Papers

Simulations of water at solid interfaces routinely disagree on basic properties, from contact angle to friction, depending on the model used, whether an empirical force field or a density functional theory (DFT) exchange-correlation (XC) approximation. Yet no accurate computational reference has been available to determine which, if any, is reliable. Here we assess XC approximations spanning the rungs of Jacob's Ladder against the random-phase approximation (RPA) at the graphene-water interface. Using machine-learned potentials (MLPs) trained on both DFT and RPA data, we achieve the extensive sampling required to characterize interfacial structure, wettability, friction, and vibrational sum-frequency generation spectra. We find that XC approximations disagree substantially with RPA and with each other. To make sense of this, we introduce a transferable, multi-observable scoring framework that distills each functional's agreement with RPA into a single measure. Six functionals emerge as reliable starting points for studies of graphene-water and other dispersion-dominated interfaces, namely B3LYP-D3(0), revPBE-D3(0), revPBE-D3(BJ), B97M-rV, r$^{2}$SCAN, and revPBE0-D3(0). Notably, this ranking follows no simple rule based on functional family or rung of Jacob's Ladder. What does hold across functionals is that dynamical and spectroscopic properties tend to be harder for DFT to reproduce than structural ones, with direct consequences for functional selection. Our framework also offers a general strategy for benchmarking any atomistic model, including the new generation of foundational MLPs, against high-accuracy references at solid–liquid interfaces. Beyond simulation, this work identifies which models can be trusted to connect measured contact angles, friction, and vibrational spectra to the molecular structure behind them.

Test-time scaling (TTS) improves the reasoning capabilities of large language models by allocating additional inference computation. Existing approaches to improving TTS efficiency largely optimize accuracy against one resource dimension at a time, advancing either the accuracy–cost or accuracy–latency Pareto frontier. Yet user requirements are multidimensional: users may specify accuracy, latency, and inference-cost requirements jointly, and different requirements can favor different controllers. We formulate Personalized Test-Time Scaling as discovering executable controllers that maximize the joint satisfaction rate of user-specific requirements. To reduce the overhead of repeated policy discovery for new user profiles, we propose PersonTTS, an amortized agentic policy-discovery framework that reuses prior search experience through requirement-matched controller initialization and source-distilled procedural guidance, while retaining target-profile evaluation for every candidate. Experiments on AIME and HMMT show that PersonTTS substantially outperforms strong TTS baselines in joint requirement satisfaction on unseen user profiles and held-out problems. Under the same candidate-evaluation budget, cross-user experience reuse further improves policy quality while substantially reducing discovery-agent time and cost.

Dual-process agents pair a fast policy with a slow deliberative model. In real-time settings the slow model usually runs continuously; in turn-based agents and robot planners it is invoked on events such as uncertainty or a detected failure. We study a fast learned actor that takes every decision and hands control to a reasoning vision-language model only when a gate opens, while the game keeps running. We use closed-loop Doom and the new open "System One" typed-decision models, served through a common llama.cpp interface. On 900 held-out questions, (i) zero-shot decision models from 0.15B to 9B parameters choose to collect items 1.6-1.8 times more often than chance among their errors, in any option order, although the order changes some models' accuracy; (ii) accuracy, calibration and sensitivity (how well confidence separates right from wrong answers) are distinct: models of similar accuracy differ widely in AUROC, and the confidence of the most sensitive one tracks which kinds of situation it fails, not which answers are wrong; (iii) offline, deferring the least confident 30% of decisions to a reasoning model gains over random deferral in proportion to the actor's AUROC (rank correlation 0.87); with actor and rate chosen on held-out games the gain is +0.13 [0.08, 0.18] with doomLaya's option order and +0.08 [0.02, 0.14] with shuffled options, and reasoning carries about half of it; (iv) in closed loop (33 games, three seeds) no variant reaches the exit. Committing to plans, the reasoner's or a fixed explore rule's, opens more doors and makes an actor that stands still play; with the rule the agent dies more often. Told that some doors need keys, the reasoner takes ordinary doors for locked ones, which the state cannot tell apart; without that knowledge it goes back to collecting. We release code, prompts, data and logs.

Broadband in-air sonar sensors with a sparse microphone array form images through delay-and-sum beamforming, which leaves high sidelobes and an angular resolution that degrades strongly towards the sides of the field of view. Deconvolving these images requires access to the point-spread function of the complete imaging chain, including the emitter spectrum and directivity and the array imperfections. In practice, these aspects are hard to calibrate for. In this letter, we propose a self-supervised method that learns a deconvolution network together with a physics-structured model of the point-spread function, requiring only the sonar data itself and the odometry of a moving robot on which the sensor is mounted. Using acoustic flow warping, the internal representation of the model is warped and subsequently rendered through the learned point-spread function. This is then compared to the measured images under a speckle likelihood in accordance with the statistical model of image formation. In simulation, the learned model recovers the unknown emitter directivity, and the deconvolved images reach a constant angular resolution of 5 to 6 degrees over the full field of view, against 12 to 30 degrees for delay-and-sum and 7 to 23 degrees for delay-multiply-and-sum with coherence factor. The method outperforms classical deconvolution with the nominal point-spread function, and even outperforms CLEAN with the true one.

Program verification establishes software correctness through machine-checkable proofs constructed in theorem provers. It's a guarantee especially valuable for code generated by large language models (LLMs), which is fluent but carries no assurance of correctness. Almost all existing provers, however, pursue pass rates alone at whatever sampling or search budget it takes, and overlook the success-vs-cost frontier; yet real software often carries hundreds of interdependent proof obligations, so what matters at scale is not whether one theorem can be proved, but how many can be proved economically. We introduce CoCo-Prover, which formalizes cost-efficient program proving as metalevel decision-making under cost, grounded on two-level proof graphs: an AND/OR proof hypergraph within each declaration is joined to a lemma-dependency graph across declarations; and at each step, it answers two questions: which open goals to select, and which actions to purchase on these goals. Selection stays symbolic as a topological pass over the proof graphs. Action choice is agent orchestration via metalevel decision-making: an agentic router treats every bounded specialist invocation as a separately priced, best-effort computation, matching heterogeneous specialist agents together with configurations, under evolved routing rules as evidence accumulates. On five program verification benchmarks in Lean 4 including function-level CLEVER, VERINA, and AlgoVeri, and repository-level NTP4VC and Vero, we show that CoCo-Prover achieves a better success-vs-cost frontier than baselines including frontier coding agents and state-of-the-art LLM-based provers: it achieves the best solve rate on every benchmark and up to 100% on two benchmarks. It also reduces cost by up to 30.9% compared to the strongest baseline with the strongest LLM in our evaluation.

We present initial-state corrections to $e^+\,e^-\to\mu^+\,\mu^-\,\gamma$ at next-to-next-to-leading-order in QED for several experimentally motivated scenarios up to a few GeV. We include photonic as well as vacuum-polarisation corrections into the Monte Carlo framework McMule that allows for the calculation of arbitrary infrared-safe observables. For the total cross sections, the corrections are typically below the percent level, but they can be locally larger in distributions. The vacuum-polarisation corrections are of similar size and must not be neglected, in particular for larger centre-of-mass energies.

Adaptive rollout methods for group-relative reinforcement learning typically allocate a fixed per-update budget across prompts. We instead study how to coordinate a finite rollout budget over the entire training horizon. We formulate this problem using a concave surrogate utility of cumulative prompt exposure and introduce CERO, an online primal dual scheduler for prompt admission and budget pacing. In our experiments, each admitted prompt receives a fixed-size response group. CERO instead adapts which prompts are selected, how often they are revisited across rounds, and how many groups are generated in each round. A compact Fenchel representation linearizes the dependence on cumulative exposure, while projected online gradient descent updates prompt-specific supporting slopes and a shared budget price using reward-variation feedback and budget deviations. We establish pathwise guarantees for the surrogate allocation objective against fixed-rate and same-path time-varying benchmarks, with explicit terms for proxy discrepancy and rate variation. Under matched training-response budgets, CERO attains the highest avg@16 macro-average on each of three backbones across five mathematical reasoning benchmarks. Mechanistic analyses link CERO's prompt choices to within-group reward contrast, while multi-seed ablations show gains from adaptive pacing over both uniform and preset spending schedules.

A KMS-symmetric quantum Markov semigroup is completely determined by the quadratic form in induces on the GNS Hilbert space. The quadratic forms that arise in this way are called quantum Dirichlet forms. In this article we give a complete description of bounded quantum Dirichlet forms in terms of derivations. In particular, we exhibit an explicit characterization of the bounded quantum Dirichlet forms on type I factors. We also show that any unbounded quantum Dirichlet form can be represented in terms of derivations. The key technical insight is that the modular group always restricts to a strongly continuous group on the domain of a quantum Dirichlet form such that the square root of the analytic generator is accretive with respect to the form norm.

Self-supervised anomaly detection is a promising paradigm for medical ultrasound, as normal images are often easier to obtain than exhaustive annotations of all possible pathologies. However, most existing evaluations are limited to a single anatomy or task, making it unclear whether models learn a robust notion of normal ultrasound appearance or only a source-specific representation. We introduce the SADUSI benchmark, a multi-source ultrasound dataset designed to train and evaluate anomaly detection methods across a broad range of anatomical regions, views, and acquisition protocols. The goal of SADUSI is to provide a diverse normal ultrasound distribution and a benchmark for visible structural anomalies that can be assessed from single images. We evaluate representative self-supervised anomaly detection methods and find that current approaches struggle in this setting. In particular, reconstruction-based diffusion methods such as AnoDDPM and DeCo-Diff achieve pixel-level AUROC values of 0.56-0.72 and maximum F1 scores of 0.10-0.26, indicating limited separation of pathology from normal image regions. Feature-based PatchCore variants perform better, reaching pixel-level AUROC values of 0.76-0.83, but remain limited with maximum F1 scores of 0.14-0.40. These findings suggest that broad multi-source ultrasound anomaly detection remains an open challenge and that SADUSI can serve as a resource for developing methods that generalize beyond anatomy-specific settings.

IceCube detections of TeV neutrinos from nearby Seyfert galaxies imply, if confirmed, that the accompanying gamma rays are substantially reprocessed down to the MeV range by electromagnetic cascades, which places multi-messenger constraints on neutrino emission models. However, gamma rays can also be produced by the primary nonthermal electrons of the X-ray corona, as recent radiative kinetic simulations of magnetized turbulence have shown. To disentangle the leptonic and hadronic contributions, we self-consistently combine results from these simulations with predictions for hadronically induced electromagnetic cascades. Focusing on NGC 4151, we find that the leptonic channel dominates the MeV band, and remains comparable to the hadronic cascade up to $\lesssim$0.1 GeV, when the nonthermal electron tail carries $\gtrsim5\%$ of the electron energy. The leptonic contribution may be detectable in soft gamma rays by the upcoming Compton Spectrometer and Imager (COSI) mission, whereas our predicted hadronic component falls below its sensitivity. We also find that the same nonthermal electron population can account for both the $0.1-1$ MeV excess of the cosmic X-ray background, generally attributed to AGNs, and the MeV tail of Cygnus~X-1 in the hard state, lending observational support to our model. Our results offer a refined view of multi-messenger signals from black-hole coronae, grounded in their nonthermal kinetic physics.

We study the accuracy of Gauss-Newton curvature in ridge-regularized nonlinear least squares. Under local regularity and persistence of level-set curvature magnitude along an exact-fit section, we prove uniform coexistence of two curvature regimes. Global minimizers exist, and every global minimizer has relative Hessian error below $(1+\sqrt2)/8$, while the same low-cost set contains a point with an indefinite Hessian and relative error at least $15/8$. One positive ridge cap works for all independent center and label perturbations in fixed neighborhoods and every positive ridge weight up to the cap. These neighborhoods do not shrink as the ridge weight tends to zero. A pointwise certificate based on the current prediction level set controls the normal, mixed, and tangent parts of the Hessian correction. We prove a sharp relative-error bound over the stated pointwise class when the prediction map and ridge vary. Analytic examples describe the roles of output alignment, curvature orientation, and persistence. A separate structural result gives full Jacobian row rank throughout low-cost sets and exact interpolation near a rank-deficient reference.

Real-time equilibrium reconstruction is essential for tokamak plasma control. Data-driven surrogates are usually trained on high-quality labeled experimental data, but producing a large amount of labeled data sufficient for surrogate learning requires many expensive discharge shots, and few exist at the start of a new campaign or the operation of a new device. To reduce this cost, we propose to pre-train on simulated equilibria, generated at low cost by physics-based simulation, then fine-tune on limited real data. We evaluate under three settings with increasing difficulties, i.e., in-distribution, out-of-distribution (shape), and cross-campaign (temporal) splits, which mirror how reconstruction is actually deployed. Experimental results validate that synthetic pre-training is highly effective when few real labels are available: fine-tuning on only 1% of the full available real dataset cuts the per-sample normalized root-mean-square error (nRMSE) of the reconstructed poloidal flux by roughly 58% under out-of-distribution shapes and 64% under cross-campaign extrapolation, compared to a model trained from scratch on the same amount of real data. These realistic evaluations show that synthetic pre-training improves equilibrium reconstruction accuracy in the data-scarce regime that real tokamak campaigns face.

Identifying and measuring topological invariants remains a fundamental challenge due to the non-local nature of momentum eigenstates in real space. Established approaches based on Bloch oscillations, quantum walks, charge pumping and quench dynamics, typically fail to provide direct momentum-dependent geometric information of each Bloch state. Here, we propose a novel framework for implementing direct band-topology tomography via momentum-selective measurements, and experimentally demonstrate it on a superconducting processor featuring a highly connected coupling architecture. Utilizing a “star-like" geometry in which an auxiliary probe couples to a one-dimensional topological chain with spatially modulated coupling strengths, we selectively address the desired momentum states of the chain. This approach enables the direct and robust extraction of the winding phase that contributes to the associated topological invariant. Leveraging non-local couplers bridged by long tantalum airbridges, we realize an extended Su-Schrieffer-Heeger (SSH) model and map its full topological phase diagram. Through a nested Fourier analysis over time and space, we directly extract the winding phase across Brillouin zone and resolve quantized winding numbers of $\nu=\{0,\pm1,2\}$ characterized by phase winding $\{0, \pm 2\pi,4\pi\}$. Our work establishes a versatile tool for probing topological characteristics and opens a new avenue for robust quantum computing architectures based on topological protection.

In [6], we introduced the notion of a Hermitian-relative curve, which is a plane curve defined by $(x^{\sqrt{q}}, y^{\sqrt{q}}, z^{\sqrt{q}})A (x,y,z)^t =0$ with $A \in GL(3, \mathbb{F}_q)$. After investigating their basic properties, we classified those curves with two or more rational inflexions. In this paper, we first complete the classification of curves whose rational points are all inflexions by demonstrating the existence of curves with exactly one rational point. As a continuation, we then classify Hermitian-relative curves that possess at least one rational point which is not an inflexion. We categorize these curves according to the ordered pair consisting of the number of rational points and the number of inflexions. Finally, we enumerate all possible pairs and prove that each pair is realized by a suitable curve.

Recent advances in reasoning-oriented large language models (LLMs) have motivated increasing evaluation of their ability to perform professional decision tasks. Insurance claim adjudication is one such task, requiring models to connect case evidence, insurance rules, intermediate judgments, and payout calculations across a structured decision process. We introduce InsClaimBench, an end-to-end benchmark for evaluating insurance claim adjudication across the decision chain. Grounded in real claim materials and structured insurance rules, InsClaimBench contains 3,780 cases in 375 case families across auto, property, and health insurance, comprising 86,656 atomic rule judgments. It evaluates each claim from atomic rules through adjudication modules to payout decisions and amounts, with controlled factual variants testing whether required changes are correctly propagated across levels. Evaluation of six LLMs reveals a progressive loss of reliability along the decision chain. Payout-decision accuracy ranges from 74.23–80.19%, while joint decision–amount accuracy drops to 47.54–73.15%. Strong local performance also fails to ensure case-level correctness: atomic-rule accuracy reaches 95.48%, whereas rule-vector exact match peaks at only 36.90%, and the most frequent module errors are not necessarily those most associated with final-decision failure. Under factual changes, these inconsistencies further become propagation failures: module updates are less reliable than rule updates, correct local judgments can still yield incorrect payouts, and correct payouts can conceal intermediate errors. These results show that reliable claim adjudication requires consistent composition and propagation across the decision chain.

Unsupervised anomaly detection, which learns only from normal images, is a central task in medical image analysis and remains an open problem. Reconstruction-based methods pass an image through an encoder-decoder network trained on normal data and detect anomalies from the residual between the image and its reconstruction. This works only if the information passed from the encoder to the decoder is limited; otherwise the network learns an identity mapping and reconstructs anomalies too. This limit is usually imposed through architectural choices, tuned per dataset, that cannot be stated in bits. We introduce the quasi-binarizing (QB) layer, which squashes each latent element into [0, 1] and adds Laplace noise of scale 1/epsilon. Each element is then epsilon-locally differentially private, and the mutual information between an image and its reconstruction is bounded by a quantity that depends only on epsilon and the number of QB elements, whatever the encoder and decoder. Placing a QB layer on every encoder-decoder path, including all skip connections, we build QBAE, a seven-level attention U-Net with 32,768 QB elements. On the seven datasets of the MedIAnomaly benchmark, QBAE with one architecture and one configuration reaches a mean image-level AUROC of 0.828, the highest among methods that do not adapt to each dataset, and the best reported results on BraTS2021 (AUROC 0.911, pixel-level AP 0.838). The noise is kept at test time, so that every reconstruction satisfies the bound. Without input corruption, the bottleneck alone prevents identity collapse (mean AUROC 0.805 vs. 0.590). Code is available at https://github.com/hanaokalog/MedIAnomalyQB.

We present a novel dual approach to characterizing the solution to optimal stopping problems for strong Feller processes on locally compact, separable metric spaces. Our characterization builds upon reinterpreting a standard result in optimal stopping theory as a dual formulation, which yields computationally tractable algorithms. Our framework is applicable to a broad class of Markov processes. First, in the one-dimensional case, while previous studies have managed to comprehensively solve only diffusion and one-sided jump Lévy processes over an infinite time horizon, our proposed method successfully extends to both infinite- and finite-horizon scenarios across all these processes, including two-sided jump Lévy processes. In addition, for multi-dimensional cases, our framework is applicable to diffusions over an infinite horizon with certain structural features such as convex stopping regions. To illustrate this versatility, we explicitly solve several challenging problems including a case where the optimal stopping region consists of two disjoint connected components – a structural feature that remains unaddressed in the literature.

The Global Ocean Observing System is key to understanding oceanic variability and climate change. The ocean is vast and often sparsely and irregularly sampled in time and space by various instruments and systems. Statistical models, especially spatio-temporal ones, are useful for enabling inferences, forecasts and decisions from sparse oceanic observations. This article focuses on the statistical treatment of two types of in situ observations in the Global Ocean Observing System: from profiling floats and surface drifters, focusing on the Argo and Global Drifter Programs. We describe the spatio-temporal models that have been developed in recent years for these data, to give a picture of the statistical challenges faced in modern physical oceanography. We will discuss in detail two types of reference frames when modeling such data: Eulerian and Lagrangian, the former of which is more appropriate for profiling floats and the latter for surface drifters.

Random selection is widely used in resource allocation and auditing, making reliable implementation essential for AI-agent systems. Behavioural tests across six reasoning models uncovered threshold and divisibility rules used in identifier-based choices. For threshold-following GPT-6 Sol and Gemini 3.8 Flash, single-agent measurements prospectively predicted correlated participation under shared identifiers and biased participation under distinct identifiers with common timestamp bits. Changing dates, formats and identifier labels revealed when these predictions held. Explicit instructions to randomize independently reduced but did not eliminate shared-input correlation. To test implications for oversight, we asked four models to select customer requests randomly for human review. GPT-6 Sol approached the target rate while selecting predictably from identifiers; the others rarely selected requests. All four closely followed supplied random draws. These findings expose collective and audit vulnerabilities that selection rates alone miss, making input-dependent bias, correlation and predictability central targets for agent evaluation.

Superconducting qubits are often based on the first two energy levels of a physical anharmonic LC oscillator. Since the anharmonicity of such systems is usually much lower than the involved energy gaps, population can easily leak to higher energy states. This is what characterizes leakage errors, and it can have multiple sources: from coherent error induced by the control platform to incoherent thermal dissipation. In this work, we adapt an optimal control method based on sub-Riemannian geodesic search in the unitary group for a system consisting of two superconducting qubits with leakage levels. We use the $R_x(\pi/2)$ rotation for single-qubit operation analysis, and the $i$SWAP operation for two-qubit analysis. For the former, we find considerable performance improvement in average gate fidelity as well as in the reduction of leakage errors when comparing the obtained optimal pulses with Gaussian and DRAG pulses, especially for short duration pulses. For the $i$SWAP gate, although leakage error remains roughly the same, we manage to obtain considerably higher fidelities with the cost of requiring a slightly longer time evolution.

On-policy post-training can improve large language models by learning from their own trajectories, but requires costly rollout generation. We ask whether fixed demonstrations can support competitive off-policy learning through better supervision. Our premise is that their usefulness depends not only on the training trajectories, but also on whether supervision provides informative preferences among continuations and connects this guidance to the reasoning decision being learned. We introduce Step-Aligned Privileged Distillation (SAPD), a rollout-free self-distillation method that turns demonstrations into step-aligned distributional supervision. Its key insight is to use the known progression of a reference solution to associate each reasoning transition with targeted privileged guidance, rather than treating the solution as undifferentiated context. On mathematical reasoning benchmarks, SAPD outperforms supervised fine-tuning and label smoothing on average while remaining competitive with on-policy reinforcement learning and self-distillation. Analyses support both the value of context-dependent distributional guidance and the benefit of aligning privileged information with the current step. SAPD also largely preserves out-of-domain coding performance and achieves approximately 2x training-loop speedups over the on-policy baselines. These findings suggest that carefully constructed supervision can make fully off-policy post-training a competitive and computationally efficient alternative. Our code is available at https://github.com/Miaow-Lab/SAPD.

Altermagnets offer a promising route to spintronic devices by combining spin-split electronic states with zero net magnetization, but how reduced dimensionality reshapes their magnetic order remains poorly understood. Here, we grow bilayer-thickness CrSb (0001) films directly on SrTiO3(001) substrates by molecular beam epitaxy and uncover an unexpected thickness-dependent magnetic transition. The one-bilayer films exhibit room-temperature ferromagnetism with strong perpendicular magnetic anisotropy, whereas ferromagnetism vanishes in two-bilayer films, contrasting sharply with the altermagnetic ground state of bulk CrSb. Density functional theory calculations reveal that the exposed Cr surface induces orbital rearrangement that strengthens intralayer ferromagnetic exchange while suppressing competing interlayer antiferromagnetic coupling, driving the dimensional crossover from altermagnetism to ferromagnetism. These findings establish dimensionality and surface termination as powerful controls of magnetic order and provide a platform for engineering ultrathin magnetic states for spintronic applications.

In marine seismic acquisition, seismic interference (SI) occurs when energy from nearby external seismic source(s) is captured. It typically appears as coherent noise with linear or non-linear movement and varying amplitudes across different sail lines. SI is commonly observed and poses a challenge for seismic data processing. We present a case history of a previously proposed deep neural network (DNN)-based workflow applied for SI attenuation across a marine seismic block in the Camie Field of Angola. This field survey covers over 345 km2 and is marked by the challenge of multiple SI types. The employed DNN-based workflow performs SI attenuation in the common shot domain based on a supervised learning framework: a small subset of the SI-contaminated data was first processed by a conventional geophysical algorithm to obtain an estimate of the SI noise, which was then manually blended with the SI-free common shot gathers from the same survey to generate the training pairs. To ensure signal fidelity, several techniques were applied to improve the DNN's performance. A key highlight of this case history is its scale: this represents a real-world, large-scale processing project and we present a comprehensive comparison of the DNN-based workflow with the conventional geophysical algorithm across the entire survey block, focusing on both processing quality and processing time. The results demonstrate the outstanding performance of the employed DNN-based workflow, which achieved higher SI removal accuracy, with less signal leakage and more complete SI removal. The promising results of this application also open up possibilities for integrating deep learning into other seismic denoising tasks. In addition, we discuss the limitations of this case history, aiming to provide insights for future research and applications in the field.

Hydrodynamic relaxation in open quantum many-body systems can be understood in terms of diffusive Nambu–Goldstone (NG) modes associated with strong-to-weak spontaneous symmetry breaking (SWSSB). Here, we find a common structure underlying weak ergodicity breaking in closed quantum systems and diffusive hydrodynamics in open quantum systems. We establish a spectral correspondence between the parent Hamiltonian of rainbow quantum many-body scar (RQMBS) states in closed systems and the effective Lindbladian of strongly dissipative open systems with strong $U(1)$ symmetry and local charge-dephasing Lindblad operators. For broad classes of spin, fermionic, and bosonic systems, we construct asymptotic RQMBS (ARQMBS) states. Within the enlarged scar subspace, the parent Hamiltonian admits a local Rokhsar–Kivelson-type frustration-free decomposition. A single-mode variational construction gives an $O(k^2)$ upper bound on its lowest excitation energy and yields states satisfying the defining ARQMBS criteria: orthogonality, vanishing energy variance, and characteristic entanglement scaling. Under vectorization, RQMBS states are mapped to maximally mixed infinite-temperature states in fixed-charge sectors, which exhibit long-range SWSSB order under finite-density conditions. Assuming uniformly bounded positive transition rates, the effective Lindbladian has the same local projectors as the parent Hamiltonian but with positive transition-dependent weights, yielding two-sided spectral bounds and the same system-size scaling of their gaps. For uniform rates, the two operators are proportional to each other and their eigenmodes coincide. Therefore, an exact quadratic parent-Hamiltonian branch maps directly to a diffusive NG mode. Our results connect different mechanisms of slow relaxation in closed and open quantum many-body systems.

Automatic Short Answer Scoring (ASAS) is central to NLP for Education. However, openly available benchmarks remain scarce, and existing datasets largely address how well students answer a question directly rather than how well they master underlying concepts (knowledge elements) such as thermal energy or epistemic activities (skills) such as reasoning or claim. To address this gap, we introduce Alice, a large-scale, rubric-based German ASAS dataset that is pedagogically aligned and comprises three subtasks: (i) learning performance (Alice-LP), (ii) knowledge elements (Alice-KE), and (iii) skills (Alice-SK). We further formulate rubric-based ASAS as a rubric-retrieval task and benchmark the dataset with a range of language models, from encoder-only models to lightweight LLMs. We also benchmark the dataset with zero-shot prompting via LLMs and a standard classification baseline. The experiments show that LLMs, in particular, struggle to score knowledge elements and skills in the zero-shot setting. They also indicate that rubric text is often useful, especially for Alice-KE and Alice-SK, while on Alice-LP gains over sample-solution-focused inputs are more modest and vary by model and input format.

Automatic Short Answer Scoring (ASAS) requires models that can score student responses against question-specific criteria while remaining efficient and transferable across rubric sets. We propose RUSPAN, a rubric-conditioned ASAS framework that treats rubric descriptions as semantic label representations. RUSPAN serialises the question context, student answer, and all candidate rubric levels into a single sequence, then scores the levels listwise from the rubric-span and whole-sequence representations produced in a single LM pass. We further introduce RUSPAN-RIM, in which a Rubric-Independent Mask prevents rubric spans from attending to one another, making rubric representations depend only on the answer and question context and preventing overfitting to rubric patterns during training for zero-shot transfer. On six ASAS benchmarks spanning English, German, and Portuguese, RUSPAN improves mono-benchmark scoring over discriminative and generative baselines, while RIM with position reindexing delivers consistent and substantial gains on PT-ASAG, the held-out benchmark with the strongest combined language and rubric-structure shift.

Distributed quantum circuit simulation enables the execution of large-scale quantum algorithms by partitioning qubits across multiple computational nodes. However, inefficient qubit placement frequently leads to excessive inter-node communication, severely limiting performance and scalability. This work introduces a communication-aware qubit layout optimization framework that systematically minimizes data exchange among nodes while dynamically selecting the optimal number of computational resources. Integrated into the QuCSLO library, our approach supports dynamic node scaling, OpenQASM parsing, and seamless integration with the QuEST simulator. Extensive evaluations on both synthetic benchmarks and real-world QASMBench circuits demonstrate that optimized layouts significantly reduce communication cuts and yield runtime speedups of over 2x compared to identity layouts, particularly in highly entangled circuits. Our results underscore the critical impact of strategic qubit placement on distributed simulation efficiency and provide a practical, resource-scalable solution for quantum workloads.

We compare two locked survival prediction models in a target population before target survival outcomes are available. At a prespecified horizon, the estimand is the target Brier-risk contrast. Under a bounded conditional log-odds shift model on a prespecified deployment summary, we derive a sharp identified set preserving the shared unidentified target outcome law. Direct identification is never wider than separately identifying the risks and subtracting their bounds, with strict tightening under a Brier-specific same-side-1/2 condition. For right-censored source data, conditional Cox censoring estimation, inverse-probability-of-censoring-weighted logistic outcome modeling, and a joint pairs bootstrap yield simultaneous confidence envelopes over a finite sensitivity grid. In simulations, separate-to-direct width ratios ranged from 1.00 to 5.73 across controlled prediction geometries. Targeted simulations showed finite-sample undercoverage of the outer envelope at the small, heavily censored non-small-cell lung cancer (NSCLC) information scale (0.847-0.861 versus 0.95 nominal), compared with 0.946 at the Rotterdam-GBSG scale. In the cross-institutional NSCLC application, all 40 prespecified evaluations resulted in DEFER despite reduced identification uncertainty. In a supporting Rotterdam-to-GBSG analysis, candidate superiority was certified under small sensitivity allowances; one locked configuration yielded ADOPT CANDIDATE under direct identification but DEFER under separate-risk subtraction. Direct identification can materially reduce identification uncertainty and change the operational conclusion when signal and sampling precision are sufficient, while retaining DEFER when directional certification is unsupported.

With the ever-growing size of deep learning models, GPU memory is prone to being insufficient during training. A prominent approach is ZeRO-Offload, which moves the optimizer states to CPU memory and performs parameter update using CPU. However, the deficiencies of ZeRO-Offload include low GPU utilization, imperfect overlapping of communication and computation, and inflexible offloading. In this paper, we leverage Direct Host Access (DHA) on the GPU that can compute data in CPU memory, forming a novel hybrid on-GPU and DHA. We design and implement MemFerry consisting of an execution scheduler and a shadow model. The scheduler strategically chooses layers of parameters for DHA computation and transmits the remaining parameters to GPU memory simultaneously to shorten forward propagation time, and further loads DHA parameters to GPU memory to reduce backward propagation time. The shadow model presents a unified memory abstraction for the parameter partitions stored separately in GPU and CPU memories. To further reduce GPU memory usage, we present MemFerry along with its dynamic programming algorithm that offloads gradients to CPU memory via DHA. We further extend MemFerry to emerging scale-up domains with ScaleUp-MemFerry, which exploits otherwise underutilized accelerator interconnect bandwidth to assist host-to-accelerator data movement through adaptive multi-path transfer. Our experiments show that \system trains up to $1.68\times$ faster and MemFerry can train $1.52\times$ larger model compared to ZeRO-Offload on a single GPU, and increase training speed by at least $28.1\%$ when scaling to data parallelism on 8 GPUs. We further extend the design to a Huawei CloudMatrix384 scale-Up node with up to 8 NPUs, and our ScaleUp-MemFerry reduces the end-to-end iteration time by up to $20.7\%$ over DeepSpeed.

First-principles studies of materials with strong electronic correlations introduce additional theoretical complexities and computational costs, even more so if lattice degrees of freedom also play a role. To address these challenges, we show how direct and phonon-mediated electron-electron interactions can be treated consistently. The key observation is that the GW and Fan-Migdal self-energies are topologically the same diagram, so that they can be unified using a single effective interaction, which is the sum of the two interactions. The Dyson equation for the electron Green function then yields normal-state Eliashberg equations, which can be viewed as a subset of the Hedin-Baym equations for coupled electrons and phonons [Phys. Rev. X 13, 031026 (2023)]. We solve these dynamical equations in a local approximation to calculate from first principles the temperature-dependent quasiparticle spectrum, with close attention to the low-energy window determining transport. We apply this extended dynamical Hubbard approach to the correlated metal Sr$_2$RuO$_4$, which has long served as a benchmark system for advanced electronic-structure methods. The resulting renormalization of quasiparticle bands is in qualitative agreement with experiments and results from dynamical mean-field theory. However, here the low-energy linewidth and associated resistivity within the Green-Kubo formalism in the bubble approximation are underestimated, suggesting the need for vertex corrections. This work highlights how simple dynamical formulations can provide an extensible framework for studying correlated materials, which treats electronic and vibrational excitations on equal footing.