Papers

Connectors are ubiquitous in technical systems and frequently combine mechanical contact with electric current transfer and heat generation. To capture these interacting processes, an electro-thermo-mechanical contact formulation based on the third-medium concept is proposed. Mechanical contact, electric current flow, and heat conduction are represented within a unified finite element framework without explicit contact-surface tracking. Deformation-dependent switching functions govern the onset of electrical and thermal transport during interface closure. Electrical conduction is coupled to the transient thermal problem through Joule heating, while the temperature dependence of the electrical conductivity provides the corresponding feedback on the current flow. The coupled displacement, electric potential, and temperature fields are solved monolithically and combined with an independently interpolated deformation-gradient-like field for robust third-medium regularization under severe compression. Numerical examples demonstrate contact-induced current transfer, subsequent Joule heating and transient temperature evolution, as well as localized current paths during progressive closure of rough interfaces.

Vision Language Models have achieved strong performance on multimodal benchmarks, yet their ability to reason about culturally grounded and metaphor-rich content remains insufficiently studied. Internet memes present a challenging setting where meaning emerges from implicit interactions between image, overlaid text, sarcasm, and shared socio-cultural knowledge rather than literal visual recognition. This challenge is amplified in low-resource languages such as Bangla, where code-mixing, stylized scripts, and culturally specific symbolism introduce substantial distribution shift. In this work, we introduce BanglaMemeX, a culturally grounded multimodal benchmark comprising 3,000 Bangla memes annotated with multi-dimensional labels (humor, sarcasm, offensiveness, motivational intent, and overall sentiment) and human-written explanations that explicitly describe textual and visual metaphors. We systematically evaluate modern VLMs on both classification and explanation generation, revealing that current models struggle to interpret implicit cultural cues despite reasonable surface-level accuracy. Our results highlight the need for culturally-aware multimodal systems capable of grounded reasoning under linguistic and cultural distribution shift.

Importance: Reports have raised concerns that AI chatbots may validate or elaborate delusional beliefs, respond inappropriately to suicidal ideation, and contribute to mental health harms, but real-world data on reported harms remain limited. Objective: To characterize psychopathological features, chatbot behaviors, timing, and outcomes in first- and second-hand accounts of mental health harm linked with AI chatbot use. Design: Cross-sectional secondary analysis of deidentified online survey responses gathered between August 7, 2025, and February 2, 2026. Main Outcomes and Measures: The primary quantitative outcome was the presence of delusional beliefs, coded by paired raters with relevant clinical experience. Additional variables included reason for chatbot use, current episode features, delusional content, chatbot validation of beliefs, harms, social and occupational consequences, healthcare use, and timing. Results: 95 first-hand and 90 second-hand accounts were analyzed. Median age was 35.0 (IQR 27.0 - 45.0). Raters coded descriptions consistent with delusional beliefs in 102 reports (55.1%), with chatbot validation of beliefs in 50/102 (49.0%). Common outcomes included isolation, relationship breakdown, hospital admission, job loss, and financial loss. Four second-hand reports described death by suicide. Conclusions: In this self-selected convenience sample, AI-chatbot-associated harms were frequently described in relation to delusional beliefs, perceived chatbot validation, intensive use, and substantial social, occupational, and clinical consequences. Because reports were retrospective, unverified, and collected from individuals seeking to report harm, our findings should be interpreted as preliminary signal detection rather than as suggesting prevalence or providing evidence of causality. Prospective surveillance and trajectory-based safety evaluations are needed.

Acoustic scene understanding in multi-room environments is a difficult task. Most existing systems use a single centralized microphone array, and they often fail because walls and doors block sound signals. To address this challenge, we propose a geometry-informed distributed acoustic scene understanding framework. Our system leverages distributed microphones and uses an audio spectrogram transformer and a topology-aware graph neural network to fuse spatio-temporal acoustic features. Then, these features are decoded into discrete semantic triplets. Finally, a frozen large language model combines these symbolic observations with the environmental geometry. This allows the system to perform spatial understanding, infer plausible missing transitions, and generate a physically consistent narrative of the scene. Experiments on a custom multi-room simulator demonstrate that our framework outperforms centralized baselines and improves spatial consistency under simulated occlusion.

Reasoning agents increasingly rely on external tools such as web search to answer complex queries. Reinforcement learning (RL) finetuning algorithms such as GRPO have improved long-form reasoning in text-only language models, particularly for coding and mathematics. Reliable tool use in multimodal agents, however, remains challenging because models must interpret text and images while integrating noisy retrieved evidence, often under sparse outcome-level supervision without explicit verification signals. We present Self-Verification via Reinforcement Learning (SVRL), an RL-only finetuning framework that trains multimodal agents to verify and filter retrieved evidence within their own reasoning traces, reducing reliance on external verifiers at inference time. SVRL also introduces a search-aware penalty that discourages unnecessary tool calls and a query-diversity reward that encourages diverse, well-formed search queries, providing fine-grained feedback on when and what to search. Finetuning Qwen-2.5-VL-7B with SVRL on only 5{,}000 visual question answering examples yields consistent gains in multi-hop VQA generalization and tool efficiency across benchmarks. Overall, SVRL narrows the gap between compact agents and much larger proprietary models while requiring substantially lower training and inference cost.

Traditional Bayesian optimal experimental design (OED) selects measurements that best inform a model's parameters. However, such measurements can be suboptimal for downstream predictions. Goal-oriented OED targets the prediction directly. However, the existing goal-oriented criteria value all reductions in predictive uncertainty equally, with no way to prioritize rare, high-consequence outcomes. In this article, we develop a risk-aware framework that composes risk at three levels, each generalizing an ingredient of classical $I$- and $G$-optimal design: a deviation measure of the posterior predictive uncertainty (generalizing the predictive variance), a risk measure across the prediction domain (interpolating $I$-optimal averaging and $G$-optimal worst-case selection), and a risk measure over datasets (generalizing the expectation). We generate each level from a regret function in the risk quadrangle, so that one triple specifies a practitioner's risk preference. We relax the design to continuous weights on the unit simplex and construct a nested-quadrature estimator that is differentiable in the design variable. This enables solving the optimal design problem with gradient-based methods, avoiding a combinatorial search over candidate designs. For a linear-Gaussian lognormal model and a nonlinear extension, we derive closed-form objectives. These give exact references against which we verify that the estimator converges. We demonstrate this framework for finding optimal sensor placements in an inverse problem governed by an advection-diffusion equation. We find that the risk-aware designs substantially outperform the expected-information-gain baseline, which is statistically indistinguishable from a random allocation.

This paper studies lower-level-constrained bilevel optimization in a response-oracle setting, where lower-level model information is unavailable and the induced response mapping is locally Lipschitz but potentially nonsmooth. In this setting, the classical response Jacobian and reduced hypergradient may fail to exist. We propose Bilevel Zeroth-Order Learning (Bi-ZOL), a structure-guided zeroth-order method for finding stationary points of the nonsmooth reduced problem. Instead of estimating the gradient of a fully smoothed reduced hyperobjective, Bi-ZOL separates the bilevel chain-rule structure: it keeps the exact upper-level partial gradients at the queried response and uses zeroth-order sampling only to estimate the response Jacobian. This construction yields an approximate hypergradient that is more directly aligned with the Clarke chain-rule subdifferential. We show that the Bi-ZOL direction admits a partial-smoothing interpretation, quantify its pointwise structural bias, and prove finite-time convergence to a $(\delta,\epsilon)$-Bi-ZOL Frank–Wolfe stationary point. The bias is $O(\delta)$ for piecewise $C^{1,1}$ responses under local regularity and vanishes for piecewise affine responses on active-cell neighborhoods. Experiments on incentive-based tracking problems show that Bi-ZOL achieves smaller stationarity gaps and lower hyperobjective values than vanilla zeroth-order smoothing under comparable response-oracle budgets.

Distributed systems support many critical business workflows, but service health is often judged through component dashboards rather than through end-to-end user outcomes. This article presents service health engineering as a practical reliability discipline that connects telemetry, workflow completion, dependency behavior, operational readiness, and recovery validation. Using a document approval workflow as a running example, it describes how service promises, service-level indicators and objectives, watchdogs, incident measures, resiliency testing, and weekly service-health reviews can reveal silent failures and stranded asynchronous work. It also presents a human-reviewed, AI-assisted reporting architecture for assembling service-health evidence without making AI an autonomous decision-maker. The approach brings established reliability practices together around whether user journeys complete as promised.

URLs in scholarly documents link to rich external resources such as datasets, software, publications, and websites. Extracting these URLs is crucial in the data preparation stage of many downstream tasks, such as link rot analysis, web crawling, and building knowledge graphs. However, existing studies often downplay this phase, simply extracting URLs from a single format, usually text directly converted from PDFs. We present a systematic study evaluating URL extraction across six input formats (text with annotation layer, LaTeX, HTML, XML, Markdown, and PNG converted from PDF). To support the evaluation, we compiled a benchmark dataset consisting of 2,338 manually annotated URLs from 200 arXiv papers spanning a wide range of domains over a 33-year period. In addition to evaluating individual file formats, we also compared 63 composite input-format combinations. Our extensive evaluations indicate that TEXTWAL achieves the best performance among single-format inputs, while TEXTWAL+LaTeX achieves the best overall URL extraction performance. The same trend is observed for URLs linking to open-access datasets and software. To further validate these findings, we apply our format-specific URL extraction pipelines to a longitudinal random sample of 364,744 arXiv papers spanning 33 years. We observe a sharp increase in URL density after 2015, along with remarkable differences in URL extraction across file formats over time. Overall, our study highlights the importance of selecting an appropriate format for URL extraction from scholarly documents. The dataset and code are publicly available at: https://github.com/lamps-lab/arxiv-url-bench .

Multi-agent debate, in which several LLMs exchange arguments before producing an answer, raises a basic question: does expressed disagreement reflect changes in the members' own positions? No single signal can settle this question, so we organize the analysis around five questions: (A) does the debater say it disagrees; (B) does its reply text actually argue; (C) does its own position change after each debate turn; (D) how much, quantitatively, does the position change; and (E) how do members' final positions compare with their initial ones? We evaluate two- and three-member committees on 50 curated opinion questions from GlobalOpinionQA, using same-model, same-family, and mixed-family configurations with friendly, neutral, and hostile instructions assigned at each turn. (A) Tone changes reported agreement: the share of replies reporting strong agreement is 70.7-96.3% under friendly instructions, compared with 9.5-19.3% under hostile ones, varying with model choice. (B) Self-reports broadly align with text judgments, but consistency varies by agreement level and model choice. (C) Position changes depend on the interaction: after a peer's leaning-disagree reply, members reporting strong agreement switch options more often than those reporting leaning disagreement. (D) For open-weight members, the probability of the option a member already holds stays near saturation, even after a peer's pushback, while endorsement of its earlier position text drops after a peer's argument relative to neutral filler, more for Qwen3.8-27B than Inkling. (E) The selected option is unchanged from members' initial to final positions in over 90% of comparisons in every configuration. Taken together, expressed disagreement need not translate into position revision, either within individual exchanges or over a complete debate.

Long-running AI agents may read state, reason, wait for tools or human approval, and perform an external action much later. The state that justified the action can change in the meantime. For example, after an agent proposes an 80 GBP refund under a limit of 100, a customer-name change affects only presentation metadata, a new limit of 90 still permits the refund, a limit of 50 invalidates it, and a refund issued by another worker must prevent a duplicate. Standard optimistic concurrency control and version checks can detect that previously read state has changed, but by themselves do not determine whether that change invalidates the pending action's justification. We call any detected version change a version conflict; when that change invalidates the action's justification, it is also a decision conflict. ATR records the explicit, executable conditions that justify a pending action and rechecks only the conditions affected by a change before releasing the external operation. It can retain the action, refresh non-decisive metadata, require replanning, or block execution; a target-side transaction or compare-and-set binds checked state to commit. Across 210,000 controlled executions over 15 mutation cases, ATR matched every developer-specified outcome with no false allows or blocks. In ten durable SQLite checkpoint/resume cells, it evaluated 0.6 conditions per change versus 6.0 for FullScan. At 4,093 recorded reads, ATR took 9.3 microseconds versus 2595.9 microseconds for FullScan. These deterministic results establish controlled feasibility, not production generality or automatic extraction of the required conditions.

Human pose estimation and keypoint-based action recognition models are increasingly deployed as components of video understanding pipelines, yet their vulnerability to adversarial attacks remains insufficiently studied. Temporally coherent black-box attacks have been previously studied in visual object tracking, where the attack feedback can be defined using bounding-box overlap measures such as Intersection over Union (IoU). However, human pose estimation produces keypoint configurations rather than enclosing boxes, making box-level similarity poorly suited for measuring pose degradation. We propose OKS Attack, a decision-based black-box attack that uses Object Keypoint Similarity (OKS) as the attack feedback signal, directly targeting the spatial structure of human poses rather than their enclosing boxes. Experiments on the Penn Action dataset show that OKS Attack consistently reduces pose quality across evaluated pose estimators, with mean OKS decreases ranging from 0.0802 to 0.1494. In a downstream cross-dataset action-recognition evaluation, the attack reduces accuracy by 6.18 to 13.86 percentage points and outperforms query-matched random-noise perturbations. The attack is effective across both top-down and single-stage pose estimation models. The source code will be made publicly available at https://github.com/KacperM33/OKS_attack

Image restoration is commonly applied before object detection under adverse conditions, yet a visually improved image need not improve the downstream task. We study this mismatch as restoration utility prediction: given a degraded image and its candidate restoration, should the restoration be used or should the original observation be preserved? We introduce TaskGuard, a post-hoc controller for frozen restoration and detection pipelines. TaskGuard characterizes the realized restoration residual through its interaction with detector sensitivity and predicts whether the intervention is task-beneficial. Exact regional counterfactuals reveal substantial within-image utility heterogeneity, while a deployable pseudo-gradient preserves statistically reliable directional information. Feature-group ablation further shows that task-conditioned evidence contributes information beyond detector-response and residual statistics. The TaskGuard utility predictor is trained only on Gaussian degradation and frozen before final evaluation, then transferred to unseen motion blur, rain, and defocus. Across these unseen families, TaskGuard reduces lossnegative interventions by 54.2% (family macro) and practical per-image detection deteriorations by 37.0% (pooled), while preserving 98.8% of the Always-Restore COCO AP. On natural-rain DAWN, it reduces loss-negative interventions by 97.9% while retaining 77.8% of the AP improvement obtained by deraining. These results support restoration utility as a task-conditioned property of the specific intervention rather than image appearance alone.

This paper presents a general framework for simulating multi-body space robots with contact. We bring efficient, large-scale robot simulation to in-space servicing, assembly, and manufacturing applications. First, we perform an empirical trade study of methods for coupling orbit propagation with existing robotics simulation frameworks. Next, we present mjorbit, a general, flexible, and performant framework built on the MuJoCo engine widely used in robotics, to which we add key spacecraft dynamics, actuators, and sensors. We provide a low-latency C++ CPU backend and a high-throughput GPU backend behind a simple Python API. We demonstrate mjorbit by solving several realistic on-orbit case studies with both model-predictive control and reinforcement learning. Open-source code and examples are available at: https://johnzhang3.github.io/mjorbit/

Otome games, a romance simulation genre primarily targeting female, have emerged as a major force in the global gaming market, attracting hundreds of millions of players and billions in revenue. Despite their popularity, otome game communities face pervasive online toxicity, which has been largely unexplored. In this work, we present the first large-scale measurement of toxicity in otome game communities across social platforms. We introduce OtomeSCAN, a framework for collecting, evaluating, and analyzing 620,045 posts from Weibo and Reddit spanning 18 months. To support robust analysis, we manually annotated a ground-truth dataset of 4,308 posts, identifying eight target groups such as players and game developers. We evaluate seven toxicity detectors on the dataset, including general-purpose models and our proposed LLM-based detectors, with our best model achieving F1-scores of 0.82 (Weibo) and 0.78 (Reddit). Our analysis reveals significant platform-based differences in toxicity: 22.20% of otome-related posts on Weibo are toxic, compared to 3.71% on Reddit. Besides, real-world events like in-community conflicts can rapidly escalate toxicity, with toxicity ratios increasing to 37.09% in just 72 hours during an external attack on Weibo. We also flag 191 potential-coordination clusters in otome game communities, 64.40% of which target game developers, with several accounts participating repeatedly across multiple clusters. We hope our work inspires further research on community-specific toxicity and contributes to building healthier online spaces for marginalized gaming communities.

Cell-to-cell electrical and thermal variations produce nonuniform state of charge (SoC) and temperature distributions in lithium-ion battery packs. This paper proposes a Koopman-operator-based framework for simultaneous SoC and temperature balancing of series-connected cells. Cell-specific Koopman predictors identified using extended dynamic mode decomposition are assembled into neighboring-cell error dynamics that retain heterogeneity and thermal-disturbance effects. A Tikhonov-regularized feedforward controller provides numerically robust compensation without excessive current amplification, while constrained Koopman-based model predictive control (KMPC) regulates the residual errors through a convex quadratic program. Three-cell simulations validate the controller under two consecutive Urban Dynamometer Driving Schedule cycles. KMPC achieves balancing performance comparable to nonlinear model predictive control (NMPC). Relative to the uncontrolled case, KMPC reduces the RMSEs of the neighboring-cell SoC and core temperature differences by 66-81% and 57-76%, respectively, while reducing the computation time per optimization from 0.0259s for NMPC to 0.0028s for KMPC.

We give a deterministic polynomial-time list-decoding algorithm for Reed-Solomon codes over prime fields that approaches list-decoding capacity for every evaluation set and every constant rate.

Behavioral foundation models have been proposed as stand-ins for human participants across settings, but it is unclear whether theories discovered on them generalize to humans or merely characterize the simulator. We ran the Automated Cognitive Scientist (AutoCog), a closed-loop discovery system in which LLM agents design theory-discriminating experiments, collect responses, arbitrate between competing theories, and synthesize successors, entirely on behavior simulated by Centaur, a foundation model of human behavior. In a multi-attribute decision-making setting, the theories AutoCog found on Centaur generalized to human data: they outperformed canonical theories on ten held-out experiments and were rivaled only by theories found by running the same loop on people. We argue that this succeeds despite the simulator's inevitable imperfections because a discovery loop that arbitrates between competing theories demands less of its simulator than estimation does. The simulator only needs to capture the regularities that distinguish the theories, and not necessarily reproduce behavior precisely. Imperfect simulators can therefore widen the search over theories, with human data then testing whether the surfaced theories generalize.

On-demand platforms face the fundamental challenge of fulfilling time-sensitive jobs with independent workers who may decline offers. To minimize delays and unfulfilled jobs, platforms frequently raise the offered wage sequentially following each rejection. However, the interaction between these dynamic price adjustments and the specific sequence in which workers are approached has been overlooked. In particular, if the best-suited workers (e.g., closest to the job) are also ranked earliest in the sequence, then those workers would see the lowest offered wages and may decline, leading to poor system outcomes where less-suited workers end up seeing the raised wages and accepting the job. We study the sequential offering problem to maximize expected welfare or platform profit by jointly optimizing the ranking of workers and the pricing trajectory. Surprisingly, our main result establishes that if the reservation wage distribution exhibits a non-increasing and convex density function (e.g., Uniform, Exponential), welfare is maximized by greedy ranking and wages optimized via backward induction. For arbitrary distributions, we prove that greedy ranking achieves a tight $n/(2n - 1)$ fraction of the prophet benchmark. Numerical results for settings beyond the distributional assumptions find welfare losses well below those allowed by the universal guarantee, even in families where greedy is provably suboptimal. This suggests that rather than sending initial "low ball” offers to worse matches, platforms should stick with greedy ranking and optimize the wage offerings by appropriately taking the continuation value of the downstream offers into consideration.

We study the control of Markov decision processes in which the quality of a policy is evaluated by a dynamic, time-consistent Markov risk measure rather than by an expected discounted cost. The main obstacle to combining such measures with reinforcement learning is that a transition risk mapping depends on the transition kernel in a nonlinear way, and therefore cannot be estimated from a single observed transition. We remove this obstacle by employing mini-batch transition risk mappings: the mapping is applied to the empirical measure of $N$ independent next-state samples, and the result is averaged. The resulting mapping is again coherent. However, as an expected value of a function of $N$ next-state values, it admits an unbiased one-sample estimator. We embed this mapping into a double deep Q-network, analyze the two sources of estimation bias that arise, and obtain a risk-averse Q-learning method applicable to state spaces far beyond the reach of tabular schemes. The method is applied to an underwater robot navigation problem, in which a vehicle must visit collection points, gather stochastic information payloads, and deliver them at transmission points, while exposed at each step to the risk of destruction. A hierarchical decomposition delegates path execution to an exact graph search and confines learning to the high-level “collect or transmit” decision. A low-dimensional feature map, invariant under the symmetries of the problem, replaces the raw state–configuration encoding. In experiments on $300$ held-out environments, the resulting policies transfer to instance sizes never seen in training, and already $N=2$ reduces the upper semideviation of the outcome distribution while simultaneously improving its mean whenever the simulator is misspecified—an empirical counterpart of the duality between coherent risk measures and distributional robustness.

We characterize the sharp structure-agnostic minimax risk for coefficient estimation in the partial linear model when the outcome and treatment nuisances are learned by two distinct black-box learners, which resolves the open problem in double machine learning posed by Gu (2025). For each nuisance \(q\in\{\mu,\pi\}\), we characterize the available learner by an approximation-error budget \(a_q\) and a stochastic-error budget \(s_q\), with the latter controlled through localized Rademacher complexity. Writing \(\mathcal E_n\) for the minimax mean-squared error, we show that \[\mathcal E_n\asymp1\wedge\left\{\frac1n+\left(a_\mu a_\pi+\min\left\{a_\pi s_\mu+s_\pi^2,\,a_\mu s_\pi+s_\mu^2\right\}\right)^2\right\}.\] The main new ingredient is a novel lower bound for the general two-learner problem. Our proof constructs four finite-mixture testing experiments using orthogonal code functions. Across these experiments, the hidden perturbations are placed outside both learner classes, outside only the treatment learner class, outside only the outcome learner class, or inside both learner classes. These four configurations capture, respectively, the interaction between the two approximation errors, the two asymmetric interactions between one learner's approximation error and the other learner's learning error, and the joint estimation difficulty of learning both nuisances. Combining the four resulting lower bounds yields the displayed rate, which matches the latest upper bound in Gu (2026). Our result shows that standard double machine learning can overstate the intrinsic difficulty of target estimation and provides a target-specific principle for learner selection: approximation error and stochastic complexity must be jointly balanced across the two nuisance learners rather than optimized separately.

Sequential reinforcement learning (RL) solvers and global diffusion model (DM) solvers for neural combinatorial optimization exhibit complementary failure modes under an optimization-regret view. The former enjoys small marginal regret in the early construction stage, but suffers from horizon-wise compounding errors with super-linear regret growth; the latter avoids horizon compounding but incurs linear or sublinear regret w.r.t. the dimension of the remaining unsolved subspace. We propose Hybrid Neural Solver for Combinatorial Optimization (HyCO), a hybrid inference algorithm that constructs a solution prefix with an RL solver and adaptively switches to a conditional DM to complete the remaining decisions. To characterize why such hybridization helps, when to trigger the handover, and how to realize it in practice, we first develop a unified error-scaling theoretical framework and prove that, under explicit error-scaling assumptions, i) the hybrid structure achieves strictly lower expected regret than either backbone alone, and ii) there exists a unique optimal trigger step that minimizes the hybrid regret. We then design a lightweight adaptive trigger that combines policy entropy and RL-DM disagreement to detect trajectory-level signals of the regime shift as a practical proxy, since the optimal trigger step is defined at the expected-regret level and is not directly computable on individual trajectories. Experimental results on diverse benchmarks demonstrate that HyCO achieves consistent improvements over both backbones and support the empirical effectiveness of adaptive triggering. Our code and checkpoints are available at https://github.com/wmd3i/HyCO.

LLM-based digital twins promise to reduce repeated human data collection by generating person- specific responses, yet existing evaluations provide little evidence about whether they can reduce human measurement while preserving valid inference. To address this, we introduce statistical substitutability, an inferential criterion that evaluates the extent to which twin predictions can reduce human measurement for a particular estimand while preserving valid inference. We develop a framework, grounded in mixed-subject and prediction-powered inference, that evaluates statistical substitutability along four dimensions: aggregate fidelity, paired respondent-level signal, finite-sample human-label recovery, and stability across populations. Across two empirical evaluations spanning behavioral experiments, multiple models, and alternative respondent representations, we find that digital twins can reproduce average human effects while providing little information about which individuals differ from those averages. Newer models and richer respondent information improve some dimensions of performance but do not reliably translate into human-data savings. Human calibration can reduce aggregate prediction error, yet limited labeled samples often fail to produce stable precision gains. Importantly, these findings demonstrate that behavioral fidelity is neither necessary nor sufficient for statistical substitutability. More broadly, they suggest that AI-generated evidence should be evaluated based on its ability to support valid scientific inference rather than its ability to reproduce human outcomes alone. Digital twins should therefore be judged for confirmatory use by whether they reduce uncertainty about human quantities, not merely by whether they reproduce human means, distributions, or effects.

Purpose: Accurate CT protocol selection is critical for diagnostic quality and patient safety, yet the current process is manual, time-consuming, and prone to inconsistencies. Prior Machine Learning methods using keywords or bag-of-words lack contextual understanding and perform poorly on rare protocols. We propose a decision support system using large language model (LLM) features to recommend protocols from free-text clinical indications, capturing clinical nuance and phrasing variation for more consistent, efficient selection. Methods: In this REB-approved retrospective study, 285,123 chest CT imaging requests from a large academic medical center (2017-2024) were split into training (228,099, 80%) and held-out test (57,024, 20%) sets. Each request included procedure names, clinical indication, HIS comments, and the selected protocol. Clinical text was embedded using a fine-tuned LLM, Meta's LLaMA-3.1-70B; these features input a logistic regression classifier predicting 18 protocol labels (e.g., PE, LDCT). Results: The pipeline achieved a weighted precision of 0.84, weighted F1-score of 0.81, and overall accuracy of 79% across 18 CT protocols. On 300 independent cases with expert consensus, the LLM reached an overall accuracy of 80% versus 83% for radiologists, with no significant difference (p = 0.263). Performance was comparable across most classes, with the LLM exceeding radiologists for some challenging categories, and entropy analyses indicated more balanced protocol use, suggesting reduced variability. Conclusion: An LLM-based recommendation system can leverage general knowledge from a large natural-text corpus to accurately assign chest CT protocols from free-text imaging requests, and may serve as a viable foundation for protocol recommendation tools where inputs require language understanding.

Process mining has long turned event logs into process knowledge: discovered models, conformance evidence, bottleneck diagnoses, and runtime predictions. Agentic AI changes the target. Process-aware agents will not only ask what happened. They will ask whether a proposed action should be taken, given the available evidence, privacy budget, organizational authority, and downstream risk. This BlueSky paper proposes event-to-action process mining: a process-mining agenda for transforming heterogeneous operational event data into governed action. The goal is not another dashboard, a generic enterprise simulator, or a language interface over logs. We argue that the community needs four mineable artifacts: event-object representations, action evidence packages, governance contracts, and benchmarks where act, defer, ask, and refuse are all valid outputs. This agenda is timely because agentic business process management (BPM), LLM-assisted process mining, object-centric event standards, causal process monitoring, and privacy-preserving learning are maturing separately. Bringing them together defines a data-mining target inside process mining: mining logged organizational behavior for accountable action, not only retrospective insight.

Physics-Informed Neural Networks (PINNs) have recently emerged as a promising approach for solving Partial Differential Equations (PDEs), offering a meshfree alternative that integrates physical principles into the learning process. This presents a new paradigm compared to traditional discretization methods and purely data-driven machine learning techniques. While promising, PINNs are not a panacea; they inherit challenges such as spectral bias and unstable convergence. Moreover, their potential in seismology remains largely unexplored. In this work, we provide a robust and critical assessment of PINNs for solving the elastic wave equation in seismology. We investigate the performance of PINNs on problems with varying degrees of complexity across various seismic sources and parameter models, from constant to highly heterogeneous settings. A pivotal aspect of our work involves investigating whether embedding physical principles directly into the network architecture enhances convergence and accuracy. We test an extensive range of neural architecture designs, from unrestricted, uninformed PINNs to highly specialized ones. We find that integrating an understanding of wave physics into the network design significantly improves accuracy. For instance, introducing a custom wavelet or plane wave layer, coupled with encoder and decoder layers, consistently yields a relative $L_2$ error approximately half that of the standard PINN, as evidenced across numerous experiments. We further demonstrate that this novel architecture enhances accuracy when applied to the acoustic wave equation, underlying the versatility of our network. Another key contribution of our research is the successful conditioning of PINNs on seismic source locations. This signifies a considerable advancement towards rapid seismic hazard detection and seismic analysis.

Standard semi-supervised learning (SSL) typically relies on labelled and unlabelled data sharing a common marginal distribution. This assumption is often violated by biased spatial sampling mechanism, when labels are collected under spatially biased or preferential site selection. We treat this marginal mismatch, spatial autocorrelation, and spatial non-stationarity as three distinct mechanisms, varied independently via a labelled-sampling concentration parameter, a spatial length scale, and a non-stationarity strength parameter, and ask how mismatch degrades SSL, whether the cluster and manifold assumptions survive it, and how the resulting failure can be diagnosed. Using a controlled synthetic framework alongside PovertyMap-WILDS, California housing, socio-economic and US air quality monitoring datasets, we systematically vary the degree of mismatch while accounting for spatial autocorrelation and non-stationarity. Through a series of analyses including a segmented-regression changepoint, we show that in the synthetic generator, SSL performance does not degrade gradually but instead exhibits a threshold-like breakdown between approximately 0.71 and 0.77 once distribution mismatch becomes sufficiently severe. We further demonstrate that spatial non-stationarity contributes to performance loss independently of marginal mismatch and that models become increasingly overconfident outside the regions where labels are available. To support practical deployment, we evaluate several distribution-divergence measures as indicators of reliability and introduce a kernel-weighted local divergence metric that provides a more stable estimate of spatial mismatch than a naïve localised approach. These findings provide empirical evidence and diagnostic tools for better documenting the risk of incorporating unlabelled spatial data into semi-supervised learning workflows.

Audio provenance attribution - which system produced a synthetic utterance - is reported at near-ceiling accuracy on clean benchmarks, yet audio reaching an analyst has usually been transcoded. We report a prospectively registered measurement of closed-set attribution after single-stage codec transport, with the analysis region fixed from fidelity metadata before any attribution model was trained. On two corpora, in-support losses reach 53.5 [43.5, 63.6] and 70.3 [63.0, 77.5] Macro-F1 points for WavLM-Base+, and 61.0 [56.8, 65.1] and 49.8 [41.6, 57.9] for W2V2-BERT 2.0, under simultaneous component-level bands. Degradation is strongly condition- and representation-dependent: within one in-support grid WavLM losses run from -0.4 to +53.5 points, and the two encoders differ beyond a prespecified +/-5-point margin at six of twelve conditions. A clean-qualified ECAPA-TDNN and a Proxy-Anchor head degrade comparably, so the effect is not confined to one representation family or a weak linear head. The registered matched-fidelity comparison was not estimable on this grid, and waveform and perceptual measures order the conditions differently: MP3 at 8 kbit/s ranks mid-grid on SI-SDR but last on PESQ-WB while causing the largest loss. For the tested tasks, corpora, representations and codec grid, a clean accuracy figure does not by itself characterise deployment robustness.

The Delaunay triangulation is a classical geometric structure on a finite set of points in the plane. Its stretch factor is the largest ratio of the shortest-path distance along its edges to the Euclidean distance between the same two points. Determining the worst-case stretch factor of planar Delaunay triangulations is a long-standing open problem. We prove a uniform upper bound of 1.5949, improving the previous bound of 1.998 and narrowing the gap to the best known lower bound of 1.5932 from 0.4048 to less than 0.0017, nearly pinpointing the worst-case stretch factor. A central difficulty is that the line segment between two points may cross arbitrarily many triangles. Our progress accounting framework addresses this by converting the global bound into a family of local inequalities, each involving only a fixed number of real parameters regardless of the number of triangles crossed. Two further geometric lemmas make computer-assisted verification tractable: the pencil lemma decomposes each local transition into two geometrically independent stages, reducing verification over infinitely many possible next disks to a single differential condition; and the corner lemmas handle degenerate boundary cases where direct verification would fail. With these reductions, the proof ends with a finite verification of an explicit potential in exact rational arithmetic and outward-rounded interval arithmetic. Numerous geometric network constructions use the Delaunay stretch factor as a black box, so the tighter bound directly improves stretch-factor guarantees for a range of downstream applications, including bounded-degree plane spanners and low-weight spanners.

Bringing multiscale geometric analysis directly to irregular point clouds remains difficult: quantities such as local dimension, anisotropy, density variation, and geometric transitions are typically estimated through explicit neighborhood, manifold, or graph constructions, or left for neural networks to infer from coordinates. We introduce Heat Field Signatures (HFS), which lift a point cloud to a multiscale family of smooth ambient heat fields, providing a direct interface from discrete samples to geometric analysis. From this field, HFS computes closed-form global and local signatures directly from pairwise distances, capturing heat concentration, intrinsic dimension, anisotropy, and scale transitions. We further introduce the Heat Dimension Spectrum (HDS), a compact summary of multiscale geometric composition. HFS can be used as a closed-form descriptor, a lightweight learned representation, or a geometric feature channel for neural point-cloud models. Across synthetic and real-world benchmarks spanning subcellular, neuronal, tree, and protein data, HFS outperforms strong point-cloud and multiparameter-persistence baselines while substantially reducing end-to-end cost. On SCOP protein-fold classification, HFS improves over the strongest deep baseline by nearly $24$ percentage points using coordinates alone, while standalone HFS representations are exactly rotation-invariant by construction. More broadly, HFS turns a classical heat field into a practical interface for multiscale geometric analysis in modern point-cloud learning.