Papers

We construct a commutative ring whose big finitistic projective dimension is $1$ and whose big finitistic injective dimension is $2$. The example contradicts the comparison suggested by Bass for commutative rings.

We prove that a complete surface of finite topological type satisfying Condition A is homeomorphic to the two-torus. More generally, we show that an isolated planar end of a complete open surface satisfying Condition A cannot be periodic under the induced action on the space of ends. Condition A includes dense periodic points, a uniformly hyperbolic splitting into one-dimensional subbundles for a complete metric, and invariant line fields that uniquely integrate to transverse stable and unstable foliations. These results give a partial answer to the question of whether a complete surface satisfying Condition A must be compact.

Let $X$ and $Y$ be real Banach spaces, and let $\cC(X)$ and $\cC(Y)$ denote the families of all nonempty bounded closed convex subsets of $X$ and $Y$, respectively, equipped with the Hausdorff metric. We prove that every surjective isometry $F:\cC(X)\to\cC(Y)$ is induced by a surjective affine isometry of the underlying spaces. More precisely, $F$ maps singleton sets onto singleton sets, and the map $T:X\to Y$ defined by $F(\{x\})=\{T(x)\}$ is a surjective affine isometry satisfying \[ F(A)=T[A]:=\{T(x):\;x\in A\},\qquad(A\in\cC(X)). \] In particular, our theorem extends the result of Gruber and Lettl for finite-dimensional Euclidean spaces to arbitrary real Banach spaces, without imposing any additional assumptions on the underlying spaces.

In order to ascertain the precise atomic characteristics, it is important to incorporate a wide range of electron correlations in the computational analysis. However, it is important to note that undertaking such studies will result in substantial increases of the CSF expansions. The present publication, as well as the series of articles that have been published by G. Gaigalas, P. Rynkun, and L. Kitovienė, are dedicated to the analysis and resolution of the problem in question. In this series, a method was developed based on second-order perturbation theory to identify the most important core-valence, core, core-core, and valence-valence correlations. The method under discussion is based on a combination of the relativistic configuration interaction method and the stationary second-order Rayleigh-Schrödinger many-body perturbation theory in an irreducible tensorial form. In this study, the method is further expanded to encompass additional core-valence electron correlations of the third and fourth types. Conversely, the correlations that are not encompassed by perturbation theory are addressed in a conventional manner. This method can be used to calculate the properties of an atom or ion with any number of valence and core electrons. As an example of its application, the atomic calculations of the energy structure and lifetime for Ne II are presented.

Time-dependent coherent control can enhance excitation transport in open quantum networks, but independently controlling every inter-site coupling creates a control space of high dimension and leads to difficult optimization problems. We introduce an edge-ranking strategy based on the control-induced change in the gradient component of the time-integrated quantum probability current, which is obtained via a graph Hodge decomposition. When our strategy is applied to the seven-site Fenna-Matthews-Olson (FMO) model, the six-edge set retains $99.83\%$ of the enhancement achieved by full control, and the four-edge set retains $97.60\%$ while reducing the pulse fluence—used here as a proxy for control effort—by $41.55\%$ relative to full control. Dephasing scans and comparisons with random edge sets and random networks provide numerical support for the relevance and potential broader utility of the ranking. These results show that edge selection guided by the quantum probability current can substantially reduce the control space while preserving high transport performance with lower control effort.

Real-world recommendation scenarios are commonly grounded in shared physical environments during user-recommender interactions. This motivates situated conversational recommendation (SCR), a complex task requiring recommender assistants to jointly reason over dialogue history, co-observed scenes, and in-scene item attributes. However, current approaches struggle with this setting due to two intertwined challenges: accurately understanding situated user preferences throughout the conversation and generating responses that simultaneously satisfy user needs and grounded situations. To this end, we propose Re2A, a framework that formulates SCR as a structured reason-then-align process. We introduce rubric-based preference reasoning, which uses automated rubrics to guide the model toward producing explicit preference states. Based on these states, we propose a preference-conditioned optimization to align response generation with dual objectives: user preference satisfaction and situation consistency. Extensive experiments on two SCR datasets demonstrate that Re2A consistently outperforms state-of-the-art methods, delivering more precise, context-aware conversational recommendations. Our code is available at https://github.com/DongdingLin/Re2A.

An advanced retrieval-augmented generation pipeline is typically assembled from three or four independently operated systems: an approximate nearest-neighbour index, a full-text search engine, a graph database, and a relational document store. Each contributes its own deployment surface, configuration model, and failure modes, and the integration logic that binds them is written anew in every project. In this work, we present Quanta, an open-source Python library, which unifies dense vector search over 4-bit quantised embeddings, BM25 full-text retrieval, and knowledge-graph traversal behind a single retrieval API. Quanta makes two design commitments, which distinguish it from existing hybrid retrieval stacks. First, signals are combined by weighted reciprocal rank fusion rather than by normalising heterogeneous scores onto a shared range, which we argue is ill-posed because such normalisations are query-dependent. Second, the graph is a candidate expander and not a relevance scorer: traversal widens the candidate pool, and the newly admitted documents are re-scored by the dense indexes under an identifier allowlist, so structural adjacency determines what is considered while content evidence determines how it ranks.

Core-collapse supernovae (CCSNe) are intense sources of feebly interacting particles (FIPs), whose visible decays beyond the stellar envelope can generate distinct electromagnetic signals. We propose the circumstellar medium (CSM) surrounding the CCSNe progenitor as a new target for 511 keV line searches. FIPs escaping the stellar interior may decay into electron-positron pairs and inject positrons into the CSM, where they slow down and subsequently annihilate, giving rise to a 511 keV line as an electromagnetic precursor to the supernova shock breakout. Using dark photons as a benchmark, we calculate the time-dependent signal for CSM profiles motivated by SN 2023ixf and SN 2024ggi. For a Galactic supernova at 10 kpc, the projected reach of COSI and AMEGO extends beyond existing supernova limits over part of the MeV-scale dark-photon parameter space. The same FIP decays can simultaneously heat the CSM and sublimate circumstellar dust, producing a rapid spectral transition from infrared excess to optical or ultraviolet emission [Y. Cheng et al., Phys. Rev. Lett. 137, 111005 (2026)]. The coincident appearance of this transition and the 511 keV line would provide a distinctive correlated multiwavelength signature of FIP decays in the dense presupernova environment.

Computing dynamical expectation values typically relies on approximations optimized for the state or observable separately, without accounting for how their errors combine in the final expression. This work explores an alternative approach in which each approximation is guided by both the state and the observable, accounting for how they jointly determine the target expectation value. This idea is implemented through a sweep algorithm with coupled loss functions for forward state and backward observable updates. Exact error relations provide an analytical rationale for how the proposed losses can improve the accuracy. Numerical tests on 30-qubit random circuits show errors two to three orders of magnitude smaller than those of variational state compression at equal bond dimensions. These results motivate further exploration of joint state and observable approximation for dynamical expectation value estimation.

A3P5 NEMESIS is a four-wheel rover intended to combine remote inspection, environmental observation and lightweight manipulation within one serviceable platform. This study develops a photo-constrained geometric reconstruction, a subsystem architecture and a reproducible analytical assessment while distinguishing physical prototype evidence from proposed functions. An exploratory search retrieved 5,000 bibliographic records across ten queries, yielding 4,897 distinct DOI records and 1,212 metadata candidates; selected primary studies and technical documents informed the design. The reconstructed configuration retains the carbon-pattern enclosure, independently steered wheel assemblies, folded manipulator, inclined camera mast and side sampling equipment. A declared 24 kg scenario predicts 3.28 newton-metres of gearbox-output torque per wheel on a 20-degree grade under equal load sharing; a separate static model shows how a 2 kg forward payload reduces the geometric front-tipping bound from 38.1 degrees to 32.7 degrees. These are design screens, not measured operating limits. A public-data calibration benchmark uses 7,344 eligible hourly observations, eight sensor/environmental predictors and chronological training, validation and test partitions. Validation-selected ridge regression achieves a held-out CO root-mean-square error of 0.502 milligrams per cubic metre, with a 95% daily-block bootstrap interval of 0.435-0.569 milligrams per cubic metre. This result concerns an external sensor array and cannot establish NEMESIS accuracy. The combined analysis identifies priority measurements, proposed control interfaces and mission-specific validation requirements. The contribution is a traceable engineering design study and evaluation framework for a prototype whose integrated field performance remains to be established.

We investigate the purity dynamics of a quantum system coupled to a continuous-variable pointer through a von Neumann-type interaction Hamiltonian of the form $\hat H_{\mathrm{int}}=g\,\hat H\otimes\hat p$. For an initially Gaussian pointer state, the interaction generates energy-dependent conditional translations whose mutual overlaps are determined explicitly by the populated spectral separations of the system Hamiltonian. After tracing out the pointer degrees of freedom, we obtain the reduced density operator and derive an exact analytical expression for the time-dependent purity. Using this expression, we establish two-sided purity bounds governed by the minimum and maximum nonzero energy gaps on the populated spectral support. These bounds provide a state-dependent spectral characterization of the loss of purity and become exact for two-level systems. We further show that the short-time decrease of purity is controlled by the Hamiltonian variance of the initial state, with $\mathcal P''(0)=-(g^2/\sigma^2) \operatorname{Var}_{\psi_S}(\hat H)$. In addition, an explicit sufficient timescale is derived for the purity to approach its asymptotic value within a prescribed tolerance, revealing the scaling $t_{\varepsilon}\propto\sigma/(|g|\Delta_{\min})$. Finally, the general results are illustrated for an equally weighted $N$-level system with an equally spaced spectrum, for which the asymptotic purity is $1/N$. The analysis clarifies the distinct roles of spectral separation, energy variance, coupling strength, and pointer width in Hamiltonian-conditioned purity loss.

Vision-Language-Action models and World-Action Models have advanced language-conditioned robotic manipulation, yet often leave metric relations among actions, objects, and scene geometry implicit. Human manipulation combines semantic understanding of task-relevant objects with spatial feedback that guides hand motion relative to objects and their surroundings. Inspired by this, we introduce a metric interaction framework that models object-level and scene-level interactions in physical Cartesian space at a shared metric scale. At the object level, Interaction-Centric Tokens (ICTs) explicitly represent end-effector pose trajectories relative to manipulated objects and are jointly denoised with actions, providing physically grounded interaction supervision. At the scene level, the Metric Action Interaction Field (MAIF) uses action and ICT queries to attend to metric scene point-cloud features and learns geometry-conditioned action corrections. Through two-stage adaptation, our framework improves diverse VLA and WAM baselines with a small number of additional parameters and training steps. Experiments demonstrate average success-rate gains of 0.80 and 3.59 percentage points on LIBERO and RoboTwin 2.0, respectively, alongside gains of 6.80 percentage points on real-world tasks and 7.45 percentage points on their out-of-distribution variants.

Force-aware Vision-Language-Action (VLA) policies improve contact-rich manipulation, but typically combine task-level motion and contact-dependent adjustment in a single action prediction. Demonstrations provide no explicit labels for decomposing that prediction into a reusable reference action and a correction. We present ForceDelta-VLA, a correction-distillation framework that constructs an explicit force-correction target using paired predictions from a frozen teacher's force-conditioned and learned force-agnostic modes. A separate delay-correction target accounts for reference-action mismatch and the change in reference state. Training uses asynchronous schedule replay with the cached task context available during execution. The resulting lightweight policy adjusts the reference actions using recent force history and robot state, responding to contact changes between reference-action updates without regenerating complete action chunks. Across nine single-arm and bimanual contact-rich tasks, ForceDelta-VLA achieves an 82.2% mean success rate, compared with 54.4% for the original ForceVLA baseline. Direct execution of our Stage-1 Temporal Teacher achieves 70.6%. Relative to ForceVLA, the complete system reduces mean peak contact force over successful trials by approximately 26% on both platforms.

We prove a form of the fundamental theorem of vector calculus for Hölder continuous vector fields with no assumption on the derivatives that exploits Schauder spaces with negative exponents. Then by developing an idea of von Wahl in the context of Lebesgue spaces, we prove a form of the Helmholtz Weyl decomposition Theorem for spaces of Hölder continuous functions, with explicit formulas for the projection operators.

The method based on the second-order perturbation theory to identify the most important configuration state functions of the various correlations is extended to include valence-valence correlations, which are described by the three-particle Feynman diagram. The extension presented in this work complements the core-valence, core, core-core, and valence-valence correlations which were developed in a series of previous papers by G. Gaigalas, P. Rynkun and L. Kitovienė. Whereas these valence-valence correlations are described by the three-particle Feynman diagram, additional developments to calculate the spin-angular parts of this diagram have been made to the program library librang of the Grasp. As an example of the application of the developed method, the atomic calculations of the energy structure for the Se III are presented. In the present work, this method was also used to select the most significant configuration state functions and to use this basis to solve the self-consistent field equations.

Structured pruning is commonly formulated as ranking individual channels, although channel responses can be complementary or cancel through downstream mixing. Motivated by these response interactions, we formulate pruning as the selection of a subset with large joint response capacity, followed by a separate functional realization step. Our unified response geometry maps each candidate set to \(M(D,R)=D^{1/2}RD^{1/2}\) and uses its determinant together with Schur-greedy residuals to select non-redundant coordinates. The same construction yields two information-conditioned instances: an unlabeled instance based on activation covariance, and a task-conditioned instance that combines activation and gradient variance for response scale with gradient correlation for complementarity. To convert the selected subset into an executable network, we fold predictable removed responses into successor weights through ridge compensation and recalibrate batch-normalization statistics, without fine-tuning the network. On ImageNet ResNet-50, the unlabeled instance reaches \(65.4\%\) and \(53.9\%\) Top-1 accuracy at 30% and 40% deletion, versus \(59.8\%\) and \(43.1\%\) for strength-only selection; the task-conditioned instance reaches \(67.7\%\) and \(56.3\%\) under the same protocol. A six-family screen shows architecture-dependent behavior, with positive relative contrasts in several convolutional and expansion-layer settings and clear boundary cases in windowed attention. These results support response geometry as a conditional principle for structured pruning, with its benefit determined jointly by the observed response and the architecture in which that response is realized.

Observational decision-support systems often expose one causal estimate as a recommendation even when plausible estimators disagree. The inherent engine of the proposed system is causal machine learning: a conditional-average-treatment-effect estimand identified by backdoor adjustment, estimated by an EconML DML causal forest and DoWhy linear regression, checked by two-way fixed effects, and converted into candidate levers by constrained optimisation. F-DACE is the decision layer on that engine. It represents precision, propensity overlap, placebo-refutation stability, interval overlap, and directional agreement as fuzzy memberships. Hard vetoes force abstention after estimand mismatch, failed diagnostics, informative sign conflict, or weak evidence. In 180 panel simulations spanning six identification conditions, F-DACE made a decision in 67.2% of runs and limited false recommendations to 17.2%; the corresponding rates were 33.3% for the causal forest and 35.6% for backdoor regression, matching deterministic unanimity rather than dominating it. Nearly all (30 of 31) false recommendations occurred under shared unmeasured confounding, which no fusion rule can diagnose when every component shares the omitted variable. The retail application aggregates a public Walmart panel to 6,435 store-weeks across 45 stores. F-DACE abstains for all five markdown indicators: some estimates are imprecise, one refutation fails, and MarkDown5 has a direct sign conflict. A LangGraph conversational agent exposes impact, what-if, and lever-optimization tools while a deterministic verifier preserves causal-layer status. On 24 live questions it achieved 100.0% tool-routing accuracy, 100.0% status fidelity, and 0.983 mean groundedness. On ten adversarial questions it resisted all injected instructions.

We prove the \(3/4\) conjecture for \(q\)-ary fix-free codes with at most three distinct codeword lengths, for every integer \(q\geq2\). Every prescribed length distribution with Kraft sum at most \(3/4\) is realized by a deterministic construction. We introduce a matrix approach based on an exact identity for the overlap between forbidden prefix extensions and suffix residuals. Natural numerical order fixes the shortest layer, and the remaining selection problem is expressed through row and column counts. A fixed-cardinality interpolation theorem supplies feasible sets of every intermediate cardinality between nested endpoints, provided their differences satisfy one-sided uniqueness and either acyclicity or integral slack. Reverse-order selection handles the uniform cases directly; in the remaining cases, layer completion and aligned groups provide endpoints for interpolation. Together, these methods extend the binary three-length result to arbitrary finite alphabets and give a deterministic procedure for constructing the code.

We prove the existence of $C^{r_{*},\alpha}$-regular sub-stable invariant manifolds and conjugacies for analytic maps near a fixed point in the presence of resonances, assuming an invertible semisimple linearization. The regularity order $r$ and the admissible Hölder exponents $\alpha$ depend on the resonant indices and the relative rates of contraction. Our approach uses functional expansions based on logarithmic polynomials, which allow the invariance equation to be solved explicitly and algorithmically to arbitrary order. The existence results are obtained via a fixed-point argument in a suitable Banach space. We further present several analytic examples that illustrate the theorem and demonstrate the loss of regularity caused by resonances.

We develop a factorization framework for magnetic induction tomography in the eddy-current regime. The induced current density is represented by a divergence-free current vector potential, and the external excitation by magnetic dipole densities on an observation sphere. This choice reveals a natural physical factorization of the near-field operator into a data operator, an interior boundary response operator, and an adjoint counterpart. By introducing the Riesz map on the flux space, we recast this factorization in a Hilbert-space setting, which yields a positive-semidefinite operator with a coercive middle factor. Douglas' range theorem then identifies the range of the square root of this dissipative operator with the range of the data operator. For inclusions with a regular boundary, this range identity yields a uniqueness theorem for the support of the conductivity. This characterization leads to a non-iterative spectral indicator based on a regularized Picard quotient computed from the Hermitian imaginary part of the near-field data matrix. The reconstruction stage requires no forward solves. Numerical experiments confirm that the indicator localizes and separates the inclusions accurately, degrades gracefully under increasing noise, and captures non-convex support geometry without any convexity prior.

Low back pain (LBP) is a leading cause of disability worldwide and affects populations ranging from working adults to students with prolonged sitting habits. Conventional lumbar support belts are generally static and non-adaptive, which limits their ability to accommodate dynamic postural changes and individualized comfort requirements. This paper presents SmartFlex, an intelligent wearable lumbar support system that integrates real-time posture recognition with an adaptive air bag array. The system uses a JY901S gyroscope sensor to detect user posture and a lightweight TinyML neural network deployed on an Arduino R4 UNO to process posture data at the edge. Based on the recognized posture state, a closed-loop pneumatic control system dynamically inflates or deflates 14 distributed air bags through four independent micro air pumps to provide targeted biomechanical support. Evaluation results show that SmartFlex achieves over 94% posture recognition accuracy and generates corresponding pressure-control commands with a sensing-to-command delay of less than 120 ms. The pneumatic system operates within a calibrated pressure range of 15-85 kPa. A user study with 20 participants produced a 4.5/5 rating for support effectiveness, suggesting that adaptive wearable support may improve daily sitting comfort and reduce lumbar fatigue.

Multi-messenger astronomy (MMA), the joint study of cosmic sources through gravitational waves (GWs), electromagnetic (EM) radiation, neutrinos, and cosmic rays, is rapidly reshaping time-domain astrophysics. Realizing the promise of MMA will require coordinating heterogeneous observing resources and automating the chain from alert to analysis to follow-up. RADAR (Radio Afterglow Detection and AI-driven Response) is a federated, privacy-enhancing framework for the radio follow-up of GW events, previously validated on GW170817. Here, we extend it along three axes. First, we benchmark three large language models (LLMs; GPT-5.5, Claude-Opus-4.7, and Gemini-3.5-Flash) against the GW170817 radio light curve dataset. GPT-5.5 attains the highest event-level $F_1$ score, the harmonic mean of precision and recall, at $0.893 \pm 0.010$, and the highest GCN-level recall, $0.794 \pm 0.013$, improving on previous GPT-4.1 results by 16% and 10%, respectively, while Claude-Opus-4.7 achieves the highest precision, $0.978 \pm 0.014$. Second, we introduce concurrent likelihood evaluation, which speeds up the MCMC computation by a factor of $40\times$ over our previous results. Third, we present an LLM-driven scheme that converts natural-language observing requests into submission-ready scheduling blocks for the Karl G. Jansky Very Large Array. Together, these developments advance RADAR toward a scalable, largely autonomous system for GW radio follow up.

Real-robot demonstrations are limited, motivating the use of human manipulation data collected without robots, including egocentric videos and handheld Universal Manipulation Interface (UMI) demonstrations. However, differences in viewpoint, embodiment, and available action supervision make it difficult to align representations across these sources according to manipulation motion rather than visual appearance. We introduce UMI-Bridge, which uses UMI as an intermediate domain to align representations according to action equivalence rather than pixel similarity. UMI action supervision anchors the latent representation to end-effector motion and gripper behavior, while synchronized head-wrist observations and paired ego-UMI clips support alignment across views and domains. We train a dual-view latent action model (LAM) on human manipulation data without robot demonstrations, then freeze its wrist teacher and dynamics model to regularize vision-language-action (VLA) post-training on UMI and robot data. The shared wrist interface enables this training-time supervision across both domains while preserving the policy's standard inference architecture. Across three real-robot tasks, UMI-Bridge achieves 91.7% mean success versus 73.3% for Naive Co-training with matched UMI and robot data. On two data-efficiency tasks, it surpasses a full-data Robot-only baseline using 25% of the robot demonstrations together with UMI data. It also achieves 85% and 90% success on two additional tasks learned from UMI demonstrations without task-specific robot demonstrations. These results support action-anchored latent alignment for data-efficient robot learning and UMI-to-robot task transfer.

We employ uncertainty quantification methods to investigate how parameter identifiability changes in the vicinity of a bifurcation point. We perform numerical experiments on the non-spatial Klausmeier vegetation model with random coefficients, which describes biomass-water interactions and exhibits a fold bifurcation. We partition the domain around the bifurcation into regions with distinct convergence behaviors. In each region, we follow a UQ workflow that includes sensitivity analysis, Bayesian inference, and Fisher information evaluation to assess parameter identifiability. The results show that proximity to the bifurcation point is decisive for parameter estimation. Reliable joint inference of both model parameters is not possible when the system exhibits bistability. Furthermore, we can reconstruct the model's bifurcation pattern by Fisher information heatmaps, presenting a practical tool for bifurcation detection. Lastly, we highlight the importance of transient data for successful parameter estimation.

We study the deterministic oracle complexity of smooth convex optimization when the algorithm receives only exact function values. The objective is a globally $\beta$-smooth convex function, all queries and the final output are restricted to the Euclidean ball of radius $R$, and the unique minimizer lies in the ball of radius $R/2$. We establish an upper bound of $O(d\sqrt{\beta R^2/\epsilon})$ using coordinate finite differences together with an error-robust accelerated projected method. Our main contribution is a matching lower bound, up to the high-accuracy saturation of the construction: any deterministic adaptive value-oracle algorithm requires $\Omega\!\left(d\min\{\sqrt{\beta R^2/\epsilon},(d/\log(ed))^{1/3}\}\right)$ queries. Consequently, the minimax oracle complexity is $\Theta(d\sqrt{\beta R^2/\epsilon})$ throughout the moderate-accuracy regime $\beta R^2(\log(ed)/d)^{2/3}\leq\epsilon\leq c\beta R^2$ for a universal constant $c>0$. The lower bound must account for the fact that a single exact real value can encode arbitrarily much information. To overcome this difficulty, we construct a single fixed smooth convex hard instance using a Moreau-smoothed biased max chain, an exact prefix-shielding mechanism, and batched delayed rotations. These techniques preserve consistency with the full adaptive transcript and establish the optimality of the square-root complexity branch for deterministic bounded-query algorithms.

This paper provides a recursive formula for computing the coefficients in a small-system-noise asymptotic expansion for nonlinear filtering. The expansion, obtained from the Kallianpur–Striebel formula, was justified in the author's previous work. Our main contribution is to reduce the coefficient calculation to a finite-dimensional system extending the Kalman–Bucy filter by applying Fubini's theorem and Wick's formula and differentiating the resulting terms. For each fixed expansion order, the number of variables grows at most polynomially, rather than exponentially, with the system dimension. To justify the construction, we define the required non-adapted integrals as limits of discrete sums and establish a generalized Ito formula. We also extend the expansion from conditional expectations to conditional characteristic functions and provide a numerical illustration of the method.

Journal Entry Tests (JETs) are a mandatory part of annual audits to evaluate and assess both highrisk audit areas and potential material misstatements. However, as JETs are designed to detect known patterns based on domain knowledge, the resulting lists are often very large and require substantial additional effort from the auditor. To ensure the economic efficiency of the audit, the number of false positives in JET result lists must be reduced. Especially machine learning (ML) methods represent a promising approach to improve anomaly detection in this field. In this research in progress paper, we investigate different approaches on how to combine JETs with ML-methods in a hybrid manner. We present specialized models to increase the detection performance and validity of anomaly detection results to improve audit efficiency. The experiments are based on synthetic data consisting of different normal and anomalous journal entries.

Wildfire smoke detection from satellite imagery is critical for early warning and rapid response. For onboard satellite deployment, detection systems must operate under strict memory and latency constraints while providing spatially informative outputs for downstream decision-making. Existing tile-level classification methods are computationally efficient but lack spatial localization, whereas pixel-level segmentation approaches provide detailed masks yet are typically too computationally demanding for real-time onboard execution. To address this gap, we propose WISE (Weakly-supervised Inference-efficient Smoke Extraction), a deployment-oriented framework for onboard fire smoke detection and localization. WISE leverages only tile-level annotations through a teacher-student distillation strategy, where an offline teacher provides soft spatial supervision to a lightweight WISE-Student optimized for efficient onboard inference. The student jointly predicts tile-level smoke presence and smoke probability maps within a single forward pass, enabling spatially informative detection under strict computational constraints. WISE was evaluated through in-orbit execution aboard the ISS-mounted IMAGIN-e payload. Three model variants achieve average inference times of 0.10 s, 0.14 s, and 0.26 s per tile, indicating near-real-time per-tile inference within onboard resource limits. Ground-based experiments on Landsat 5 and Landsat 8 imagery further indicate effective detection and spatially informative localization. The best-performing variant achieves a mean tile-level F1 score of 0.964 and a mean pixel-level F1 score of 0.750 across 10 runs, while containing only 0.12M parameters and requiring approximately 3 GFLOPs. Together, these results indicate that WISE is a practical candidate for low-latency wildfire smoke monitoring from space under onboard resource constraints.

Galactic cosmic rays (CRs) are thought to be accelerated by diffusive shock acceleration in supernova remnants (SNRs), but observations of SNRs aged ~100-1000 yr suggest maximum CR energies below the PeV level. Recently, Inoue et al. (2021) showed with kinetic-MHD simulations that CRs can be accelerated to ~3 PeV when a blast wave shock propagates through dense circumstellar material (CSM) within tens of days after explosion. Such CSM can be produced by a red-supergiant (RSG) wind with an observationally motivated mass-loss rate of ~10^-3 Ms/yr. 100 TeV gamma-rays from neutral pion decay can test PeV CR acceleration. However, hadronic gamma-rays from very young SNRs can be significantly attenuated by supernova photospheric photons and the cosmic background radiation. Using the CR distribution functions and CSM density profiles from Inoue et al. (2021), we calculate time-dependent gamma-ray fluxes from Type II-P SNe, including both attenuation processes. We find that photospheric photons attenuate the intrinsic 1-100 TeV gamma-ray flux by about one order of magnitude. For the dense CSM case with dot{M}_RSG ~ 10^-3 Ms/yr and an outer CSM radius of ~ 3x10^15 cm, CTA can detect 100 TeV gamma-rays with a 50 h observation out to ~ 4 Mpc (this is reduced to 3 Mpc if the spatial extent of the CSM is shrunk to 2x10^15 cm). These results suggest that early 100 TeV observations of nearby Type II-P SNe can probe PeV CR acceleration in sufficiently dense and extended CSM.

Can quantum entanglement continue to grow even after the entanglement entropies of subsystems have saturated? We address this question using genuine multi-entropy, which probes multi-partite entanglement structure beyond bipartite entanglement. In a two-sided BTZ black hole, this quantity grows linearly with time after every subsystem entropy of a fixed tripartition has saturated, extending the interval in which entanglement detects Einstein-Rosen bridge growth. The result refines the ER=EPR relation by identifying multi-partite structure invisible to ordinary entanglement entropies. Some spin-chain models and the spatially partitioned Sachdev-Ye-Kitaev (SYK) model exhibit distinct transients, while the late-time plateau values in holography, the SYK model, and Haar-random states share a universal behaviour.