Numerical simulation of wave propagation and run-up is a cornerstone of coastal engineering and tsunami hazard assessment. However, applying these forward models to inverse problems, such as bathymetry estimation, source inversion, and structural optimization, remains notoriously difficult due to the rigidity and high computational cost of deriving discrete adjoints. In this paper, we introduce AegirJAX, a fully differentiable hydrodynamic solver based on the depth-integrated, non-hydrostatic shallow-water equations. By implementing the solver entirely within a reverse-mode automatic differentiation framework, AegirJAX treats the time-marching physics loop as a continuous computational graph. We demonstrate the framework's versatility across a suite of scientific machine learning tasks: (1) discovering regime-specific neural corrections for model misspecifications in highly dispersive wave propagation; (2) performing continuous topology optimization for breakwater design; (3) training recurrent neural networks in-the-loop for active wave cancellation; and (4) inverting hidden bathymetry and submarine landslide kinematics directly from downstream sensor data. The proposed differentiable paradigm fundamentally blurs the line between forward simulation and inverse optimization, offering a unified, end-to-end framework for coastal hydrodynamics.

In the human-AI collaboration area, the context formed naturally through multi-turn interactions is typically flattened into a chronological sequence and treated as a fixed whole in subsequent reasoning, with no mechanism for dynamic organization and management along the collaboration workflow. Yet these contexts differ substantially in lifecycle, structural hierarchy, and relevance. For instance, temporary or abandoned exchanges and parallel topic threads persist in the limited context window, causing interference and even conflict. Meanwhile, users are largely limited to influencing context indirectly through input modifications (e.g., corrections, references, or ignoring), leaving their control neither explicit nor verifiable. To address this, we propose Mixed-Initiative Context, which reconceptualizes the context formed across multi-turn interactions as an explicit, structured, and manipulable interactive object. Under this concept, the structure, scope, and content of context can be dynamically organized and adjusted according to task needs, enabling both humans and AI to actively participate in context construction and regulation. To explore this concept, we implement Contextify as a probe system and conduct a user study examining users' context management behaviors, attitudes toward AI initiative, and overall collaboration experience. We conclude by discussing the implications of this concept for the HCI community.

AI coding assistants are now used to generate production code in security-sensitive domains, yet the exploitability of their outputs remains unquantified. We address this gap with Broken by Default: a formal verification study of 3,500 code artifacts generated by seven widely-deployed LLMs across 500 security-critical prompts (five CWE categories, 100 prompts each). Each artifact is subjected to the Z3 SMT solver via the COBALT analysis pipeline, producing mathematical satisfiability witnesses rather than pattern-based heuristics. Across all models, 55.8% of artifacts contain at least one COBALT-identified vulnerability; of these, 1,055 are formally proven via Z3 satisfiability witnesses. GPT-4o leads at 62.4% (grade F); Gemini 2.5 Flash performs best at 48.4% (grade D). No model achieves a grade better than D. Six of seven representative findings are confirmed with runtime crashes under GCC AddressSanitizer. Three auxiliary experiments show: (1) explicit security instructions reduce the mean rate by only 4 points; (2) six industry tools combined miss 97.8% of Z3-proven findings; and (3) models identify their own vulnerable outputs 78.7% of the time in review mode yet generate them at 55.8% by default.

The transformative potential of large language models (LLMs) in education, such as improving accessibility and personalized learning, is being eclipsed by significant challenges. These challenges stem from concerns that LLMs undermine academic assessment by enabling bypassing of critical thinking, leading to increased cognitive offloading. This emerging trend stresses the dual imperative of harnessing AI's educational benefits while safeguarding critical thinking and academic rigor in the evolving AI ecosystem. To this end, we introduce AI-Sinkhole, an AI-agent augmented DNS-based framework that dynamically discovers, semantically classifies, and temporarily network-wide blocks emerging LLM chatbot services during proctored exams. AI-Sinkhole offers explainable classification via quantized LLMs (LLama 3, DeepSeek-R1, Qwen-3) and dynamic DNS blocking with Pi-Hole. We also share our observations in using LLMs as explainable classifiers which achieved robust cross-lingual performance (F1-score > 0.83). To support future research and development in this domain initial codes with a readily deployable 'AI-Sinkhole' blockist is available on https://github.com/AIMLEdu/ai-sinkhole.

This study presents a conditional flow matching framework for solving physics-constrained Bayesian inverse problems. In this setting, samples from the joint distribution of inferred variables and measurements are assumed available, while explicit evaluation of the prior and likelihood densities is not required. We derive a simple and self-contained formulation of both the unconditional and conditional flow matching algorithms, tailored specifically to inverse problems. In the conditional setting, a neural network is trained to learn the velocity field of a probability flow ordinary differential equation that transports samples from a chosen source distribution directly to the posterior distribution conditioned on observed measurements. This black-box formulation accommodates nonlinear, high-dimensional, and potentially non-differentiable forward models without restrictive assumptions on the noise model. We further analyze the behavior of the learned velocity field in the regime of finite training data. Under mild architectural assumptions, we show that overtraining can induce degenerate behavior in the generated conditional distributions, including variance collapse and a phenomenon termed selective memorization, wherein generated samples concentrate around training data points associated with similar observations. A simplified theoretical analysis explains this behavior, and numerical experiments confirm it in practice. We demonstrate that standard early-stopping criteria based on monitoring test loss effectively mitigate such degeneracy. The proposed method is evaluated on several physics-based inverse problems. We investigate the impact of different choices of source distributions, including Gaussian and data-informed priors. Across these examples, conditional flow matching accurately captures complex, multimodal posterior distributions while maintaining computational efficiency.

Deploying Large Language Model (LLM) services at the edge benefits latency-sensitive and privacy-aware applications. However, the stateless nature of LLMs makes managing user context (e.g., sessions, preferences) across geo-distributed edge nodes challenging. Existing solutions, such as client-side context storage, introduce network latency and bandwidth overhead, undermining edge deployment advantages. We propose DisCEdge, a distributed context management system that stores and replicates user context in tokenized form across edge nodes. By maintaining context as token sequences, our system avoids redundant computation and enables efficient data replication. We evaluate an open-source prototype in a realistic edge environment. DisCEdge improves median response times by up to 14.46% and lowers median inter-node synchronization overhead by up to 15% compared to a raw-text-based system. It also reduces client request sizes by a median of 90% compared to client-side context management, while guaranteeing data consistency.

This research proposes an extensive technique for monitoring and controlling the industrial parameters using Internet of Things (IoT) technology based on wireless communication. We proposed a system based on NRF transceivers to establish a strong Wireless Sensor Network (WSN), enabling transfer of real-time data from multiple sensors to a central setup that is driven by ARDUINO microcontrollers. Different key parameters, crucial for industrial setup such as temperature, humidity, soil moisture and fire detection, are monitored and displayed on an LCD screen, enabling factory administration to oversee the industrial operations remotely over the internet. Our proposed system bypasses the need for physical presence for monitoring by addressing the shortcomings of conventional wired communication systems. Other than monitoring, there is an additional feature to remotely control these parameters by controlling the speed of DC motors through online commands. Given the rising incidence of industrial fires over the worldwide between 2020 and 2024 due to an array of hazards, this system with dual functionality boosts the overall operational efficiency and safety. This overall integration of IoT and Wireless Sensor Network (WSN) reduces the potential risks linked with physical monitoring, providing rapid responses in emergency scenarios, including the activation of firefighting equipment. The results show that innovations in wireless communication perform an integral part in industrial process automation and safety, paving the way to more intelligent and responsive operating environments. Overall, this study highlights the potential for change of IoT-enabled systems to revolutionize monitoring and control in a variety of industrial applications, resulting in increased productivity and safety.

Colorectal cancer liver metastasis (CRLM) exhibits high postoperative recurrence and pronounced prognostic heterogeneity, challenging individualized management. Existing prognostic approaches often rely on static representations from a single postoperative snapshot, and fail to jointly capture tumor spatial distribution, longitudinal disease dynamics, and multimodal clinical information, limiting predictive accuracy. We propose DyPro, a deep learning framework that infers postoperative latent trajectories via residual dynamic evolution. Starting from an initial patient representation, DyPro generates a 12-step sequence of trajectory snapshots through autoregressive residual updates and integrates them to predict recurrence and survival outcomes. On the MSKCC CRLM dataset, DyPro achieves strong discrimination under repeated stratified 5-fold cross-validation, reaching a C-index of 0.755 for OS and 0.714 for DFS, with OS AUC@1y of 0.920 and OS IBS of 0.143. DyPro provides quantitative risk cues to support adjuvant therapy planning and follow-up scheduling.

The continuous monitoring of the interactions between cyber-physical components of any industrial control system (ICS) is required to secure automation of the system controls, and to guarantee plant processes are fail-safe and remain in an acceptably safe state. Safety is achieved by managing actuation (where electric signals are used to trigger physical movement), dependent on corresponding sensor readings; used as ground truth in decision making. Timely detection of anomalies (attacks, faults and unascertained states) in ICSs is crucial for the safe running of a plant, the safety of its personnel, and for the safe provision of any services provided. We propose an anomaly detection method that involves accurate linearization of the non-linear forms arising from sensor-actuator(s) relationships, primarily because solving linear models is easier and well understood. We accomplish this by using a well-known water treatment testbed as a use case. Our experiments show millisecond time response to detect anomalies, all of which are explainable and traceable; this simultaneous coupling of detection speed and explainability has not been achieved by other state of the art Artificial Intelligence (AI)/ Machine Learning (ML) models with eXplainable AI (XAI) used for the same purpose. Our methods explainability enables us to pin-point the sensor(s) and the actuation state(s) for which the anomaly was detected. The proposed algorithm showed an accuracy of 97.72% by flagging deviations within safe operation limits as non-anomalous; indicative that slower detectors with highest detection resolution is unnecessary, for systems whose safety boundaries provide leeway within safety limits.

Denoising diffusions sample from a probability distribution $μ$ in $\mathbb{R}^d$ by constructing a stochastic process $({\hat{\boldsymbol x}}_t:t\ge 0)$ in $\mathbb{R}^d$ such that ${\hat{\boldsymbol x}}_0$ is easy to sample, but the distribution of $\hat{\boldsymbol x}_T$ at large $T$ approximates $μ$. The drift ${\boldsymbol m}:\mathbb{R}^d\times\mathbb{R}\to\mathbb{R}^d$ of this diffusion process is learned my minimizing a score-matching objective. Is every probability distribution $μ$, for which sampling is tractable, also amenable to sampling via diffusions? We provide evidence to the contrary by studying a probability distribution $μ$ for which sampling is easy, but the drift of the diffusion process is intractable -- under a popular conjecture on information-computation gaps in statistical estimation. We show that there exist drifts that are superpolynomially close to the optimum value (among polynomial time drifts) and yet yield samples with distribution that is very far from the target one.

We study the kernel instrumental variable (KIV) algorithm, a kernel-based two-stage least-squares method for nonparametric instrumental variable regression. We provide a convergence analysis covering both identified and non-identified regimes: when the structural function is not identified, we show that the KIV estimator converges to the minimum-norm IV solution in the reproducing kernel Hilbert space associated with the kernel. Crucially, we establish convergence in the strong $L_2$ norm, rather than only in a pseudo-norm. We quantify statistical difficulty through a link condition that compares the covariance structure of the endogenous regressor with that induced by the instrument, yielding an interpretable measure of ill-posedness. Under standard eigenvalue-decay and source assumptions, we derive strong $L_2$ learning rates for KIV and prove that they are minimax-optimal over fixed smoothness classes. Finally, we replace the stage-1 Tikhonov step by general spectral regularization, thereby avoiding saturation and improving rates for smoother first-stage targets. The matching lower bound shows that instrumental regression induces an unavoidable slowdown relative to ordinary kernel ridge regression.

Mobile apps are essential in daily life but frequently employ deceptive patterns, such as visual emphasis or linguistic nudging, to manipulate user behavior. Existing research largely relies on manual detection, which is time-consuming and cannot keep pace with rapidly evolving apps. Although recent work has explored automated approaches, these methods are limited to intra-page patterns, depend on manual app exploration, and lack flexibility. To address these limitations, we present AppRay, a system that integrates task-oriented app exploration with automated deceptive pattern detection to reduce manual effort, expand detection coverage, and improve performance. AppRay operates in two stages. First, it combines large language model-guided task-oriented exploration with random exploration to capture diverse user interface (UI) states. Second, it detects both intra-page and inter-page deceptive patterns using a contrastive learning-based multi-label classifier augmented with a rule-based refiner for context-aware detection. We contribute two datasets, AppRay-Tainted-UIs and AppRay-Benign-UIs, comprising 2,185 deceptive pattern instances, including 149 intra-page cases, spanning 16 types across 876 deceptive and 871 benign UIs, while preserving UI relationships. Experimental results show that AppRay achieves macro/micro averaged precision of 0.92/0.85, recall of 0.86/0.88, and F1 scores of 0.89/0.85, yielding 27.14% to 1200% improvements over prior methods and enabling effective detection of previously unexplored deceptive patterns.

As more applications of large language models (LLMs) for 3D content for immersive environments emerge, it is crucial to study user behaviour to identify interaction patterns and potential barriers to guide the future design of immersive content creation and editing systems which involve LLMs. In an empirical user study with 12 participants, we combine quantitative usage data with post-experience questionnaire feedback to reveal common interaction patterns and key barriers in LLM-assisted 3D scene editing systems. We identify opportunities for improving natural language interfaces in 3D design tools and propose design recommendations for future LLM-integrated 3D content creation systems. Through an empirical study, we demonstrate that LLM-assisted interactive systems can be used productively in immersive environments.

State-of-the-art large language models (LLMs) are typically deployed as online services, requiring users to transmit detailed prompts to cloud servers. This raises significant privacy concerns. In response, we introduce ConfusionPrompt, a novel framework for private LLM inference that protects user privacy by: (i) decomposing the original prompt into smaller sub-prompts, and (ii) generating pseudo-prompts alongside the genuine sub-prompts, which are then sent to the LLM. The server responses are later recomposed by the user to reconstruct the final output. This approach offers key advantages over previous LLM privacy protection methods: (i) it integrates seamlessly with existing black-box LLMs, and (ii) it delivers a significantly improved privacy-utility trade-off compared to existing text perturbation methods. We also develop a $(λ, μ, ρ)$-privacy model to formulate the requirements for a privacy-preserving group of prompts and provide a complexity analysis to justify the role of prompt decomposition. Our empirical evaluation shows that ConfusionPrompt achieves significantly higher utility than local inference methods using open-source models and perturbation-based techniques, while also reducing memory consumption compared to open-source LLMs.

We derive general trade-off relations among the power, efficiency, and constancy for two-terminal thermoelectric systems in the linear response regime. Constancy, which quantifies the steadiness of the heat engine, is measured by its fluctuations. The bounds of the efficiency, power and fluctuations are valid even when time-reversal symmetry is broken, revealing how such a symmetry breaking alters the fundamental constraints on steady-state energy conversion. Our results extend and refine previously established universal trade-offs, offering deeper insight into the performance limits in nonequilibrium thermodynamics. Guided by this bound, heat engines with broken time-reversal symmetry can be operated at near-Carnot efficiency while maintaining finite power output and fluctuations, enabling them to outperform their traditional counterparts.

This paper develops a rich theory of cardinality in the paraconsistent and paracomplete set theory $\mathrm{BZFC}$, where sets can be inconsistent ($A$ such that ``$x\in A$'' is both true and false for some $x$) or incomplete ($A$ such that ``$x\in A$'' is neither true nor false for some $x$). We carefully analyze what it means for two potentially incomplete or inconsistent sets to have ``the same size'', construct the corresponding cardinal numbers, and develop the basic theory of cardinal arithmetic. A surprising result is that the cardinality of any set can be expressed as a linear combination of three fundamental cardinal numbers with classical cardinals as coefficients. In that sense, our cardinal numbers form a three-dimensional space over the usual cardinals, much like how the complex numbers form a two-dimensional space over the reals.

We postulate that the excesses of di-dijet events observed at the LHC are attributed to the production of four fourth-generation quarks $b'$ with a mass $m_{b'}\approx 2$ TeV at few-TeV scales. The di-dijet signals around the four-jet invariant mass $m_{4j}\approx 8$ TeV arise from a resonant $b'b'\bar b'\bar b'$ tetraquark production, where the dijet resonances of masses about 2 TeV correspond to $b'\bar b'$ first excited states (color-octet scalars with the principal quantum number $n=2$) in a Yukawa potential created by Higgs boson exchanges. Those around $m_{4j}\approx 3.6$ TeV originate from a non-resonant $b'b'\bar b'\bar b'$ production, where the dijet resonances of masses 0.95 TeV correspond to $b'\bar b'$ ground states (color-octet vectors with $n=1$). It is shown that a $b'\bar b'$ system with $m_{b'}\approx 2$ TeV in the Yukawa potential does generate the aforementioned bound state spectrum. We then illustrate that the observed excesses can be accommodated in our setup by translating the fourth-generation model to the effective theories containing color-octet scalars and vectors available in the literature. The di-dijet events at $m_{4j}= 6.6$ TeV and 5.8 TeV with dijet masses about 2 TeV can also be interpreted in the same framework. Simply speaking, our scenario can be viewed as a TeV-scale version of the search for a fully charmed tetraquark via the four-muon channels $X(6900)\to (c\bar c)(c\bar c)\to 4μ$ at a GeV scale.

We report MINERvA's first measurement of inclusive charged-current $\barν_μ$ cross sections on carbon, hydrocarbon, iron, and lead, and their ratios to the cross section on hydrocarbon, as functions of the antimuon transverse momentum, $p_{\mathrm{T}}$. Using a wide-band $\barν_μ$ beam with mean energy $\sim 6~\text{GeV}$, these measurements probe all interaction modes, including the transition from resonance production to deep-inelastic scattering. The total uncertainties are typically $5-10\%$ for the absolute cross sections and $2-5\%$ for the ratios. Comparisons with multiple neutrino interaction models reveal significant discrepancies in the $p_{\mathrm{T}}$ dependence, particularly for heavier nuclei. The disagreements are most pronounced at low $p_{\mathrm{T}}$ but extend across the full $p_{\mathrm{T}}$ range, indicating missing or mis-modelled nuclear effects.

In this study, energy backflow in the Poynting vector, as well as its orbital and spin current density components, has been examined for a 2-dimensional causal unidirectional vector-valued monochromatic electromagnetic wave. Linear transverse electric (TE), transverse magnetic (TM), and circular polarization cases are considered and studied in detail, including both electric and magnetic contributions to the current density components. Spin current backflow has been found to be unexpectedly strong. A study of the energy backflow is also presented in the scalar version of the 2-dimensional monochromatic wave. A detailed study has been carried out of the correlation of the positions of energy backflows with local wavenumbers and their signs, the zeros of appropriate intensities and the presence of vortices.

Homeomorphisms of the Cantor set play a central role in topology, dynamical systems and descriptive set theory. In parallel, several classes of fence-like spaces - such as the hairy Cantor set, hairy arcs, Cantor bouquets in complex dynamics, the Lelek fan in topology and Fraïssé fence in descriptive set theory - have recently been studied for their rich structural and dynamical properties. In this paper, we introduce a general construction that associates to each homeomorphism of the Cantor set a canonically defined homeomorphism of a corresponding fence space. This construction lifts dynamical properties from the Cantor set to these fence-like spaces, allowing one to systematically transfer features such as minimality, recurrence, and orbit structure. As a consequence, we obtain a unified framework for studying dynamics on a broad class of fence-like spaces and establish new connections between their topological structure and induced dynamical behavior.

The Holevo limit bounds the channel capacity of a communication channel in which information is encoded in quantum states in a Hilbert space at the transmitter and decoded using quantum measurements at the receiver. Saturating the Holevo limit requires quantum-limited transceivers that either generate quantum states of light or employ quantum-limited measurements. Here, we demonstrate an integrated photonic-electronic quantum-limited coherent receiver (QRX) achieving 14.0 dB shot noise clearance (SNC), 520 $μ$W knee power, 2.57 GHz 3-dB bandwidth, 3.50 GHz shot-noise-limited bandwidth, and 90.2 dB common-mode rejection ratio ($\mathrm{CMRR}$). We scale this design to a 32-channel QRX array with median 26.6 dB $\mathrm{SNC}$, and automatic $\mathrm{CMRR}$ correction yielding a median 76.8 dB $\mathrm{CMRR}$ at minimum. Using the integrated QRX and fiber-optic transmitter, we measure $0.15\pm0.01$ dB of squeezing below the shot noise limit, limited by off-chip losses. We propose a squeezed light communication scheme that can surpass the Shannon limit, with a path toward the Holevo limit.

Neural rendering paradigms have recently emerged as powerful tools for radio frequency (RF). However, by entangling RF sources with scene geometry and material properties, existing approaches limit downstream manipulation of scene geometry, wireless system configuration, and RF reasoning. To address this, we propose a physically grounded RF inverse rendering (RFIR) framework that explicitly decouples RF emission, geometry, and material electromagnetic properties. Our key insight is an RF-aware bidirectional scattering distribution function, embedded into the Gaussian splatting paradigm as an RF rendering equation. Each Gaussian primitive is endowed with intrinsic physical attributes, including surface normals, material electromagnetic parameters, and roughness, and leveraged by a customized ray-tracing scheme to represent RF signal synthesis. The proposed RFIR generalizes three typical RF tasks: radar cross-section synthesis, received signal strength indicator prediction, and wireless scene editability. Experiments demonstrate significant performance advantages, underscoring the potential for wireless world modeling.

Horizontal stirred bed reactors (HSBRs) are used in gas--phase polyolefin production, where efficient solids mixing and controlled residence time distributions are essential for product quality and stability. Despite their industrial relevance, the influence of operating conditions on granular flow and mixing in HSBRs is not well understood. Discrete Element Method (DEM) simulations are used to study the effects of rotation speed and fill level on particle motion, mixing, and axial transport in a lab--scale HSBR. An industrial--grade polypropylene powder is modelled using calibrated contact parameters. Mixing is quantified using the Lacey index in axial (z) and cross--sectional (xy) directions. Particle circulation is characterised via cycle--time analysis and a coarse--grained angular velocity field. Axial dispersion coefficients are obtained from particle trajectories using both Einstein--type and cycle--based approaches, and validated with a diffusion model predicting the axial Lacey index. Results show that axial mixing depends strongly on rotation speed and fill level: higher rotation speeds accelerate homogenization, while higher fill levels slow mixing. Cross--sectional mixing is mainly sensitive to rotation speed, with fill--level effects diminishing at higher speeds. Cycle time decreases with increasing rotation speed and fill level, indicating enhanced circulation. Axial dispersion increases with rotation speed but decreases with fill level, with consistent results across methods. These findings reveal trade--offs between axial mixing, circulation, and dispersion, highlighting the need to balance operating conditions and demonstrating the capability of DEM to support HSBR optimisation.

We provide a detectability analysis for nonlinear large-scale distributed systems in the sense of exponential incremental input/output-to-state stability (i-IOSS). In particular, we prove that the overall system is exponentially i-IOSS if each subsystem is i-IOSS, with interconnections treated as external inputs, and a suitable small-gain condition holds. The analysis is extended to a Lyapunov characterization, resulting in a different quantitative outcome regarding the small-gain condition, which is further analyzed within this work. Moreover, we derive linear matrix inequality conditions posed solely on the local subsystems and their interconnections, which guarantee exponential i-IOSS of the overall distributed system. The results are illustrated on a numerical example.

Machines are designed, assembled, and programmed to convert power into predetermined dynamics and functions. In contrast, living systems such as interacting cells and animal groups self-organize, synchronize, and perform complex tasks without predefined patterns. Inspired by these decentralized architectures, experiments have shown that small assemblies of elastically coupled self-propelled robots can achieve two fundamental functionalities observed in nature: collective motion and oscillatory deformations. However, biological inspiration has steered research toward translational self-propulsion, while active rotation remains an underexplored route to designing broader animate materials. Here, we study the self-organization of microscopic metamachines composed of thousands of 3D-printed rotary motors. We first demonstrate and explain how motors precessing in unspecified directions collectively arrange their dynamics into a pristine antiferromagnetic phase. Next, we elucidate the emergence of spatiotemporal order in the form of phase coherence in the rotors' precession. Finally, we show how quenched disorder initiates the free propagation of phase waves across self-organized regions with mismatched rotation speeds. Our results suggest that spinner-based metamachines could illuminate metachronal-wave formation in living systems, and signal propagation in synthetic animate materials.

Multimodal retrieval over text corpora remains a fundamental challenge: the best vision-language encoder achieves only 27.6 nDCG@10 on MM-BRIGHT, a reasoning-intensive multimodal retrieval benchmark, underperforming strong text-only systems. We argue that effective multimodal retrieval requires three tightly integrated capabilities that existing approaches address only in isolation: expanding the query's latent intent, retrieving with a model trained for complex reasoning, and reranking via explicit step-by-step reasoning over candidates. We introduce \textbf{MARVEL} (\textbf{M}ultimodal \textbf{A}daptive \textbf{R}easoning-intensi\textbf{V}e \textbf{E}xpand-rerank and retrieva\textbf{L}), a unified pipeline that combines LLM-driven query expansion, \textbf{MARVEL-Retriever} -- a reasoning-enhanced dense retriever fine-tuned for complex multimodal queries -- and GPT-4o-based chain-of-thought reranking with optional multi-pass reciprocal rank fusion. Evaluated on MM-BRIGHT across 29 technical domains, MARVEL achieves \textbf{37.9} nDCG@10, surpassing the best multimodal encoder by \textbf{+10.3 points} and outperforming all single-stage baselines in 27 of 29 domains and matching or approaching the best baseline in the remaining two highly-specialized domains (Crypto, Quantum Computing), demonstrating that reasoning-intensive multimodal retrieval is best addressed through a unified expand-retrieve-rerank framework. https://github.com/mm-bright/multimodal-reasoning-retrieval

The study of stronger-than-quantum phenomena (i.e., postquantum) has enabled a deeper understanding of the scope of quantum theory. Much is known about the case of correlations in Bell scenarios, where the device-independent framework allowed us to explore its possibilities independently of the formalism of quantum theory. However, less is known about the phenomenon of Einstein-Podolsky-Rosen steering. Here, the `characterised parties' are assumed to describe their systems locally through the quantum formalism, which inconveniences a theory-independent description. In addition, a theorem by Gisin and Hughston, Josza and Wootters further hindered the discovery of the phenomenon. The study of postquantum steering, initiated about a decade ago, has been quite fruitful, including: the development of mathematical formalisms that frame the effect, resource theories that quantify it as a resource, and activation protocols that relate it to Bell correlations. However, all these results have a limitation in common: they apply to scenarios with only one quantum party. Here we articulate the concept of postquantum steering for scenarios with multiple quantum parties, bringing in the missing piece to the puzzle. We provide an algorithm to certify postquantumness, which in some cases also certifies quantumness. We also define a hierarchy of semidefinite programs that bounds the set of quantum assemblages from the outside. Moreover, we show that the study of postquantum steering is fundamentally relevant since it is not just a mere mathematical curiosity allowed by the no-signalling principle, but it may arise within compositional theories beyond quantum theory. Our work further discovers a peculiarity of steering: its theory-independent description fundamentally prevents a direct connection with Bell nonlocality -- e.g., nonclassical Bell correlations do not imply nonclassical steering.

The solid electrolyte interphase SEI critically dictates the cyclability and Coulombic efficiency of sodium-metal batteries, yet its dynamic formation mechanisms and atomic-scale evolution during electrochemical cycling remain elusive due to the spatiotemporal limitations of existing techniques. Here, an "Interface Reactor" sampling strategy is proposed to construct a charge-aware neuroevolution potential (qNEP). This approach overcomes the instability bottlenecks of conventional machine learning potentials, enabling stable, first-principles-accurate molecular dynamics simulations of complex electrode-electrolyte interfaces on the hundred-nanosecond scale. Fundamentally distinct SEI formation mechanisms are revealed during the early stage: carbonate-based electrolytes form heterogeneous organic-inorganic matrices via "mixed co-formation," whereas ether-based electrolytes generate dense, self-limiting inorganic barriers through "surface-energy-controlled" NaF crystallization. Metadynamics simulations further elucidate that these compositional disparities govern sodium-ion storage dynamics: NaF-rich SEIs facilitate efficient metallic deposition, while carbonate-dominated interphases induce irreversible sodium trapping and continuous electrolyte decomposition. These findings establish a comprehensive atomic-scale framework linking solvation structure, interfacial reaction networks, and electrochemical performance, providing mechanistic guidelines for rational SEI engineering in next-generation alkali-metal batteries. Crucially, a general and robust computational framework is established for simulating complex interfacial reactions in electrochemical systems.

We present a JWST/NIRSpec-IFU study of metallicity gradients in seven low-metallicity systems at $z=7.2-9.5$. The main sample spans stellar masses of $\rm \log(M_*/M_{\odot}) \sim 7.8-9.5$, star formation rates (SFRs) of $\rm \log(\text{SFR} / M_{\odot} \text{yr}^{-1}) \sim 0.5-2.5$, and gas-phase metallicities of $4\%-15 \%~Z_\odot$. Within our sample, we also identify three low-metallicity satellite galaxies associated with two of our sources, providing a rare view of early-epoch interactions. The three satellites exhibit even more primordial properties, with metallicity $3\% -4\% ~Z_\odot$ and low star-formation activity ($\rm \log(\text{SFR} / M_{\odot} \text{yr}^{-1}) \sim -0.5$ to $-0.9$). We find that our galaxies, and especially the satellites, are significantly offset from the local Fundamental Metallicity Relation (FMR), with deviations reaching $Δ\text{FMR} \approx -0.9$ dex. This indicates that these galaxies are likely experiencing strong accretion of pristine gas. Overall, we observe a large scatter in radial metallicity gradients, ranging from positive to negative with an average metallicity gradient of $\rm -0.02 \pm 0.04 \ dex \ kpc^{-1}$. Flat gradients are found in systems with confirmed satellites, suggesting that tidal interactions and mergers drive the radial mixing necessary to homogenise the interstellar medium. The (tentative) presence of an AGN in two of our sources suggests that strong feedback may also be responsible for the observed flat gradients. Conversely, the detection of a positive gradient in one source points toward a direct funnelling of metal-poor gas inflow into the central region of the galaxy. These results show that galaxies in the first billion years grow through diverse, episodic processes, suggesting that early evolution is characterised by structural variety rather than a single, predictable path.

We present two methods to estimate bottom topography in a shallow water flow using only surface deformation measurements. One is based on Physics-Informed Neural Networks (PINNs) and the other on the Adjoint State Method. We test both methods using synthetic data in 1D and 2D cases. Both are able to successfully reconstruct not only the bottom topography but also the surface velocity. Both also show robustness against noise and data sparsity up to reasonable levels.