Science & Mathematics

Blurred vehicle contours and low contrast limit detection performance in infrared remote sensing imagery, while existing fusion methods may introduce polarization noise and fail to integrate complementary information effectively. To address these limitations, we propose a Mamba-based infrared polarimetric remote sensing image fusion model. Infrared polarization information reflects surface material properties and geometry, providing complementary information to infrared intensity. The proposed method first combines long-wave infrared total intensity S0 and the degree of linear polarization (DoLP), using a physics-inspired polarization-response proxy and spatial structural continuity to extract guidance features. It then employs bidirectional state space scanning along horizontal, vertical, and both diagonal orientations to enable cross-modal feature fusion across the entire image. Prior-guided multi-window structural similarity, reconstruction constraints, and a relative lower-bound loss for polarization saliency are jointly introduced to suppress noise while preserving global background structures and target polarization features. In addition, we construct AIR-LWIR-Pol, a dataset of 1,901 accurately registered image pairs acquired using an airborne division-of-aperture long-wave infrared polarimetric imaging system, and evaluate downstream vehicle detection using YOLO11. The experimental results show that the proposed method outperforms the comparison methods on four fusion metrics, including entropy and structural similarity. Vehicle detection mAP50 and mAP50-95 reach 55.0% and 29.4%. These results demonstrate the effectiveness of combining physical priors with state space modeling for long-wave infrared polarization image fusion and support its use for target detection in complex remote sensing scenes.

In agent workflows, a brief human instruction can alter subsequent search, reasoning, and tool execution, yet final task outcomes rarely reveal the contribution of an individual instruction to human-agent joint production. We propose a Human Turn-level framework for measuring human contributions, using task quality and token usage to characterize the collaborative production outcomes associated with each human instruction. We compare continuations from adjacent checkpoints under matched models, budgets, and termination conditions, and assess stability through repeated measurements. Across four archived and simulated interaction settings, we conduct 29,124 comparisons between predicted and observed directions of quality and computational change at both Human Turn and session levels. On HumanBench, we obtain 506 valid comparisons across 85 real-user scenarios. Results show that quality and computational changes associated with human guidance depend on task conditions. Expert-style and ordinary-style guidance differ in information provision, correction, and repetition. Even when task quality after continuation is identical, remaining computational demands can differ substantially across checkpoints. Our framework advances the measurement of human labor contributions from tasks and sessions to individual turns, providing a repeatable, fine-grained foundation for studying human contributions and their allocation in human-agent joint production.

Achieving a detailed picture of the nature and dynamics of phase boundaries remains a major challenge, particularly solid-solid transitions central to phase-change materials. Here, we use stroboscopic optical scattering microscopy and finite-element simulations to study the nonlinear thermally driven motion of the mesoscale insulator-metal phase boundary in VO2 thin films. We visualize the lateral spatiotemporal evolution of a diffraction-limited photoinduced metallic phase domain, resolving its outward propagation, contraction, and ultimate disappearance. The phase boundary dynamics are governed by heat flow, which can be controlled by excitation fluence, film thickness, and substrate thermal conductivity. Parameter-free agreement between the data and nonlinear thermal transport simulations identify the phase boundary transport as locally diffusive under this regime. We quantify how the latent heat acts as a thermal inertia, suppressing transport and prolonging the phase coexistence. Paradoxically, despite being driven by the underlying thermal gradient locally described by Fourier's law, the mesoscale phase front propagates faster than expected from thermal diffusion. This phenomenon arises from the distinction between temperature-field evolution and phase-boundary isotherm motion. These results establish a link between heat dissipation and mesoscale phase-front motion, providing a framework for controlling phase-transition kinetics relevant to photonics applications of VO2. We anticipate that this approach generalizes to a broad class of first- and second-order transitions.

The arithmetic regularity lemma of Green is a fundamental tool in additive combinatorics, providing a decomposition of arbitrary subsets of finite Abelian groups into pseudorandom pieces with respect to Fourier analysis. Over the years, it has found numerous applications in combinatorics, property testing, and theoretical computer science. However, Green's proof is existential, and many algorithmic applications require an efficient procedure for constructing the underlying regular decomposition. De, Nadimpalli, and Servedio (SODA 2022) developed a query-efficient constructive regularity framework over $\mathbb{F}_2^n$ and explicitly raised the problem of extending their dimension-independent sumset-estimation algorithm to other finite Abelian groups. In this work, we resolve this question for finite Abelian groups of constant torsion. We develop the first constructive version of Green's {\em Arithmetic Regularity Lemma} in this general setting. Given oracle access to a subset of the group, our algorithm computes an implicit description of a regular decomposition together with an oracle representation of a structured subset satisfying the guarantees of Green's lemma. The algorithm runs in time and query complexity polynomial in the regularity parameters and the tower bound, while remaining independent of the size of the ambient group. As applications, we obtain efficient algorithms for removing linear configurations and approximately counting linear images, including weighted sumsets and suitable matrix systems of Cauchy--Schwarz complexity one, over finite Abelian groups of constant torsion. Our results provide a unified algorithmic framework extending dimension-independent constructive arithmetic regularity and its applications beyond $\mathbb{F}_2^n$.

Integrated Remote Operation Centres (IROCs) are emerging as the architectural frontier of Industry 4.0 for capital-intensive sectors such as mining, oil and gas, and energy. Yet two intertwined gaps constrain their evolution towards full autonomy. First, no integrated agentic architecture has been proposed for mining IROCs that orchestrates Supervisory Control and Data Acquisition systems, digital twins, predictive analytics and human-machine interfaces under safety-critical constraints. Second, no evaluation framework jointly addresses the real-time, Health-Safety-Environment, regulatory-auditability and cost-per-decision requirements that govern industrial agentic deployments. This paper makes four contributions. It synthesises industry needs and quantified gains documented in analogous industries that motivate an agentic approach to mining IROCs. It introduces EI2ROC, a three-tier multi-agent architecture combining a Large Language Model orchestrator with specialised monitoring and production-capacityprediction agents communicating through a Model Context Protocol-style tool interface. It provides a critical review of agentic evaluation along six orthogonal dimensions and identifies the industrial blind spot of current benchmarks. Finally, it proposes a multi-dimensional benchmarking framework for industrial agentic AI comprising scenario library, harness, metrics and humancalibration protocol. The paper states explicitly which components are implemented in the current prototype, which belong to the proposed reference design and which remain future work, and reports preliminary cost and latency observations from an operating site. Together, EI2ROC and the benchmarking framework constitute a methodological blueprint transferable across CAPEX intensive industries.

Parameter-efficient fine-tuning (PEFT) methods are critical for adapting large models under resource-constrained settings. However, as mixture-of-experts (MoE) architectures, which offer efficient inference at massive parameter scales, gain increasing prevalence in the community, a critical gap arising from the lack of systematic recipes for applying PEFT to MoE models. To narrow this gap, we conduct a large-scale systematic study that investigates a set of essential design recipes, including which modules should be adapted, which granularity should be used to adapt MoE modules, which PEFT method to use, and to what extent should the load-balancing objective be used, across three MoE backbones on downstream tasks and general knowledge tasks. Our experiments yield several key insights. First, adapting MoE experts alone achieves the highest downstream performance with competitive general capability retention under a comparable parameter budget. Second, native expert-weight updates offer a better downstream--retention trade-off than adapters that bypass experts or entire MoE modules. Third, LoRA and DoRA together define the strongest operating region for high downstream accuracy among the evaluated PEFT methods. Fourth, no nonzero load-balancing coefficient Pareto-dominates the zero baseline. We finally evaluate a weight-level redesign of MoE$^2$-LoRA on Qwen3. Moving its module-level bypass pool into native expert weights improves all four upstream endpoints by 2.4 to 17.9 points with 17% fewer trainable parameters and a 3.2-point average downstream cost.

Cardinality estimation is essential for query optimization, but most existing learned estimators focus on a single predicate type, such as string LIKE clauses or numerical range filters. In practice, analytical workloads often combine both types, while the accuracy of string estimators tends to decrease with longer patterns. We introduce MultiCard, a comprehensive learned cardinality estimator that unifies LIKE predicates and numerical filters within a single end-to-end architecture. By combining a Transformer for long-range sparse patterns with an LSTM for short, frequent patterns, MultiCard achieves robust pattern modeling. In parallel, a numerical encoder integrates catalog histograms and structural predicate features, enabling the model to dynamically assess histogram estimation accuracy across different operators and data skew conditions. A final Transformer layer captures interactions among heterogeneous predicates before a multilayer perceptron predicts query cardinalities. We also introduce UPCI, a workload generator that produces length-balanced LIKE patterns, multi-column numerical predicates, and exact cardinality labels for training and evaluation. Experiments on five real-world datasets show that, for string-only workloads, MultiCard achieves the best or near-best Q-error across all datasets compared with state-of-the-art methods. For mixed multi-attribute queries, MultiCard reduces PostgreSQL's mean Q-error by up to 16 times while maintaining an inference latency below 5 ms.

Irregular multivariate time series forecasting (IMTSF) requires capturing temporal dynamics and cross-variable dependencies from sparse, asynchronous observations. Existing methods often rely on learned temporal encoders, coupling historical representation construction with model optimisation and repeatedly processing the same observations. We propose CTLite, a lightweight forecasting framework that decouples historical summarisation from learned prediction. Specifically, a parameter-free continuous-time moment encoder summarises local levels and trends at multiple temporal scales while suppressing poorly supported slope estimates. A compact readout combines damped local extrapolation with a low-rank cross-variable correction that accounts for observation freshness, target availability, and forecast horizon. Since the historical summaries are independent of trainable parameters and forecast queries, they can be reused across optimisation steps and repeated predictions for an unchanged history without altering predictions or parameter gradients. Experiments on four real-world benchmarks show that CTLite achieves the best or second-best mean MSE and MAE among the evaluated methods with only 88-2,537 trainable parameters. Under matched GPU workloads at batch size 32, CTLite reduces inference latency by 18.6-87.3% compared with the latest models, including the cost of historical encoding. Reusing summaries further reduces total runtime by approximately 52% over 16 forecast calls on the same history compared with recomputing the encoder for every call, including the initial encoding cost. The code is at: https://github.com/SSMa913/NICLab-CTLite.

This paper considers a decision-maker tasked with identifying the $k$ best items from a large pool of candidates, such as job applicants, LLMs, award nominees, or new product ideas. The decision-maker learns about the items' quality by soliciting human evaluations. Each evaluation can be costly, so the decision-maker aims to use the limited evaluation budget efficiently by allocating evaluations adaptively. A key consideration is the form of evaluation to solicit, which should be something evaluators are comfortable providing. Two popular forms are ratings and pairwise comparisons, and we allow for both, as they may be especially useful in combination given the limited budget. Ratings can be used early on to get a coarse picture of which items are good and which are not. As that picture forms, comparisons can refine the ordering of the good items, narrowing in on the ultimate top-$k$ selection. To solve the allocation problem, we develop an adaptive algorithm built from two allocation policies. The first, our primary contribution, is based on a surrogate objective that is naturally forward-looking, valuing evaluations that help inform what to evaluate in future rounds, not just those that directly improve the selection decision. The decision rule is interpretable, and we prove that it is asymptotically consistent. Late in the evaluation process, when being forward-looking is less productive, we construct a myopic policy and a criterion for switching to it. In simulation, the method gives insight into how to combine these two types of evaluations, and beats the (limited) existing benchmarks.

Test-time scaling can improve LLM reasoning, but existing approaches typically adapt generation or apply answer selection after a predetermined generation schedule. We study within-instance allocation: under a hard per-instance prompt-plus-completion token cap, when should inference stop, generate another candidate, or spend residual tokens adjudicating candidates already observed? BOSS is a training-free policy for this three-way decision, with worst-case budget checks before every generation and judge call. Its key contribution is online generation-adjudication allocation under one shared token ledger. Across two Llama-3.1 models and four reasoning benchmarks, BOSS attains the highest accuracy in 22 of 24 model-task-budget cells under the common hard cap. At Llama-3.1-8B and B=4k, it reaches 60.5 macro-average accuracy versus 58.6 for Adaptive-Consistency, 59.1 for TALE-EP, and 58.9 for fixed-schedule League. On identical Stage-I candidate pools, randomized A/B judging with repeated voting improves plurality from 58.4 to 60.3, with the full BOSS selector reaching 60.5. Approximately matched-cost controls further characterize how the benefit of adjudication changes with available inference budget. Together, these results establish generation-adjudication allocation as a distinct design axis for reasoning under strict per-instance caps.

An inference budget is useful only when the system delivers an answer within it. We study a fixed-budget policy combining first-valid-box stopping with a charged finalization cue. On 100 fresh, independently audited GSM8K test questions per model, it improves delivered endpoint accuracy over first-box stopping by 8.0 percentage points for Qwen3-4B-Thinking-2507 and 7.0 for DeepSeek-R1-Distill-Llama-8B. A prelocked repeat on the same questions gives 7.0 and 10.0 points; neither seed shows a gross harm relative to first-box stopping. The point estimates of incremental physical-slot savings are below one percent. A finalization-only ablation separates cue recovery from stopping, and a fixed scalar reader tests extraction as an alternative explanation. The cue is prior art; the contribution is a component and readout analysis with charged injections and verified paths. Results cover two checkpoints, one task, one prompt family, and two fixed seed conditions.

In analog coded computing, encoding basis functions are typically chosen to guarantee exact reconstruction and improve numerical stability. However, the same basis functions also determine the statistical geometry of the observations available to colluding workers, and hence directly influence the privacy of the underlying data. This suggests that basis selection should not be viewed solely as a reconstruction or numerical conditioning problem, but rather as a joint privacy and accuracy co-design problem. Motivated by this observation, we revisit Analog Lagrange Coded Computing (ALCC) and show that the full orthogonality constraints imposed by the conventional Lagrange basis are stronger than what is required for exact recovery. We retain only the interpolation constraints necessary for the data bearing basis functions and require the noise bearing basis functions merely to vanish at the data recovery points. This yields a parameterized family of generalized polynomial encoding bases that preserves exact reconstruction while introducing additional design degrees of freedom. We then derive a basis-dependent characterization of differential privacy for colluding workers and formulate a privacy-calibrated numerical optimization problem that jointly selects the data and noise basis functions to reduce reconstruction error while satisfying a prescribed \((ε,δ)\)-DP requirement. The resulting framework therefore treats the encoding basis itself as a privacy-design variable, rather than selecting a fixed closed-form basis a priori. The same design philosophy is also compatible with alternative privacy metrics such as mutual-information leakage. Numerical results show that, at the same privacy level, optimized basis functions substantially reduce floating-point reconstruction error relative to conventional ALCC.

Every driving scene is logged once, but driving is one-to-many: the scene determines which behaviors are feasible, while an instruction selects among them. A log records only the action taken, leaving no direct supervision for alternative feasible actions or instructions the scene cannot support. Consequently, a policy that ignores language and one that follows it without regard to scene constraints can fit the same data. We present Cartender, which learns from the drives not taken. Counterfactual Branching constructs matched and mismatched instruction-action pairs from real recordings without a simulator: a vision-language model proposes alternative futures and their selecting instructions, while unsupported instructions provide negative labels. A learned scene-language compatibility score regulates language conditioning in trajectory generation, and a reward-trained language residual re-ranks generated candidates over a scene-based score. The branches are used for training; instruction following is evaluated by swapping instructions under identical recorded inputs and measuring behavior rather than agreement with generated targets. On WOD-E2E, the best of four proposals achieves an RFS of 8.642 across 3,603 feasible instruction conditions. Across 8,974 admitted proposals, mean RFS is 8.203 versus 7.906 for matched logged trajectories, and 67% score higher than their logged counterpart. Compared with matched-only imitation using the same backbone, Cartender improves same-scene paired instruction success from 23.67% to 57.00% and RFS from 6.32 to 7.10; its gate reduces infeasible-instruction execution from 53.36% to 14.18%.

Bringing legacy video back to contemporary screens does not necessarily make it look contemporary. Earlier imaging pipelines often leave color casts, muted chroma, weak con- trast, and tonal imbalance, while correcting these character- istics still relies heavily on expert colorists and scales poorly to long, multi-shot videos. We formulate long video color modernization, a task that updates the chromatic and tonal appearance of legacy color footage while preserving scene content and coherence across frames and shots. We introduce a scalable, training-free framework that jointly modernizes grids of temporally adjacent keyframes using a pretrained text-guided image editing model, producing coherent con- temporary references without paired supervision. Compact shot-specific color transforms then propagate these adjust- ments throughout each shot, preserving spatial structure and temporal stability. We conduct a multidimensional evaluation on diverse legacy videos spanning five decades. We combine a human evaluation of perceived modernness with quantita- tive measures of content fidelity, temporal stability, percep- tual quality, and efficiency, and introduce Source-Weighted Inter-shot Grade Consistency (SW-IGC) to assess grade co- herence across shots. On our evaluation set, our frame- work achieves the highest perceived modernness while re- maining competitive across complementary measures, open- ing a path toward the systematic study of long video color modernization. Code is available at https://github.com/jiwoomood/ Video_Color_Modernization.git

Architectural roles describe how responsibilities are distributed within software systems. Entity--Boundary--Control (EBC) separates domain information, external interaction, and workflow coordination, but its proportions across diverse software remain unclear. This study examines file-level classifier-derived EBC proportionality in 1,556,985 Java files from 732 project instances and 78 functional categories. A classifier trained on labels generated by a large language model assigned one representative role per file. The labels were aggregated at project and functional-category levels and analysed using descriptive statistics and density-based clustering. The results show that Entity was less common than Boundary and Control. The dominant cluster covered 48 categories, 472 project instances, and 47.75% of classified files; its mean composition was 13.25% Entity, 42.65% Boundary, and 44.10% Control. The remaining 30 categories form one alternative and 11 smaller specialised patterns. Some selected variations were consistent with category responsibilities, while others received mixed support. Agreement with independent expert labels was limited (31.37% accuracy; macro-F1 0.31; Entity recall 13%). All reported proportions and clusters are therefore patterns in classifier-assigned roles, not verified architectural ground truth. Within this measurement boundary, the study provides a large-scale empirical reference of EBC distributions at the file level, and it documents the limitations of the automated approach.