Papers

A government may ask for a high tax because it faces a genuine expense. The same demand can also be made by a government that intends to keep the proceeds. We study this ambiguity in a two-period reputation model with privately observed spending needs and an endogenous tax base. The opportunist chooses between outright confiscation and levies that an honest government might impose. We characterize the three possible regimes through a single equilibrium equation. A mean-preserving spread of legitimate needs weakly raises the opportunist's lifetime value, with a strict increase precisely when the spread changes the upper tail relevant for mimicry. We then give a necessary and sufficient curvature condition for this ordering to extend to a general reputational continuation prize. Verification reduces the value of concealment, but also changes how the opportunist extracts. In the benchmark economy, greater auditing weakly lowers current citizen welfare, even before audit costs, while improving the future allocation through better information. The welfare case for verification consequently depends on the balance between these two effects. Examples yield no auditing, an optimum within a regime, and an optimum at the boundary between regimes.

This paper studies social learning when buying access to previous decisions also determines whose decision enters the record. Individuals first receive private information and then decide whether to pay for access. A positive price selects relatively weak private beliefs. With one founding observation, every buyer copies it, and the record acquires no further information. With a richer initial history, buyers can disagree and the record can improve; nevertheless, a price bounded away from zero prevents complete learning, even with unbounded private beliefs. We then consider a rule under which a buyer makes a preliminary choice before access and a revised choice afterward. A positive probability of implementing the preliminary choice makes it follow the private signal. Retaining both choices restores complete learning under a fixed effective fee or pooled myopic pricing, including with bounded beliefs. The rule reduces willingness to pay and sacrifices some current decision accuracy. A worked example shows how the subsequent improvement in information can outweigh both initial losses. The results show why preserving a judgment made before exposure can change what later users learn from the same selected population.

This 30 years old book is posted for free download by the generosity of Oxford University Press that granted permission to allow access to future generations of young scientists. A printed copy can be purchased from the publisher. The authors received for this book "The 1999 INFORMS Computing Society Prize for Research Excellence in the Interface Between Operations Research and Computer Science" from the Institute for Operations Research and the Management Sciences (INFORMS). For a book review by Janos D. Pinter go to: Journal of Global Optimization, 16: 107-108, 2000.

Federated learning (FL) has emerged as a leading privacy-preserving framework for collaborative machine learning across decentralized environments. While considerable progress has been made in horizontal federated learning (HFL), where data with common features is distributed across sites, vertical federated learning (VFL), where sites share observations across distinct feature sets, remains less explored. Advancing Bayesian high-dimensional multivariate reduced-rank regression methods for VFL poses unique challenges: (a) stringent privacy regulations preventing local site data sharing, and (b) fitting local regressions overlooks essential modeling aspects like inter-variable correlations. In contrast HFL allows each site to fit a comparable model independently. We present a novel Bayesian VFL framework for multivariate high-dimensional reduced-rank regression, termed BayesVFLReg, which enables precise coefficient estimation while safeguarding both feature and response privacy. Participating sites use a shared random sketching matrix to compress local variables into privacy-preserving sketches. A central server collects these sketches where Bayesian multivariate reduced-rank regression uses Gaussian scale mixture priors. For feature selection, we introduce a single-step post-processing strategy based on mixture-model clustering of the absolute posterior coefficient means to distinguish signal from noise per response variable. BayesVFLReg is computationally scalable for large, high-dimensional datasets and facilitates efficient variable selection. Theoretically, we establish sharp non-asymptotic bounds on the posterior probability that the fitted density falls within a Hellinger ball centered at the true data-generating density. Comparative simulation studies and real-world data analyses show that BayesVFLReg reliably identifies sparse feature effects, even under feature correlation.

With the ability to directly image Earth-like exoplanets and search their atmospheres for biosignatures, the upcoming Habitable Worlds Observatory (HWO) will also collect high signal-to-noise ratio (S/N) reflected-light photometry and spectra of nearby gas giant exoplanets. Such high-quality data would allow investigations into the formation and evolution of low-mass, low-temperature gas giants, and could enable the detection of exomoons around these planets. We use the EXOSIMS direct imaging mission simulator to model HWO observations of Jupiter-radius gas giants at Earth-like and Jupiter-like instellations around the 164 stars in the ExEP target list. We find that HWO should be able to achieve S/N $\geq$ 5 broadband visible-light detections of gas giants in this instellation range within 5 minutes of integration. 10 hours of R=1000 near-IR spectroscopy with HWO should reveal water, methane, and ammonia absorption features in the atmospheres of Jupiter-like gas giants. HWO time-series photometry should exceed 1% flux precision in one hour for any Earth-instellation gas giants around ExEP stars, and for Jupiter-like gas giants at $d\leq$ 7 parsecs. Time-series light curves at this cadence and precision could, over tens of hours, reveal rotation-induced variability comparable to Jupiter's. Eclipses of Mars-sized exomoons may be detectable in high-cadence light curves of Jupiter sized planets in the habitable zones of ExEP stars at $d\leq$ 10 parsecs. For any hypothetical Earth-like exomoons with oxygen-rich atmospheres at $d\leq$ 7 parsecs from the Solar System, HWO might be able to detect the spectral signature of molecular oxygen amid the parent planet's photon noise in deep ($\sim$400 hour integration) spectroscopic HWO observations at R=1000. Such moons, if they exist, represent additional habitable worlds that HWO could investigate for biosignatures.

We study the design of priority pricing systems with heterogeneous agents in environments in which improving quality for some agents reduces the average quality that can be provided. Contrary to the equity-efficiency tradeoff emphasized in public debates, we show that under economically natural conditions priority pricing can Pareto-improve on an equal-allocation benchmark. Three priority tiers suffice for such an improvement, combining higher quality for a fee, lower quality with compensation, and an intermediate tier at the benchmark quality; two tiers are never enough. Our results provide a framework for overcoming equity-efficiency tensions in applications such as lane pricing, waiting-line design, public provision, and insurance.

We study online allocation with convex costs (\OACC), a generalization of limited-supply models in which additional resources can be produced dynamically at convex cost. Prior work largely focuses on constructing a single algorithm that achieves strong competitive guarantees. In contrast, our main contribution is a structural characterization of the optimal design space of reserve functions (i.e., normalized pricing rules) for \OACC, yielding a family of optimal online algorithms. Using a principled dynamical-systems approach, we show that optimal online algorithms arise as solutions of a nonlinear eigenvalue problem: the optimal competitive ratio corresponds to the dominant eigenvalue, and the associated eigenfunction determines the optimal reserve function. This perspective establishes the existence of infinitely many optimal designs attaining the best possible competitive ratio, thereby unifying and generalizing a broad class of prior algorithms. Beyond unification, our characterization uncovers new structural phenomena, including the ability to flexibly mix dynamic and static pricing without loss of optimality and the construction of universally competitive algorithms under unknown or stochastic costs. We further extend the analysis to more structured settings such as supply-oblivious arrivals and hard supply constraints, obtaining sharper guarantees that improve upon previous bounds.

Selective area growth of PbSe has the potential to realize deterministic placement of high-density, defect-tolerant nanostructure networks toward a scalable quantum platform. PbSe is a narrow bandgap semiconductor with advantageous (opto)electronic properties that has been shown to have a desirable defect-tolerance, enabling bright emission even when grown on highly-dissimilar platforms, and could be leveraged in combination with selective growth for site-selective quantum emitters and low-disorder hybrid nanowire networks. In this work, we achieve site-selective growth of well-faceted and ordered PbSe nanostructures, despite a large 8% lattice mismatch. Structural and morphological characterization reveal that $>$99% of selectively grown islands within 100 nm opening are single-orientation and cube-on-cube oriented with sub-nm root-mean-square surface roughness. Defect analysis showed that 74% of the islands and 86% of the coalesced regions between different mask openings were free of threading dislocations. These PbSe islands achieved equivalent emission in the mid-infrared despite having a higher surface-to-volume ratio than the planar control. Further, we present a gate-tunable two-terminal Josephson junction fabricated from the PbSe nanowires grown with conditions identified in this study. The combination of the accessible selective growth regime, morphological control of island growth and demonstrations of optical and electrical transport properties indicates promise for PbSe SAG as a defect-tolerant scalable quantum platform.

Reference signal received power (RSRP) is commonly used to assess 5G cellular coverage, but it may substantially overestimate the connectivity available to cellular-connected unmanned aerial vehicles (UAVs) under strong inter-cell interference (ICI). As a UAV ascends, reduced blockage can strengthen the serving signal, while increased line-of-sight visibility to neighboring sectors can intensify ICI. This paper develops a trajectory-aware, 3GPP-based system-level framework for evaluating UAV radio serviceability during a three-dimensional flight in an urban 5G New Radio network. The framework jointly evaluates RSRP, reference signal received quality (RSRQ), and signal-to-interference-plus-noise ratio (SINR) in a 19-site, 57-sector deployment. We introduce the concept of coverage-serviceability gap to quantify the difference between RSRP availability and joint service availability. Under RSRP, RSRQ, and SINR thresholds of -100 dBm, -18 dB, and 0 dB, respectively, an inter-site distance (ISD) of 500 m provides 99.9% RSRP availability but only 13.6% joint service availability, yielding a coverage-serviceability gap of 86.3 percentage points. With RSRQ and SINR thresholds of -22 dB and -6 dB, respectively, joint availability increases to 55.0% and 63.1% for ISDs of 500 m and 1000 m, respectively. These results demonstrate that RSRP-based coverage substantially overestimates joint UAV serviceability under both evaluation configurations, with SINR remaining the limiting metric.

Post-growth defect engineering via ion implantation provides a powerful pathway for spatial optical profiling and colour patterning in III-V alloying photonic integration. However, the comprehensive recombination kinetics governing deep-level defect saturation and excitonic recovery under high-temperature annealing remain insufficiently understood. Here, we present a systematic study on the optical dynamics, rate-equation kinetics, and chromaticity evolution of indium-implanted InGaN/GaN multiple quantum wells (MQWs) across implantation doses (5E14 to 5E16 ions cm^-2) and subsequent thermal annealing stages (500 to 1100 deg C). Photoluminescence (PL) spectrum analysis reveals that a dose of <= 5E14 ions cm^-2 induces a modification of the optical response that does not further change upon high-dose implantation. A two-channel coupled rate-equation model is fitted to the data, optimised via differential evolution, allowing the extraction of transition rate constants. This demonstrates that thermal processing at 1000 deg C suppresses the carrier capture rate into deep-level states and also reduces its radiative recombination rate. This kinetic bottleneck drives an order-of-magnitude extension in the channel-specific radiative lifetime. Leveraging the excitation power density dependence of the differential recombination kinetics, where deep defects saturate whilst MQW emission scales near-linearly, we achieve a universal and power density-tunable chromaticity trajectory from warm yellow to cool white-blue emission. These insights enable microscopic defect physics to be linked to the macro-scale colour tailoring observed.

Visual representations can help lower-primary learners understand Math Word Problems, but generating classroom-usable visuals remains difficult. Existing symbolic systems are controllable but limited in coverage, while end-to-end text-to-image systems often fail to satisfy exact mathematical constraints. This paper presents a symbolic visual generation framework for lower-primary MWP generation with broader problem coverage and more scalable asset generation. The framework includes an LLM-based routing layer, three worksheet-oriented generation modules, and two fallback mechanisms for open-world SVG asset acquisition. A human evaluation comparing Math2Visual-X with Stable Diffusion XL, Nano Banana, and GPT Image showed that the proposed method achieved the strongest overall performance. The results indicate that the framework offers a scalable and pedagogically grounded approach for automatic MWP visual generation.

Loss of Load Expectation (LOLE), originally calculated as the expected number of days of load loss, has also been used as a frequency metric. This paper shows that LOLE can indeed be calculated as a frequency metric, but only with restricted conditions, and that it differs fundamentally from the standard frequency metric employed in the frequency and duration (F&D) approach. A lack of understanding of these conditions can lead to incorrect applications and interpretations.

Scaling analog and digital Ising machines to larger problems requires overcoming finite device capacity and the cost of communication between devices. We present cluster mean-field theory (CMFT), a framework that partitions a large interaction graph into clusters sized to fit available hardware. Each cluster performs local updates independently, while interactions across cluster boundaries enter through periodically updated effective biases computed from boundary-spin averages. To reduce the error introduced by fixed cluster boundaries, we introduce dynamic partitioning, which cycles through multiple partitions so that interactions approximated by mean fields at one stage can act through instantaneous spins at another. On three-dimensional spin glasses and planted Pegasus instances, dynamic CMFT exhibits power-law decay of residual energy over sweep budgets. This shows that solution quality continues to improve with computational effort despite the mean-field approximation. An automated graph partitioner combined with a weighted recovery ratio provides a practical heuristic for selecting partition combinations on graphs without natural cut directions. We demonstrate CMFT on four GPUs with approximately four million p-bits, reaching comparable energy densities up to 15 times faster than a single-GPU implementation of the full graph. By coupling locally evolving clusters through programmable effective biases, CMFT provides a route to extreme-scale Ising machines on both analog and digital hardware.

Black holes ring when perturbed, whereas their response to a stationary gravitational-wave background is a real-frequency scattering problem, not a source of additional quasi-normal-mode lines. We make this standard distinction quantitative by treating a black hole as a frequency-, angle-, and polarization-dependent filter. For an isotropic stationary background around Schwarzschild holes, elastic scattering produces no net monopole signal, so horizon absorption is the only population-level spectral distortion. We calculate this transfer function for Schwarzschild holes in detail, identify its absorptive and phase-delay signatures, and connect the stationary response to the causal ringdown excited by a finite wave packet. We then promote the single-hole result to an angular and polarization transport kernel for a cosmological population. The resulting optical depth is negligible for realistic black-hole populations, including asteroid-mass primordial black holes comprising all dark matter. We then extend the analysis to Kerr holes, for which superradiance allows genuine amplification in selected co-rotating modes, but isotropic incidence and random spin orientations strongly dilute the diffuse signal. Observable effects are therefore more likely in rare, nearby, aligned, rapidly spinning, or transiently illuminated systems than through cumulative cosmological propagation.

Forecasting a long horizon from only the first observations of a sequence is ill-posed: many trajectories are consistent with the same short history. We study this problem in oil and gas production forecasting, where forecasts made after roughly the first fifth of a well's producing life drive development and abandonment decisions, and where a usable forecast must describe a monotone decline. We present Physics-SIMS-TS, a conditional diffusion forecaster that combines negative guidance against synthetic artifacts, decline-curve constraints and an isotonic projection applied during sampling, spatial training augmentation, and an ensembled stochastic sampler yielding a full predictive distribution. Across three jurisdictions and more than 35,000 wells, under a shared-space, validation-frozen protocol, Physics-SIMS-TS is the most accurate diffusion forecaster in the comparison and is competitive with, but not superior to, ensembled transformer forecasters. Its forecasts are monotone by construction at a cost of at most 0.5% in mean squared error, and its trajectory ensemble yields calibrated intervals after one dispersion factor is fitted per jurisdiction. On six standard benchmarks a reversible-instance-normalization variant of the backbone is the leading diffusion baseline. We also quantify four protocol choices on which the measured ranking depends. Code and evaluation artifacts are released.

We develop inference for manifold-valued population minimizers when each observation belongs to a different participant and only locally private messages reach the analyst. The method releases randomized tangent gradients and combines them through Riemannian stochastic approximation and Polyak-Ruppert averaging. Directly inserting a private data surrogate into a nonlinear loss can shift its population target, whereas conditional centring of the released gradient preserves the first-order equation. We introduce symmetric-pair regression (SPR) to estimate the asymptotic variance from the same private messages used for point estimation, without holding out participants or requesting a second release. We prove the central limit theorem and consistency of the fully transcript-based sandwich covariance and intrinsic Wald region under local differential privacy. Simulations across various statistical problems and manifolds support the predicted decrease in estimation error and near-nominal coverage under moderate privacy. An application to NHANES anthropometric data illustrates private estimation of a leading body-size direction and its uncertainty.

Autoregressive video world models enable temporally coherent generation for a single observer. Extending them to multiple agents requires consistency across independently controlled views and temporal gaps under causal streaming. We present ConsistWorld, a multi-agent world model that generates camera-controlled video streams of a static scene from one shared image. We formulate consistency as routing evidence from committed multi-agent history and concurrently generated peer views to the tokens being generated. Pose Conditioned Memory Retrieval selects relevant historical observations from all agents, recovering evidence beyond the recent context window. Visibility-Gated Peer Sharing regulates current peer information according to estimated historical coverage and current-view overlap. Together, they determine which historical observations enter the context and where concurrent peer information contributes, supporting long-term recall and coordinated exploration. Both mechanisms use camera geometry and maintain a bounded active context for a fixed agent count and retrieval budget. Experiments on evidence sharing cases and video length and agent number generalizations show that ConsistWorld achieves a strong cross-time and cross-agent consistency while preserving competitive generation quality.

We study the complexity of determining the interval numbers of tournaments in the $\overrightarrow{P_3}$ and $\overrightarrow{P_3^*}$ convexities, denoted by $\overrightarrow{\mathrm{in}}_{P_3}(T)$ and $\overrightarrow{\mathrm{in}}_{P_3^*}(T)$ on a tournament $T$. For each $\overrightarrow{\mathcal{X}} \in \{\overrightarrow{P_3}, \overrightarrow{P_3^*}\}$, we show that determining whether $\overrightarrow{\mathrm{in}}_{\mathcal{X}}(T) \leq k$ is W[2]-complete when parameterized by $k$. Moreover, under ETH, we show that there is no parameterized algorithm for that problem with running time $f(k)\, n^{o(k)}$ on an $n$-vertex tournament, where $f$ is any computable function. For the $\overrightarrow{P_3}$-convexity, we also show that $\overrightarrow{\mathrm{in}}_{P_3}(T) = \mathcal{O}(\log n)$, which yields a simple quasi-polynomial $n^{\mathcal{O}(\log n)}$ brute-force algorithm. On the other hand, under ETH, we show that the problem is NP-intermediate, that is, it is neither NP-hard nor in P. For the $\overrightarrow{P_3^*}$-convexity, the same brute force algorithm is not quasi-polynomial, since we present a family of instances with $\overrightarrow{\mathrm{in}}_{P_3^*}(T) = \Theta(n)$. We conjecture that this problem is NP-complete.

Pairs trading exploits mean reversion in the relationship between related assets. We adapt this idea to political betting markets by modelling the combined implied probability of the two major-party nominees with a latent Ornstein-Uhlenbeck process whose mean-reversion level varies over time and whose observations contain additive noise. Model parameters are estimated from regularly sampled odds data using a state-space likelihood, with consecutive repeated values represented by a single retained observation and the elapsed number of sampling intervals preserved in the continuous-time transition. Parametric-bootstrap upper prediction bounds identify signal times at which the combined implied probability is likely to decline, and a no-intercept Bradley-Terry-type model selects the candidate-specific odds quote. The candidate-selection model is trained on 2020 U.S. presidential-election data and evaluated out of sample on 2024 data. The 2024 analysis produced 130 signals, empirical one-step coverage of 95.1%, a mean synthetic odds-price return of 1.86%, and an unannualized per-trade Sharpe-type ratio of 1.12. These returns are frictionless descriptive quantities rather than executable betting-exchange profits. The results support the integrated framework as a proof of concept for two-candidate electoral markets.

Light, weakly coupled particles can be copiously produced in the hot and dense interior of stars, and the resulting energy loss can leave observable imprints on entire stellar populations. For instance, additional energy loss from red giant cores delays helium ignition and brightens the tip of the red giant branch (TRGB). We revisit the TRGB bounds on light scalars coupled to electrons, of which the Higgs portal is the prototypical example. In the degenerate cores of red giants, scalar emission is dominated by resonant conversion of longitudinal plasmons, which we compute using finite-temperature field theory. Rather than rescaling neutrino emission rates at fixed benchmark core conditions, we implement this emissivity in the MESA stellar evolution code and self-consistently simulate the evolution of red giants, including the backreaction of the scalar energy loss on the stellar structure. Comparing our predictions with the observed bolometric TRGB magnitudes of 27 Milky Way globular clusters, we exclude scalar-electron couplings $\alpha_{\phi ee} = g_e^2/4\pi > 1.6 \times 10^{-34}$ at 95% confidence for sub-keV masses, improving on previous bounds by nearly two orders of magnitude. In the context of the Higgs portal, this corresponds to an exclusion on values of the mixing angle $\sin\theta > 2.2\times 10^{-11}$, providing the tightest bound for scalar masses in the $\sim 1\,\mathrm{eV} - 10\,\mathrm{keV}$ range.

Policy learning methods based on conditional average treatment effects can obscure subpopulation heterogeneity when applied to ordinal outcomes. We develop a policy learning framework for ordinal outcomes with heterogeneous utilities for individuals who strictly benefit from treatment and those who do not. Since the probability of strict benefit is only partially identified without further conditions, we adopt a minimax strategy which minimizes the worst-case regret over the identification region. To estimate the resulting nonsmooth objective, we combine smooth approximations with Neyman orthogonalization to remove first-order bias from nuisance estimation. We also derive an excess worst-case regret bound over a restricted policy class. The proposed method is validated through extensive simulations and an application to the 2022 Survey of Income and Program Participation (SIPP) data.

Data prefetching is an established technique to mitigate cache miss latency and keep the processor saturated with data. Software exists in a unique position to issue prefetches, having algorithmic knowledge of the workload at hand. However, inserting software prefetches is a time consuming, unpredictable task, with a high degree of sensitivity to microarchitecture or future code changes. While several compiler- and FDO-based approaches insert and tune software prefetches automatically, they use heuristics that do not generalize and do not consider issues that arise in large codebases. In this paper, we show that semantic understanding and reasoning about a program is a vital component in inserting effective software prefetches. We also identify that prefetching at the scale of large codebases poses new challenges, including prefetch codependence and interference. To remedy this, we build Presage, a system that leverages the semantic reasoning of Large Language Models to insert effective software prefetches. Presage processes a workload through a custom agent harness specifically designed to navigate the wide space of potential prefetches on large-scale workloads. First, a proposer agent with access to performance metrics identifies promising software prefetching candidates. Then, several long-horizon optimizer agents optimize prefetching candidates in a loop, working towards discovery of performance wins. Finally, a combiner agent handles composition of individually effective prefetches. Through this method, Presage is able to insert prefetches that improve performance by a geomean of 10% across 81 workloads by exploring tens of different prefetching alternatives per workload, including a 2.7% runtime reduction on SPEC CPU 2026 where prior SOTA fails. When compared against prior SOTA on its evaluation suite, Presage achieves a geomean runtime improvement of 17% compared to SOTA's 10%.

Three-dimensional fetal brain ultrasound offers non-ionizing, cost-effective imaging with rich spatial information for comprehensive anatomical assessment, yet its development remains limited by scarce data and annotations. In contrast, fetal brain magnetic resonance imaging has advanced further, supported by larger datasets and mature analysis methods. To leverage these resources, anatomical label maps provide a promising modality-invariant bridge for transferring knowledge from fetal brain magnetic resonance imaging to ultrasound. However, generating ultrasound images from label maps remains challenging because anatomy is tightly entangled with ultrasound appearance. To address this challenge, we separate anatomical correspondence learning from ultrasound appearance adaptation at both the data and model levels. Specifically, we propose the anatomy wavelet residual network, a two-stage framework combining wavelet diffusion and residual refinement. The first stage learns to generate volumes from label maps using atlas pairs in the wavelet domain, while the second stage learns bounded residual corrections from real clinical ultrasound in the image domain. This separation enables realistic synthesis with consistent preservation of normal and abnormal anatomy. Experiments on real fetal brain ultrasound show that our method outperforms representative synthesis methods, with normalized cross correlation of 0.518 versus 0.482 and Fréchet Inception Distance of 8.905 versus 13.319 for the strongest baseline. Beyond synthesis quality, volumes generated from fetal brain magnetic resonance imaging label maps also improve downstream segmentation, particularly for severe abnormal cases. Overall, these results highlight the potential of the proposed framework to leverage rich fetal brain magnetic resonance imaging resources for advancing three-dimensional ultrasound analysis.

Designing observatory and instrument architectures for complex missions like the Habitable Worlds Observatory (HWO) requires a rapid and rigorous mathematical framework to evaluate trade-offs across bandpass, spectral resolution, and signal-to-noise regimes. While full Bayesian atmospheric retrievals are the gold standard for atmospheric inference, they are too computationally expensive to explore this full parameter space. Here, we present a sensitivity analysis and retrieval diagnostic toolkit within the open-source code PICASO. The core addition to the code base is a flexible solution for computing finite-difference Jacobian matrices, which provides a local linearization of the forward model. These Jacobian matrices characterize the sensitivity of the spectrum (either reflected, thermal, or transmission) to perturbations in any atmospheric state parameter, such as molecular abundances. Given the Jacobian for a certain reference atmosphere, our toolkit provides a suite of analytic diagnostic quantities under Gaussian assumptions for the prior, likelihood, and posterior. These include Fisher-matrix-based estimates of 1$\sigma$ constraint intervals of atmospheric parameters, averaging kernels, singular-value-decomposition diagnostics for parameter degeneracies, and an information-content metric. We validate the results of our toolkit by comparing them to that of a full Bayesian retrieval of an Earth-like planet with an HWO-like facility. We show that our metrics can provide rapid understanding of model parameter sensitivity and that local linear-Gaussian approximations can diagnose dominant degeneracies seen in retrievals as well as provide first order estimates of posterior probability distributions. Ultimately, this provides the community with a complementary toolkit for optimizing instrument design and conducting spectral sensitivity diagnostics.

Bilingual language models (LMs) offer a controlled setting for studying how training conditions shape second-language (L2) behaviour, but prior work typically varies exposure structure, scale, and architecture at once, making it difficult to attribute effects to any single factor. We introduce Beetle, a controlled language model pretraining framework in which tokeniser, target language, training budget, and exposure structure are each independently manipulable, enabling systematic and comparable experimentation of training conditions. Using Beetle, we train and release 285 bilingual and 45 monolingual open-source LMs with rich checkpoints across a range of exposure schedules, data scales and first languages (L1s) to study multilingual pretraining and computational modelling of bilingualism and second language learning. Evaluating models on human bilingual and second language reading-time prediction and grammaticality judgement tasks, we find that staged and temporally structured curricula consistently improve alignment with language learner reading time compared to balanced bilingual training, with the largest gains at smaller data scales and for typologically closer language pairs. The Beetle models are well suited tools to help move computational psycholinguistics beyond its prevailing monolingual, English-centric focus toward models of human bilingual processing, to study cross-lingual learning dynamics, while supporting community-based development of controlled model families.

We study binary classification with abstention under separate class-conditional error constraints, with the objective of minimizing abstention while keeping both errors below prescribed thresholds. We characterize the distribution-free minimax rate of excess abstention risk, up to logarithmic factors, in terms of the complexity of the hypothesis class and the sample size. To make the framework amenable to computation with models such as neural networks, we introduce surrogate-loss formulations and derive finite-sample guarantees for excess surrogate ambiguity risk. We formulate the resulting learning task as a constrained optimization problem and characterize its computational complexity in the convex setting. Finally, we evaluate our approach on various datasets and compare its performance with a competing method for this problem.

Accurate automated interpretation of electrocardio- grams (ECGs) is essential for early detection of cardiac condi- tions such as myocardial infarction and rhythm abnormalities. However, many high-performing deep learning models remain difficult to deploy in clinical settings due to limited transparency and lack of reliability validation. In this work, we present X- Beat, an explainable and reliability-aware benchmark framework for ECG image classification designed to support trustworthy AI systems in healthcare. The proposed framework combines transfer learning with post-hoc explainability and systematic reliability evaluation across four cardiac classes: Abnormal Heartbeat, History of Myocardial Infarction, Myocardial In- farction, and Normal Heartbeat. Multiple ImageNet-pretrained CNN backbones, including EfficientNet-B0, ResNet-50, DenseNet- 121, and MobileNetV3-Large, are evaluated under a unified training protocol. Beyond standard performance metrics, we incorporate Grad-CAM-based visual explanations together with additional analyses, including explanation stability, regional sen- sitivity, and confidence-based reliability assessment, to examine whether model predictions are supported by clinically meaningful evidence. Experimental results show that ResNet-50 achieves the best performance, reaching 91.94% accuracy and a macro F1- score of 0.9098, with strong class separability (AUC up to 0.995). Explanation analyses indicate that the model primarily focuses on waveform-relevant regions, while reliability evaluation suggests that most incorrect predictions occur with lower confidence. Overall, this work provides a structured and reproducible bench- mark for evaluating both predictive performance and explanation reliability in ECG image classification, contributing toward the development of trustworthy and interpretable AI components for clinical decision support systems.

Subjective user experience is important to human-robot interaction, but the outcomes users value, and how those preferences vary across individuals and contexts, are often unknown. While inverse learning approaches using human data can help identify user rewards, in many assistive settings the experimental costs of executing a control policy, measuring biomechanical or physiological outcomes, and collecting user feedback often limit the number of queries and optimization iterations. Here, we present Simultaneous Forward and Inverse Human-In-the-Loop Optimization (SFIHILO), which efficiently infers individual-specific reward functions from preferences over human outcomes and identifies a final control policy that maximizes the learned reward. SFIHILO bootstraps a forward model to predict user outcomes from control policies, uses this model for active querying to accelerate inverse reward learning, and then optimizes the final control policy without additional user trials. We score candidate policies by their expected reduction in uncertainty across both forward and inverse beliefs, targeting a regional preference boundary to robustly inform this simultaneous learning process. In simulation, we show that SFIHILO was effective across user heterogeneity, outcome dimensionalities, precision requirements, noise levels, and nonstationarity; compared with mutual information approaches, the proposed active querying strategy significantly improved sample efficiency in inverse learning while preserving forward model accuracy. This approach demonstrates the potential to infer latent human goals, enabling more transferable and effective human-robot interaction.

Surface nanoscale axial photonic resonators (SNAPRs) provide an axially structured whispering-gallery-mode platform in which nanoscale effective-radius variations produce spatially localized optical resonances. Here, we demonstrate cascaded Brillouin-Kerr microcomb generation in a SNAPR fabricated from 660 micrometer diameter silica fiber. Under continuous-wave pumping, cascaded stimulated Brillouin scattering generates first- and second-order Stokes fields that seed Kerr four-wave mixing within the cavity. The resulting nonlinear spectrum contains two comb families with a common azimuthal spacing of 100.2 GHz and a relative offset of 9.6 GHz, consistent with the Brillouin shift in silica. A low-power transmission spectrogram provides a spatially resolved modal analysis of the resonator, revealing localized resonances with loaded quality factors on the order of ten million, and optical resonance pairs compatible with the measured Brillouin spacing. These results establish SNAPRs as a platform for studying cascaded stimulated Brillouin scattering, Kerr four-wave mixing, and comb formation in axially structured whispering gallery resonators.

Every document QA system begins with a choice that is rarely studied on its own: whether to feed the model page images, extracted text, or both. We isolate this choice, holding the prompt, judge, and scoring pipeline fixed, across four commercial model endpoints, two corpora, and two context regimes (gold evidence pages and the full document). On documents that fit the image budget, page images lead on accuracy at every document length on both corpora, but this advantage carries a growing latency and cost premium: text latency stays roughly flat as documents lengthen while image latency rises steadily. Text and images also fail on different questions, with exactly one representation correct on 19–25% of items across the reported cells, so neither subsumes the other. Exploiting this complementarity, a lightweight TF-IDF router that reads only the question text gains 2.6 points over always-text while cutting median latency 30% relative to always-vision, on a document-disjoint held-out split.