Read the full abstract
Robotics
Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Hands
Tactile sensing provides contact information that can be difficult to infer from vision alone, but tactile hardware for dexterous hands has not converged to a common design. Dexterous hands differ in finger structure, contact surfaces, and sensor layouts, while simulated tactile signals still differ from measurements produced by physical sensors. These factors make it difficult to study visuo-tactile manipulation across diverse dexterous hands within a consistent experimental setting. We present Bench2Dex, a simulation benchmark for visuo-tactile bimanual manipulation across 12 dexterous hands. We adapt existing robot models with a shared simulated tactile interface that converts local contact geometry into image-like tactile observations. The interface provides a consistent observation format across different hand morphologies without attempting to reproduce the output of a specific physical tactile sensor. Bench2Dex includes 26 bimanual manipulation tasks that involve tool use, articulated-object interaction, and multi-stage manipulation, together with about 1.3K human-teleoperated demonstrations. The benchmark provides synchronized visual, tactile, proprioceptive, action, and object-state observations, together with executable task metrics. For robustness, we group seven perturbation types into invariance axis, where the correct action does not change, and equivariance axis, where the correct action changes together with the perturbation. We evaluate ACT, Diffusion Policy, pi0.5, and GR00T N1.5 on Bench2Dex and report their performance and failure modes. Bench2Dex is meant as a platform for studying visuo-tactile learning across dexterous hands. It does not assume that simulated tactile observations can replace real tactile sensing; it offers a shared setting for algorithm development while tactile hardware and simulation models are still evolving.
Read the full abstract
Probability
On the Malliavin calculus on product spaces and an infinite de Jong theorem
We extend the Malliavin theory for $L^2$-functionals on product probability spaces that has recently been developed independently by Decreusefond and Halconruy (2019) and by Duerinckx (2021), by characterizing the domains and investigating the actions of the three Malliavin operators in terms of the infinite Hoeffding decomposition in $L^2$, which we identify as the natural analogue of the famous Wiener-Itô chaos decomposition on Gaussian and Poisson spaces. We further explore the corresponding Ornstein-Uhlenbeck semigroup in terms of its Mehler representation and prove new moment bounds for iterated gradients. As an illustration of the abstract framework, we prove an infinite version of the quantitative de Jong CLT that has recently been proved by G. Peccati and the author (2017) and by the author (2024).
Read the full abstract
Instrumentation and Methods for Astrophysics
Automatic astrometric calibration of low-cost all-sky cameras
All-sky cameras support weather monitoring, amateur astronomy, fireball and space-debris detection, and measurements of light pollution. These applications require reliable astrometric calibration despite imperfect camera alignment and variable observing conditions. We present an open-source procedure that combines a compact fisheye model, an exact rotation to account for camera tilt, and automatic star matching. It can initialise a calibration from a single frame given the observing site coordinates and observation time. We evaluate the procedure on nine cloud-free, moonless frames, obtained with a Raspberry Pi camera and a 1.55 mm lens. Fitting on eight frames and evaluating on the ninth in turn gives a median residual of 0.70 px (95% interval 0.64-0.80 px) across 4172 matches, reduced to 0.60 px by two optional decentering parameters. All nine single-frame calibrations pass the quality criteria and predict positions on other frames with a median residual of 0.77 px. We also apply the clear-sky calibration to 73 frames with varying lunar illumination or cloud cover, obtaining comparable residuals for accepted matches on moonlit frames (0.62 px). These results support reusing a calibration while the camera geometry remains unchanged and enough stars remain visible for reliable matching. To test the generality of the single-frame procedure, we calibrated one night frame from each of 23 different all-sky systems, including fireball-network cameras, observatory sky monitors and amateur stations, with images of 0.3 to 77 megapixels in JPEG, FITS and camera-raw formats. All 23 are calibrated from the site and time alone, with a median fit residual of 0.42 px. Applied to a second frame of the same camera without refitting, the calibrations give a median residual of 0.66 px over 18 systems. The procedure is available as the Python package ascal.
Read the full abstract
Machine Learning
Solving Finite-sum Coupled Compositional Optimization via Multi-block-Single-probe Estimator
Traditional variance reduction methods (e.g., SPIDER, SARAH, STORM) have been extensively investigated for improving the convergence rates of stochastic optimization. These techniques typically maintain a sequence of estimators for a single function (or gradient) across iterations. However, what if we need to track multiple functions, but can only access stochastic samples of $\mathcal{O}(1)$ functions at each iteration? This scenario arises in an important emerging family of finite-sum coupled compositional optimization (FCCO) problems of the form $\frac{1}{m}\sum_{i=1}^m f_i(g_i(\mathbf{w}))$, where each $g_i$ is accessible only through a stochastic oracle. The key challenge is to track $\mathbf g(\mathbf{w})=(g_1(\mathbf{w}), \ldots, g_m(\mathbf{w}))$ over time, where $\mathbf g(\mathbf{w})$ has $m$ blocks but only $\mathcal{O}(1)$ blocks can be probed for their stochastic values at each step. To address this challenge, we propose a novel Multi-block-Single-probe Variance Reduction (MSVR) estimator to efficiently trace $\mathbf g(\mathbf{w})$ under partial block sampling. Building on the MSVR estimator, we develop several algorithms for FCCO problems, achieving improved sample complexities for non-convex, convex, strongly convex, and Polyak-Łojasiewicz (PL) objectives. We further obtain an improved dependence on $m$ when the outer function gradients $\nabla f_i$ are linear. Empirical studies on multi-task deep AUC maximization further demonstrate the superior performance of the proposed estimators.
Read the full abstract
Artificial Intelligence
Data storytelling meets interpretable machine learning: Decoding AI decisions for non-experts without revealing sensitive data and model details
AI-driven automated decision-making requires both predictive performance and interpretability. Recent advances in interpretable machine learning (IML) provide tools for explaining model predictions, but the technical complexity of these explanations may hinder accessibility to non-experts. To address this challenge, this study integrates data storytelling with IML to enhance the explainability of AI-generated decisions for a broader audience. Following the design science research (DSR) paradigm, this study proposes a formal definition of data storytelling in IML, introduces the DIST Pyramid to align data storytelling with IML, and presents the I-P-O Model to describe their interactions. It further develops an architecture to explain AI decisions through distinct "What-if" and "Why-not" event-generation processes. The architecture also employs data desensitization to protect sensitive input data. To validate the approach, a case study is conducted with the Boston Housing dataset, using SHapley Additive exPlanations (SHAP) values and large language models (LLMs) to generate data stories with And-But-Therefore (ABT) structures. An empirical evaluation shows that 76.4% and 74.3% of respondents rated the "What-if" and "Why-not" data stories as more comprehensible, with significantly higher accessibility scores than traditional SHAP visualizations. The paper concludes with the presentation of a narrative interpretation framework that integrates IML and data storytelling, thereby expanding the research scope as well as the practical applicability of AI decision-making.
Read the full abstract
Machine Learning
Assembling the CREW: A Collaborative Multi-agent Reinforcement Learning Framework for Automated Related Work Generation
Automatic Related Work Generation (RWG) significantly reduces the human time and effort required to author the Related Work Section (RWS) of a research paper. However, prior methods leveraging multi-agent Large Language Models (LLMs) typically rely on a predefined workflow, where each agent is responsible for a specific step in the entire process. This rigid, static inter-agent coordination limits the adaptive collaboration required to synthesize complex scientific literature. To address this limitation, we propose CREW (Collaborative Reinforcement Learning for Related Work Generation), a novel framework where LLM agents bypass heuristic pipelines to dynamically coordinate by autonomously selecting actions, such as Retrieve, Disseminate, Compose, and Critique, driven by a policy optimized via Independent Proximal Policy Optimization (IPPO). Extensive experiments on a standard RWG benchmark demonstrate that our approach yields substantial quality improvements over strong existing baselines, while significantly reducing token costs. Code is available at https://github.com/YenPBao/CREW-Collaborative-MARL.git
Read the full abstract
Differential Geometry
The stable Bernstein theorem in $\mathbb{R}^{7}$
We give a Green-function proof of the stable Bernstein theorem in $\mathbb R^7$ for smooth, connected, complete, two-sided stable minimal hypersurfaces, thus resolving the last case of the stable Bernstein problem.
High Energy Astrophysical Phenomena
Iron Nuclei with Energy of 10^19 eV in Extragalactic Cosmic Rays Near Earth
The paper discusses the fraction of extragalactic cosmic iron nuclei with energies of 10^19 eV that reached Earth from distant sources without fragmentation in intergalactic space. It is shown that the fraction of particles with energies E = 10^19 eV is significant (about 10%), even if the redshifts of the sources reach 0.2, and not only for nearby sources.
Read the full abstract
Statistical Methodology
Dynamic Policy Evaluation and Learning with Spatio-temporal Data
Although sequential decision-making is ubiquitous across domains, policy evaluation and learning with spatio-temporal data remain challenging due to spatial spillover and temporal carryover effects. We develop methods for evaluating and learning individualized dynamic policies under spatio-temporal interference. Under a semiparametric additive outcome model that allows for complex spillover and carryover effects, we consider a family of stabilized estimators for evaluating the performance of a given individualized policy. From this family, we select a data-adaptive optimal estimator that minimizes the asymptotic variance. We then derive the asymptotic distribution of the proposed policy evaluation estimator, and establish the finite-sample regret bounds of our policy learning estimator. We further propose a statistical test to select the complexity of the semiparametric additive model by determining the appropriate order of interactions. Through simulations, we assess the finite-sample performance of our estimators and the validity of the proposed test. Our motivating application examines the optimal allocation of economic aid in Iraq from February 2007 to July 2008. Drawing on declassified conflict data, we study the weekly assignment of aid projects across districts and find that reallocating aid away from regions with a persistently high level of violence can substantially reduce insurgent attacks.
Read the full abstract
Materials Science
The Case Against Hall-Petch Hardening in High Entropy Carbide Ceramics
Grain size is often used with the Hall-Petch relationship to justify differences in hardness in ceramic materials. Herein, hardness does not vary systematically with grain-size for fully dense, single-phase (Cr,Mo,Ta,V,W)C1-{\delta} high entropy carbide ceramics for grain sizes that varied by a factor of three. Fully dense, single-phase ceramics with grain sizes from 9.3+/-0.3 to 28.8+/-0.7 microns exhibited a pronounced indentation size effect, with Vickers hardness decreasing from ~28-30 GPa at 0.49 N to ~20-21 GPa at 9.81 N, and Berkovich nanohardness ranging from 26 to 30 GPa at 10 mN. However, at a given load, hardness remained within a narrow range across the grain-size series, and no consistent Hall-Petch dependence was resolved. The lack of grain-size dependence likely indicates that the deformation volume sampled by the indenter was not controlled by grain-boundary interactions; instead, hardness was governed primarily by indentation load and local response of the rock salt carbide matrix.
Read the full abstract
Robotics
Continuous Manifold-Decomposed Impedance Retargeting for Contact-Rich Imitation Learning
CMDIR extends Manifold-Decomposed Impedance Retargeting (MDIR) to transform fixed-impedance demonstrations into continuous variable-impedance controllers, which can also serve as structured supervision for imitation learning. Continuous Task-Manifold Impedance Representation (TMIR) pairs an evolving task frame with controller instructions. Demo-relative Compromise dynamics retain moving-basis transport and control/physical metric mismatch, yielding displacement, reaction-impulse, and perturbation-sensitivity criteria. Quality-to-Fast automatically compiles a solver structure from development paths within a predefined finite space, re-instantiates that structure for each demonstration, and certifies the resulting candidate by multi-resolution evaluation. Across 225 retargeted-controller trials in three real contact tasks, full CMDIR improves mean task-proxy retention and reduces mean pose deviation, force fluctuation, and peak force relative to discrete MDIR. FastMPO achieves a $5.8$–$9.4\times$ speedup over C-MPO with comparable closed-loop outcomes. Downstream experiments demonstrate learnability of the complete TMIR supervision interface; lower force fluctuation and peak force are observed among successful executions, while completion reliability remains uneven across tasks and environments.
Read the full abstract
Combinatorics
Improved Bounds for the Bilu-Linial Conjecture via Spectral Recovery from Mixed Determinantal Polynomials
Bilu and Linial conjectured that every finite $d$-regular graph with $d \geq 2$ admits an edge signing $\sigma$ whose signed adjacency matrix $A_\sigma$ has spectral radius at most $2\sqrt{d-1}$. We prove that every signing meeting the mixed-root condition $r_{A_\sigma}\leq\sqrt{2(d-1)}$ satisfies \[ \rho(A_\sigma) < \frac{3+\sqrt5}{2}\sqrt{d-1}, \] where $r_{A_\sigma}$ is the largest root of the mixed determinantal polynomial $\chi[A_\sigma,-A_\sigma]$. The interlacing theorem of Ravichandran and Srivastava guarantees a signing satisfying the mixed-root condition, so our result improves the coefficient $2\sqrt2$ in their two-sided spectral bound. In the proof, we construct a positive matrix-valued probability measure supported on the roots of $\chi[A_{\sigma},-A_{\sigma}]$. The second moment gives a simple matrix inequality $A_{\sigma}^2 + dI \preceq 4r_{A_\sigma}^2I$, which yields a preliminary coefficient $\sqrt{7}$. Estimates for the fourth moment use information about short walks to obtain the coefficient $(3+\sqrt{5})/2$. For triangle-free graphs, the coefficient improves to $\sqrt6$, and for graphs with girth at least five, it improves to $\sqrt{35/6}$. As a result of independent interest, we extend the construction to $\chi[A_1,\ldots,A_k]$ for Hermitian matrices $A_1,\ldots,A_k$ with zero diagonal, and compute the first two moments explicitly. Finally, an explicit signing of $K_8$ shows that the mixed-root condition alone cannot guarantee a coefficient below $(4+\sqrt5)/\sqrt6$.
Read the full abstract
High Energy Physics (Phenomenology)
Hadrophilic inelastic freeze-in dark matter in $q_1-q_2$ gauge extension and the high energy LZ event
The recent observation of a single event at nuclear recoil energy $248\pm23\pm23$ keV by the LZ collaboration has prompted dark matter (DM) model builders to explore new physics explanations. While most of the existing literature invokes an endothermic scattering by a TeV-scale DM to explain the event, the favored DM interaction strength is in tension with the indirect detection constraints for most of the models. To evade this, this work explores exothermic DM scattering in a hadrophilic $q_1-q_2$ gauge extension setup with a new $Z'$ boson as a mediator to the dark sector. This can naturally escape the indirect searches in the absence of tree-level coupling of the mediator to neutrinos, $W,\tau$, and $b$ quark. We consider freeze-in production at a low reheating temperature to populate DM in the early universe, with a comparatively smaller coupling than the thermal freeze-out scenario. We identify the $Z'$ parameter space that can explain the observed relic density and LZ event simultaneously. We also discuss the relevant constraints coming from flavor mixing and FCNC.
Read the full abstract
Machine Learning
Knowledge-Enriched Structured EHR Features for 30-Day Hospital Readmission Prediction on MIMIC-IV
Recent approaches to 30-day hospital readmission prediction rely on pre-trained language models applied to discharge summaries. Although these methods achieve strong performance, they depend on the availability of clinical notes, incur substantial computational costs, and yield representations that lack interpretability. We propose a knowledge-enriched feature representation that augments structured Electronic Health Record (EHR) data with four medical knowledge sources: disease ontology mapping, procedure classification, drug ingredient vocabulary, and organ system laboratory aggregation, without using clinical notes. Each feature dimension corresponds to a named clinical concept, yielding a sparse and interpretable patient representation. The approach is evaluated with six classifiers on a MIMIC-IV v2.2 cohort. Under 20-fold cross-validation, the best configuration achieves an AUROC of 0.743. This performance is comparable to that of previously reported methods on this dataset, including both those using only structured data and those incorporating clinical notes, while requiring considerably less computational cost. Interpretability analysis shows that demographics, organ system labs, drug ingredient features, and first-level ontology disease categories drive prediction, while deeper hierarchy levels contribute negligibly. These findings indicate that knowledge-enriched structured features offer a competitive and efficient alternative to embeddings from clinical notes for 30-day readmission prediction.
Read the full abstract
Superconductivity
Wrinkles and Magnetic Flux Trapping in Graphite Nanoflakes: A Possible Source and Manifestation of Room-Temperature Superconductivity
The paper reports on a possible manifestation of local high-temperature superconductivity in graphite nanoflakes obtained from the bulk pyrolytic graphite by an extended grinding and subsequent annealing at 670 K in the air. Analysis of transmission electron microscopy (TEM) images shows that such treatment leads to a formation of high-density wrinkle-type defects at the basal-plane surface. The sample reveals the magnetic flux trapping - one of the key signatures of superconductivity - that persists up to 390 K and above. At the same time, both the as-ground (unannealed) and annealed in the vacuum samples manifest negligible flux trapping, and no wrinkles in their TEM images were found. The direct correlation between the appearance of high-density wrinkle arrays and the detection of the trapped magnetic flux above the room temperature strongly suggests that wrinkled regions serve a source of local high-temperature superconductivity in graphite.
Read the full abstract
Systems and Control
Model-Free PID Tuning by Step-Response Inspection
Industrial PID loops are still tuned by hand: step the setpoint, look at the response, change a gain. This paper turns that procedure into an algorithm, SPIN, for Step-response Phase-portrait INspection. From a single closed-loop step response we form three phase portraits of the control deviation and count how many times each trajectory winds around its settling point. The counts are dimensionless and independent of the step size and of the clock. They show which channel is under-damped: integral, proportional, or derivative. A triangular rule reads the three counts, cutting the gain of the one band whose fault is certain and raising that of every band the same test shows to be within its limit. No model is identified and no optimization is solved. The iteration is deterministic and its gains stay inside a fixed box. Because a ringing count is itself the evidence that the loop is in the regime where cutting that band's gain reduces the count, the iteration settles into a cycle at the edge of the well-damped set. It needs no excitation beyond a routine setpoint step, so it can run on a process in operation, and it may start from arbitrary gains, including destabilizing ones. SPIN assumes a stable, self-regulating process whose gain and phase both fall with frequency, which covers the lag-plus-dead-time models common in process control, and a controller from the nested family $\mathrm{I} \subset \mathrm{PI} \subset \mathrm{PID}$ with integral action present. We validate it on a battery of process-typical plants with one fixed step size and one set of dimensionless constants throughout. Every decision is visible: the operator sees what the algorithm sees. A reference implementation and an interactive demonstration are public.
Read the full abstract
Artificial Intelligence
Predicting build orientation for SLM dental parts: a comparison of rotation representations and direct vector regression
Build orientation for selective laser melting (SLM) manufacturing of dental parts is usually chosen manually by technicians. We treat orientation prediction as supervised machine learning of the part's up-axis from technician-labeled production data, and test which rotation representations produce the best results. Using $n\approx2400$ patient-specific dental parts, we trained a ResNet-50 multi-view image backbone and a PointNeXt-S point-cloud backbone, both pretrained and fine-tuned end-to-end, on 13 up-axis representations spanning six classical $SO(3)$ parameterizations and seven representations defined directly on the unit sphere $S^2$. We report the geodesic angular error between predicted and ground-truth up-axis on a test set, with and without test-time augmentation (TTA) over $K=21$ known rotations. With TTA, the octahedral map achieves the lowest mean angular error ($10.6^\circ$, ResNet-50). The three lowest-error results overall are direct $S^2$ representations, though this may reflect label noise in the unsupervised in-plane component of the $SO(3)$ targets rather than a topological advantage. von Mises-Fisher collapses to a near-constant prediction when trained with PointNeXt-S but not with ResNet-50. TTA reduces mean angular error by 31-73 % across almost every representation and backbone. Overall, test-time augmentation over a small set of known rotations is the most consistent driver of accuracy, whereas the best-performing representation is strongly backbone-dependent.
Read the full abstract
Information Theory
The Capacity of the Relay Channel
The capacity of the relay channel is characterized by three schemes: decode-forward, coordinated compress-forward, and uncoordinated compress-forward. The proof relies entirely on properties of typical sequences.