Papers

Growing vulnerability disclosure and widespread software reuse increase security teams' need for reproducible evidence to diagnose vulnerabilities, validate patches, and build regression tests. Producing such evidence at scale requires automated end-to-end CVE reproduction. Existing methods typically process different CVEs through a uniform pipeline, but differences in runtime form, trigger interfaces, and prerequisite state impose different execution requirements on individual stages, making fixed workflows difficult to adapt to diverse reproduction needs. To address this problem, we present TAILOR, a type- and state-aware multi-agent framework specialized for complex vulnerability reproduction. TAILOR converts static vulnerability information into auditable reproduction evidence and packages reconstructed environments and trigger evidence into reproduction artifacts. Its first-level type-aware mechanism adaptively matches each vulnerability to an execution path. Within the Web path, its second-level state-aware mechanism constructs the required prerequisite state before exploitation, decouples prerequisite-state construction from core vulnerability triggering, and shares execution constraints across exploitation and verification. We construct a dataset of 200 CVEs with an emphasis on cases with complex execution requirements. TAILOR successfully reproduces 59.24% of Web vulnerabilities and 44.19% of traditional vulnerabilities. Further ablation experiments show that the two control levels respectively mitigate execution-path mismatch and missing Web prerequisite state. Overall, TAILOR broadens the coverage of automated CVE reproduction and provides auditable evidence for vulnerability diagnosis and defense.

Invariant manifold theory for partial differential equations (PDEs) is technically challenging as we often lack spectral gaps. Motivated by recent progress using renormalization in various areas of mathematics, we introduce a spectral renormalization method for semilinear parabolic equations that creates an artificial spectral gap at high frequencies while leaving the low modes unchanged. For the resulting family of renormalized equations, classical inertial-manifold theory yields finite-dimensional invariant manifolds whose dimension diverges as the renormalization parameter tends to zero. We show that the reduced dynamics on these manifolds approximate the original semiflow through an approximate semi-conjugacy, with an $O(\epsilon^2)$ error on bounded finite-time intervals. Under global incremental dissipativity, the semi-conjugacy is uniform in time and yields an $O(\epsilon^2)$ Hausdorff estimate for the global attractors, if they exist. We also establish further dynamical properties of the spectral renormalization such as persistence of hyperbolic equilibria and their Morse indices. We illustrate the construction in settings with and without a natural spectral gap and embed it into a two-parameter fast-slow PDE framework, where the artificial gap enables us to construct slow manifolds for systems on bounded domains with arbitrary dimension. Thus our new approach provides a finite-dimensional dynamical approximation even in regimes where an exact inertial manifold for the original equation is not available. In particular, this provides a practical balance between fully invariant manifolds and direct estimates.

We present World2Motion, a framework that generates scene-aware 3D human motion and corresponding video from a single image and a text prompt. While existing 3D motion generators learn from motion datasets, their generalization is constrained by limited coverage of environments. In contrast, video world models such as Cosmos 3 offer broader environmental priors but are not designed for full-body motion generation; recovering motion from their generated videos requires costly two-stage inference. To address these, we turn Cosmos 3 into a single-stage 3D motion generator. This adaptation has two challenges: the scarcity of paired video–motion data and temporal instability in the generated motion. First, we construct a training dataset combining synthetic video–motion pairs with real videos paired with estimated 3D motion. Second, we propose a shift-decoupled noise schedule that assigns different noise levels to video and motion through shared denoising progress. This design accommodates the different denoising requirements of the two modalities, reducing motion jitter. Experiments on a multi-source interaction benchmark show that World2Motion has better motion–text alignment and scene interaction compared with the evaluated 3D motion generators. It also matches the interaction success rate of the two-stage baseline while achieving approximately 3.3$\times$ faster inference. Our project page is available at https://fyantu.github.io/World2Motion/.

Vessel perception from space is crucial for a wide range of maritime applications, from traffic monitoring to environmental protection. However, most existing datasets predominantly focus on general object detection tasks in optical remote sensing (RS) images. Relying solely on single-modality optical RS images proves inadequate for effectively perceiving vessel objects in complex maritime scenarios, where ever-changing weather conditions (e.g., clouds and rain), the need for day-and-night coverage, and the inherent limitations of a single imaging modality pose significant challenges. To fill this gap, we introduce VesselBench-800K, the largest-to-date benchmark dataset on a global scale for vessel perception in multimodal RS images. As its name suggests, VesselBench-800K comprises 800,000 images, each at a resolution of 512x512 pixels, specifically curated for vessel perception tasks such as detection, counting, and density estimation. These multimodal image pairs (i.e., optical, SAR) are collected from diverse platforms, sensors, scenes, shooting heights, and synthetic sources, spanning spatial resolutions from 4.5m to 0.1m. Furthermore, we evaluate numerous state-of-the-art detection, counting, and density estimation models on VesselBench-800K through both qualitative and quantitative comparisons. By revealing previously unrecognized cues, this dataset holds immense potential to significantly advance our understanding of marine traffic. Our VesselBench dataset will be publicly available at https://github.com/danfenghong/IEEE_TGRS_VesselBench to support and contribute to community development.

High-resolution visual question answering often fails because a multimodal model does not acquire the small, spatially localized evidence needed to answer a question. Sequential zooming can recover detail, but it asks the main model to choose a region before obtaining a reliable overview. We introduce VPS, a visual parallel-search framework in which a main agent first invokes grid_search to inspect image tiles in parallel with question-conditioned sub-agents, and then adaptively invokes zoom_in on a precise or merged region. The same interface supports both training-free inference and post-training of the main and sub-agents. Across five benchmark splits and three model sizes, VPS improves mean accuracy over dedicated zoom-only search in 14 of 15 same-model comparisons, with gains up to 8.0 points and especially strong improvements for smaller main models. ZoomBench retains an approximately 3.2-point gain at every tested size. We further develop a supervision pipeline with hint-free verification and a paired role-specific GRPO surrogate for learning the controller and tile-reader roles. SFT improves observed accuracy on all five benchmark splits, including a 4.17-point gain on HR-Bench 4K. Role-specific RL further reshapes search behavior: main-only RL reduces mean tool use from 2.65 to 2.11 with similar pass@1 in an internal four-response evaluation, while external accuracy changes are mixed. Joint training reveals an asymmetry between local evidence reading and global search control. Together, these results support VPS as an effective inference-time scaffold and a trainable decomposition for visual evidence acquisition.

Diffusion Transformers (DiTs) enable high-quality video generation but suffer from substantial inference latency, primarily attributable to the computationally expensive full spatio-temporal attention. While sparse attention methods offer potential solutions, existing approaches face an inherent flexibility–efficiency dilemma: predefined masks lack the flexibility to capture diverse attention patterns, while runtime-determined masks introduce overheads and sacrifice hardware efficiency. We identify the lack of a unified structural characterization of DiT attention as a key limitation of existing methods, and establish that video DiT attention exhibits periodic diagonal stripe structures along both temporal and spatial dimensions. To formally encode these structured patterns within a single efficient kernel, we present PSA, a parameterized stripe attention that formalizes the observed stripe regularity, unifying diverse attention patterns for efficient mask generation. This unified representation enables a single hardware-efficient CUDA kernel to process all sparse patterns, achieving FlashAttention-3-level Model FLOPs Utilization. To determine optimal sparsity configurations, we propose a training-free offline search algorithm that automatically maximizes sparsity under a specified error tolerance for each attention head. Experiments on HunyuanVideo and Wan~2.1 demonstrate that PSA achieves 1.57$\times$ and 1.37$\times$ end-to-end speedups over FlashAttention-3 baselines, with acceptable visual quality degradation.

Cross-organizational collaboration is widely regarded as a key promise of SysML-based Model-Based Systems Engineering (MBSE), yet practitioners still face persistent challenges when exchanging and integrating system models. In parallel, Large Language Models (LLMs) raise expectations for AI-assisted model understanding and integration, while reliability and required human oversight continue to pose challenges. This paper reports the results of an online questionnaire survey with 29 MBSE stakeholders involved in cross-organizational collaboration. Respondents rated eight predefined integration challenge categories and six AI-supported task types on five-point Likert scales. The results indicate that stakeholders perceive model integration as a multi-dimensional alignment problem across semantics, behavior, traceability, and exchange interoperability. These perceptions vary by organizational role and frequency of integration involvement. AI is rated highly useful for analysis tasks such as semantic structure analysis and inconsistency detection, and respondents predominantly prefer human-in-the-loop use with mandatory verification. These findings motivate AI support that enhances, rather than replaces, engineering responsibility in SysML-based integration.

This paper develops a Quantum optimisation Unit Commitment (UC) framework that benefits from Pauli Correlation Encoding (PCE) as a qubit reduction preprocessing technique and Sample-based Quantum Diagonalisation (SQD) as a hybrid post-processing refinement method. The UC problem is formulated in binary form, then transformed into a QUBO problem and mapped onto an Ising representation suitable for adiabatic quantum evolution within the Qibo framework. This study analyses how PCE and SQD techniques affect qubit scalability, computational effort, feasibility, and probability of recovering the optimal solution as the number of generators increases. The results show that PCE reduces the effective size of the encoded problem, SQD improves the concentration of probability on feasible and optimal solutions, and their combined application provides the best overall behaviour in several scalable case studies, achieving more reliable solutions with lower effective computational cost than the unreduced formulation. The repository is available open source in Github (link is not yet available until work is published. However, if requested by the reviewers, we can make it open for a scific time window, if requested).

Modularity landscapes are often flat near the top: many distinct partitions achieve indistinguishable scores, and repeated runs of heuristic algorithms can still miss accessible optima. We introduce a two-phase workflow that (i) samples partitions until novelty saturates, then (ii) localises instability to a small subset of nodes and enumerates only that residual ambiguity under a locked stable core. Across a simple 50-nodes benchmark network, and a real weighted collaboration network, constrained enumeration systematically expands plateau coverage and can improve modularity beyond extensive Louvain restarts. Finally, we show that edge-weight rounding qualitatively reshapes plateau structure, making precision sensitivity an essential part of solution-space reporting.

Saddle-point problems play an important role in modern machine learning, including robust optimization and algorithmic fairness. We investigate structured convex–concave minimax problems where the primal variable lies in a high-dimensional, unconstrained space $\mathbb{R}^d$, while the dual variable is confined to a constrained, low-dimensional space $\mathbb{R}^J$, with $J\ll d$. Although existing theoretical frameworks establish that an accelerated rate of $\mathcal{O}(L/T^2)$ is possible, solving the associated subproblems efficiently can become a computational bottleneck in high dimensions. To address this, we introduce a decoupling technique. By restricting second-order updates to the low-dimensional dual space and employing first-order methods in the high-dimensional primal space, our framework combines Newton methods with Nesterov acceleration. For simplex-constrained dual variables and regularizers compatible with a self-concordant barrier, the resulting algorithm achieves a convergence rate of $\mathcal{O}(L/T^2)$ in primal objective suboptimality after $T$ outer iterations, without strong dual concavity. The additional linear-algebra cost per outer iteration is $\widetilde{\mathcal{O}}(dJ^2+J^{3.5})$, alongside one evaluation of the component losses and their Jacobian.

Knowledge Graph Embedding models have been extensively used to learn representations of entities and relations in Knowledge Graphs for predicting missing links. However, the quality of the learned representations varies a lot across different areas of the graph. If previous research has loosely linked the problem to relation types or degree bias, we show that it is more widespread and it correlates with the degree imbalance of the entities in test triples. In particular, the prediction of a target entity that has a degree much smaller than the degree of the anchor entity is extremely problematic. This is critical in recommender systems and other use cases, where these triples represent important corner cases. To address this issue, we propose an inference-time latent search optimization method capable of significantly improving model predictions on the most imbalanced triples. Built on top of a pre-trained model, it explores the embedding space at evaluation time, blending known and out-of-band information to mitigate the degree imbalance bias. We show the value of our approach on imbalanced triples from common benchmark datasets, where we outperform conventional methods, opening the door to the successful adoption of Knowledge Graph Embedding models on these critical corner cases.

Few-step streaming audio–video generation requires both causal modeling and step distillation, yet standard training recipes face two context-related challenges. Teacher forcing pairs clean history with a noisy target, but supervises predictive contextual representations only indirectly through velocity prediction. Meanwhile, directly reusing bidirectional score models in causal Distribution Matching Distillation (DMD) creates a mismatch between generation and scoring contexts. We address these challenges with Salt++, a two-stage post-training framework comprising Causal Self-Flow (CSF) and context-aligned autoregressive DMD. CSF exploits contextual information asymmetry by varying the history while keeping the noisy target fixed: a noise-mixed-history student aligns its intermediate representations with those of a clean-history exponential-moving-average teacher. This self-supervised signal encourages the student to extract semantic information and improves cross-modal alignment. Context-aligned AR DMD shares the causal mask and prefix across generator sampling, fake-score training, and real-score evaluation to match generated and reference distributions under a block-conditional KL objective. With calibrated teacher guidance, it performs clean-prefix few-step distillation and then adapts to generated histories without switching objectives or requiring separate consistency distillation. At 480p, Salt++ improves visual and motion quality by 57% and 45% over OmniForcing on JavisBench under the same 4-step causal setting. A separate scale-wise post-training stage extends Salt++ to 4-step $1664\times960$ generation, outperforming bidirectional LTX-2 on six of seven reported metrics. Project page: https://xingtongge.github.io/Saltpp

Weakly breaking a continuous symmetry in an ordered phase can convert its Goldstone modes into topological collective bands. We establish this mechanism in a neutral dipole condensate whose conserving parent supports two independent Goldstone phase modes. Weak real flavor mixing pins the phases and selects a chiral condensate. The resulting gyroscopic coupling acts on their winding polarization texture, producing a pseudo-Goldstone sector containing a topological lowest band. Restoring dipole conservation forces the two modes to meet at zero frequency, where their separate band Chern numbers cease to be defined. The same parent dipole charges define a tensor current whose circular response permits model-assisted reconstruction of low-energy embedded local Berry curvature in the benchmark regime. Analytical stability bounds and a large-occupation limit provide a controlled realization of this mechanism.

Complex tasks for underwater robots remain limited by the capabilities of their controllers. Learning a better one for a soft, underactuated robotic fish trades simulator cost against fidelity. We show that an intentionally low-fidelity simulator is enough: a stateless, quasi-steady fluid model with no wake and no added-mass history suffices to learn a general, closed-loop controller that transfers to hardware without tuning. Our platform is a soft, single-motor, tendon-driven fish whose policy observes only what the hardware can measure. A staged pipeline grounds the simulator in two independent identifications, fixing the tail dynamics and a stateless fluid model; the policy then acts through a band-limited rhythmic trajectory generator rather than commanding the tail directly. Deployed unchanged in an outdoor pool, a single policy performs closed-loop target reaching, disturbance rejection, and out-of-distribution target acquisition and tracking. The transfer rests on the constraint rather than the fidelity: the generator cannot leave the band over which the fluid was identified. This raises the question of how much of the physics can reside in the controller rather than in the simulator.

We investigate the motion of a spherical microswimmer in a planar Poiseuille flow using analytical calculations and numerical simulations. Our results show that the microswimmer's orientation undergoes periodic oscillations governed by pendulum-like dynamics. The oscillation period is proportional to the channel width and to the complete elliptic integral of the first kind evaluated at the maximum orientation angle, and inversely proportional to the square root of the product of the maximum flow speed and the self-propulsion speed. Based on the net displacement along the flow direction within one oscillation period and the signs of the maximum and minimum velocities, we identify five motion states: upstream, oscillatory upstream, zero-drift oscillatory, oscillatory downstream, and downstream motion. The net displacement is jointly determined by the ratio of the self-propulsion speed to the maximum flow speed and the complete elliptic integrals of the first and second kinds. The signs of the maximum and minimum velocities are determined by the speed ratio and the cosine of the orientation angle. Flow nonuniformity promotes upstream migration, whereas inertial lift produces negligible radial displacement during one oscillation period. These findings identify the key parameters governing the motion of spherical microswimmers and provide significant implications for understanding microorganism motility and designing microrobots with prescribed motion states.

Accurate identification of yarns in X-ray computed tomography volumes remains a critical and complex step in generating high-fidelity numerical models of woven composites. This work introduces a tracking framework that leverages the intrinsic periodicity of woven architectures to transform a complex, large-scale segmentation problem into the annotation of a single representative unit cell. The approach first exploits the periodic nature of the weave to extract a representative unit cell from the volumetric data and segment it to provide a reference description of the yarn geometry. Digital Volume Correlation (DVC) is then performed between an idealised periodic volume obtained by replication of the unit cell and the real composite volume, yielding a three-dimensional displacement field that captures geometric deviations from ideal periodicity. The unit-cell yarn segmentation is propagated to the full volume by exploiting the periodicity of the reference volume, and the annotations are transported to the actual volume using the DVC-derived displacement field. Results obtained on real composite datasets demonstrate the ability of the method to accurately recover warp and weft yarn architectures with minimal input, opening new perspectives for efficient and scalable textile composite characterisation.

We develop a two-stage multiple-scale method for hierarchical triple systems, using osculating orbital elements and time as the independent variable. We examine the disputed cross terms at orders $\delta\varepsilon^{3/2}$ and $\delta\varepsilon^{7/2}$, for which the multiple-scale results of Lim and Rodriguez (LR) disagree with the effective field theory results of Kuntz, Serra, and Trincherini (KST). Here, $\delta$ is the first post-Newtonian (1PN) parameter, and $\varepsilon=a_*/a_2$ ($a_*\simeq a_1$) measures the orbital hierarchy. These terms arise from couplings between relativistic interactions and the tidal field of the third body. Our results agree with KST: there is no secular term at order $\delta\varepsilon^{3/2}$, while the $\delta\varepsilon^{7/2}$ contribution consists of three parts: a direct term, a first-stage coupling between the inner binary's center-of-mass motion and the tidal field, and a second-stage coupling between the 1PN dipole and the quadrupole. Their sum reproduces the inner-orbit secular dynamics generated by KST's magnetic 1PN–quadrupole Hamiltonian, including its exact mass dependence, for nonsingular eccentricities and inclinations in the nonresonant, hierarchical weak-field regime. The analysis includes mean-anomaly perturbations and resubstitution of short-period corrections in both directions. We clarify how the choice of averaging variable affects the normalization of the mean elements and establish agreement with canonical perturbation theory. In particular, osculating and canonical mean elements give the same inner-orbit secular rates at this order.

In this paper, we study the hydrodynamics stability of the 3-D plane Couette flow $v_s=(y,0,0)$ with constant temperature state $\rho_s$ for the Boussinesq system in a finite channel $\mathbb{T}\times [-1,1]\times \mathbb{T}$ with non-slip boundary condition in $y$ direction. We prove that if the initial velocity field $v_{in}$ and initial temperature $\rho_{in}$ satisfy \begin{align*} \|v_{in}-(y,0,0)\|_{H^2} \leq c_0 \nu, \ \ \|\rho_{in}-\rho_s\|_{H^2} \leq c_0 \nu^2, \end{align*} for some $c_0>0$ independent of the viscosity and thermal diffusion $\nu$, then the solutions to 3D Boussinesq system do not transit from the Couette flow and constant temperature state. To our best knowledge, this is the first result about the nonlinear stability of plane Couette flow for 3-D Boussinesq system in the domain with non-slip boundary condition. This result, together with [18], implies that the strong boundary layer effect does not lead to strong instability.

Fast Radio Bursts (FRBs) are spatially and temporally compact sources at cosmological distances. As background sources, they will be uniquely sensitive to the gravitational lensing of exoplanets outside of our own galaxy. We present a method to detect extragalactic exoplanets using planetary echo nanolensing-periodic, nanosecond-scale variations in the relative time of arrival (ToA) between an initial FRB and its microlensed echo as the planet orbits its host star. For a simple binary lens model, we demonstrate that measuring the amplitude and period of these ToA fluctuations enables a determination of the exoplanet mass and orbital parameters. FRBs will be most sensitive to echo nanolensing by lenses within ten megaparsecs of the Milky Way as well as within ten megaparsecs of the FRB host, opening the door to detections of planets at a range of redshifts. We estimate that at least one in a million FRBs will exhibit this phenomenon (i.e. an optical depth of $\tau_l = 10^{-6}$). While current burst rates and low repeating-FRB fractions make a detection challenging in the immediate term, upcoming high-volume surveys and next-generation arrays (such as the SKA Phase 2) will bring the unambiguous detection of extragalactic exoplanets across cosmological redshifts into possibility.

Graph databases are increasingly queried through natural language, yet every existing benchmark evaluates isolated single-turn queries rather than the multi-turn sessions through which analysts actually work. We introduce CypherTurn, the first benchmark for conversational Text-to-Cypher evaluation, comprising 721 sessions and 5,927 turns across 7 knowledge graphs and 13 conversational phenomena. We evaluate 15 models under a guided oracle protocol and a fully autonomous agentic protocol, yielding four findings. First, the best model reaches only 64.7% execution accuracy, and session-level correctness remains below 5%. Second, despite strong overall rank correlation, frontier models exhibit a consequential reordering of the top of the leaderboard under autonomous operation, a phenomenon we term the Autonomy Divergence, which reveals error-management as a partially independent capability from raw generation skill. Third, scaling action budgets from x3 to x10 fails to close the autonomy gap, as the strongest frontier models self-limit to approximately two actions per turn regardless of available budget. Fourth, single-turn Cypher fine-tuning degrades multi-turn instruction following, while architecture-appropriate specialization outperforms several frontier models. These results establish CypherTurn as an open challenge for conversational graph database reasoning. Code and data are available at https://github.com/BarryQ/CypherTurn.

Federated LoRA fine-tuning enables parameter-efficient adaptation of pre-trained models without sharing private data, but suffers from two fundamental mismatches under heterogeneous client data: a structural aggregation mismatch caused by independently averaging LoRA factors, and a statistical collaboration mismatch caused by enforcing a single global adapter across divergent clients. To address these issues, we propose CF-LoRA, a clustered federated LoRA fine-tuning framework that combines decoupled factor aggregation with adaptation-aware client clustering. CF-LoRA first learns a globally shared $A$ factor while retaining personalized $B_i$ factors, then identifies clients with similar adaptation patterns based on the cosine similarity of their learned $B_i$ factors, and finally performs intra-cluster $B$-factor aggregation with a frozen $A$ factor. By decoupling LoRA factor aggregation, CF-LoRA preserves the low-rank structure and mitigates the structural aggregation mismatch, while adaptation-aware clustering promotes collaboration among clients with similar adaptation patterns and reduces negative transfer caused by statistical heterogeneity. Experiments on four language tasks and four vision datasets with RoBERTa and ViT show that CF-LoRA achieves the highest average accuracy in both modalities while communicating only one LoRA factor per optimization round.

The central challenge of world modeling is to learn representations that capture how the world evolves. However, existing world models predominantly represent future states without explicitly capturing the latent causes underlying their evolution, limiting their ability to reason about why and how the world changes. To address this limitation, we propose Abductive World Modeling (AWM), a framework that learns structured causal representations by abductively inferring latent causes from predicted futures. Specifically, we realize AWM through the Hierarchical Abductive State Pyramid (HASP), which organizes the inferred world state into three complementary components - Entity, Dynamic, and Relation - capturing what exists, how it changes, and how entities interact, respectively. By jointly reasoning over the current observation and its predicted future, HASP abductively infers these latent factors and integrates them into a structured state representation for downstream reasoning. To the best of our knowledge, AWM is the first framework to introduce abductive state inference into latent-space world modeling for learning structured representations of world dynamics. Experiments across physical prediction, causal reasoning, and action understanding demonstrate the effectiveness of our approach. Compared with V-JEPA, a state-of-the-art latent-space world model, AWM improves physical prediction AUROC by 10.7%, causal reasoning accuracy by 16.8%, and action Top-1 accuracy by 68.0%.

Complex questions often require multi-hop reasoning that connects facts distributed across sources or distant regions of a long context through intermediate steps. Benchmarks commonly evaluate this ability with questions built around predefined reasoning chains, treating a correct answer as evidence that the intended composition was used. Yet answer correctness alone leaves open whether success depends on the evidence associated with each intended step: models may instead rely on memorized associations, shorter paths, or partial evidence. We examine this dependence using the Behavioral Necessity Rate (BNR), which measures how often targeted evidence removal prevents answer recovery on initially correct instances. Across five existing benchmarks, panel-mean BNR ranges from 16.6% to 48.9%, exposing a substantial gap between annotated structure and observed dependence. Guided by this diagnosis, we introduce REALHOP, a diagnose-construct-verify framework that rebinds entities, factorizes selected relations, adds complete competing paths, and places evidence at traceable locations. Structural and semantic checks precede freezing; behavioral interventions follow. On 790 paired MuSiQue questions, REALHOP raises panel-mean BNR from 27.4% to 94.4% while retaining high Full accuracy. It also yields high BNR on REALHOP-FRAMES and REALHOP-LONGBENCH. On 216 long-context questions, the matched multiple-choice spread across 16 models grows from 13.9 to 59.2 points and persists under repeated open-ended evaluation. Together, these results show that a conceptually coherent chain and a correct final answer do not by themselves establish multi-hop reasoning. Verifying that success depends on every intended hop is therefore as fundamental to multi-hop evaluation as measuring answer accuracy itself.

Day-night solid-state power generation is experimentally demonstrated with thermoelectric generators integrated with a tunable radiative coating, which switches between a selective solar absorber during daytime and a radiative cooler during nighttime by temperature without any external control. The self-adaptive radiative coating is carefully designed with thermochromic VO2 thin film on heavily doped silicon substrate along with a phase-shifting silicon spacer in between, whose physical mechanism is elucidated. The fabricated coating exhibits a high solar absorptivity of 0.90 and a low infrared emissivity of 0.23 during daytime, and a high emissivity of 0.85 at night, resulting in large emissivity change of 0.62 within the atmospheric transparency window upon metal-to-insulator phase transition. From outdoor tests in a high vacuum apparatus, the tunable radiative coating in 35-mm squared size achieves 1.64 W/m2 power generation around noon time from a stack of commercial thermoelectric modules of 15 mm squared, which is 32% more from that with a black absorber. At night, it produces 26 mW/m2 power, outperforming the black emitter by 63%. Even exposed the ambient with convective loss, the 2-inch-round tunable coating generates power of 0.48 W/m2 at noon time, 14% more than the black absorber of the same size. The tunable radiative coating covered thermoelectric modules produces open circuit voltages from 433 mV to -62 mV in vacuum and from 300 mV to -51 mV in ambient through 24-hr continuous outdoor tests.

Knowledge concept tagging aims to assign specific concept or topic labels to educational content, which is essential for both educators and learners in traditional and online teaching practices. Recent work has explored large language models (LLMs) for this task, achieving promising performance. However, LLMs still struggle to select the correct concept from a large-scale candidate set due to the high dimensionality of the decision space. In this paper, we propose a novel three-stage Select-Reason-Judge (SRJudge) framework, which empowers LLMs with selective reasoning capability for fine-grained knowledge concept tagging. Specifically, the Selector in Stage 1 first narrows the candidate concepts to a top-K shortlist by fine-tuning a small language model (SLM), e.g., BERT, since the top-$K$ predictions hit the correct concept in most cases, thereby reducing the decision space of correct candidates. Next, the Stage 2 Reasoner employs a lightweight LLM for refined reasoning over the shortlisted candidates. It further integrates an improved reinforcement learning strategy with a dynamic task-specific reward function and a pruning mechanism to better align with human reasoning preferences. Finally, a larger LLM acts as a judger that evaluates the overall rationality of the reasoning process and its explanations to determine the final output. In addition, we construct two high-quality datasets for further validation, i.e., the biology dataset S_Bio and the physics dataset S_Phy. Experimental results demonstrate that our method consistently outperforms state-of-the-art baselines across benchmark datasets, verifying its effectiveness and superiority. Resources are available at: https://github.com/Nicozwy/SRJudge.

Bellissard's framework uses covariant observables to describe aperiodic solids. We examine how butterfly-like spectra in twisted bilayers can be related to bulk gap labels and topological response. A bulk gap permits a spectral projector and an integrated density of states. A topological response requires an additional pairing of that projector. In an illustrative finite honeycomb bilayer, an exact second-moment identity isolates the interlayer contribution to spectral variance. Angle-resolved spectra, finite state counts, an uncoupled reference, and size and broadening checks show how coupling redistributes spectral weight without assigning a bulk gap label. We then examine the requirements for atomistic electronic-structure calculations, including nonorthogonal orbitals, projected densities of states, and basis functions that move during relative sliding. The resulting framework specifies the steps needed to connect atomistic spectra with bulk state counting and topological response.

Despite recent advances in accuracy and efficiency, coarse matching remains an indispensable yet costly stage in existing semi-dense matchers due to dense token-level matching. We present UltraMatch, an ultra-efficient and scalable semi-dense matching framework that bypasses the quadratic computation and memory cost of dense token-level matching by routing only a small fraction of candidate matching paths. At its core, a lightweight Transport Path Router operates on coarse block representations to rank candidate target blocks for each source block and retain only a small set, restricting subsequent token-level matching to the selected paths and avoiding the construction of the full token-to-token matching matrix. We further design a sparse global Dual-Softmax that performs matching only over the routed block candidates while retaining global competition across the sparse matching space. Beyond matching acceleration, UltraMatch employs deployment-oriented structural reparameterization for feature extraction and a tiny fine matching head with shared parameters, further reducing inference cost and memory consumption. UltraMatch achieves competitive accuracy among semi-dense matchers, while running 1.67$\times$ faster than SuperPoint+LightGlue with only 0.44 GiB peak inference memory. Its scalability enables inference at up to 6K resolution on a single RTX 3090, whereas existing semi-dense matchers run out of memory before reaching 2K. Our routing strategy is also transferable, delivering about 2$\times$ end-to-end speedup in EDM and ELoFTR without accuracy loss. The project repository is available at https://github.com/JiajunLe/UltraMatch.

LLM-as-a-judge is widely used for evaluating text, but extending this paradigm to speech requires models to interpret acoustic as well as linguistic evidence. This introduces a modality-specific risk: a speech judge may treat a perceptually salient cue as evidence of quality even when that cue is irrelevant to the target criterion or receives more weight than human listeners give it. We call this behavior an acoustic shortcut. To study it, we audit six speech LLM judges using controlled manipulations of intensity, content richness, and emotional delivery. We evaluate both pointwise scoring and pairwise comparison, using human preference calibration to interpret the results. The judges consistently reward louder audio, prefer content-rich speech more strongly than human listeners do, and map emotional delivery into quality preferences. These effects are most visible in pairwise comparison, while pointwise scores often obscure them. More concerningly, the accompanying rationales rarely identify the acoustic cue that changes a judgment and instead repeatedly rely on a limited vocabulary, leaving them acoustically underspecified. Together, these findings show that reliable speech judges must both resist acoustic shortcuts and ground their rationales in the acoustic evidence behind their decisions. To support reproducibility and future audits, we also release SpeechJudgeAudit, the controlled stimuli and evaluation tools used in this study.

Existing analyses for the transient stability of grid-forming (GFM) converters under current limiting mainly address single-converter systems, while network-level transient stability analysis remains challenging. This paper develops a network energy construction that incorporates current limiting while retaining a provable dissipation property. This structure enables Lyapunov-based analysis consistent with classical direct method for synchronous-machine based systems, without switching between mode-dependent energy functions. A two-GFM example using a reactive circular current limiter and impedance insertion illustrates the proposed framework. Electromagnetic transient simulations support the theoretical results.

Community-detection methods search over possible partitions of a network, but the structure of this partition space is rarely examined directly. This paper introduces two complementary tools that make~$\mathcal{P}$ analytically and visually accessible. First, a canonical labelling scheme based on the Restricted Growth Sequence~(RGS) is adopted, assigning each partition a unique, permutation-invariant identifier and eliminating the need for pairwise similarity measures such as the Normalized Mutual Information. Second, granularity~$\Gamma$ is defined and combined with modularity~$Q$ to produce a two-dimensional projection: the Partition Space Map~(PSM). Within the $(\Gamma,Q)$ plane, only a discrete set of integer values of~$\Gamma$ is attainable, and upper and lower bounds on~$Q$ are derived directly from the deviation matrix. These tools are demonstrated through complete enumeration of~$\mathcal{P}$ for small benchmark graphs, revealing the fine structure of modularity degeneracy and the combinatorial constraints that shape the feasible region. The framework provides an exact, assumption-free reference for evaluating heuristic community detection algorithms and lays the groundwork for geometry-aware exploration of partition space in larger networks.