arXiv:2606.24842v1 Announce Type: new Abstract: In the big-world regime, agents cannot be universally capable and their ability is inevitably specialized across a world model in pieces. Consequently, standard uniform guarantees fail to distinguish between the understanding of critical bottlenecks and irrelevant failures. We first formalize this limitation by proving that general agents are not universal, rendering standard worst-case analysis uninformative. To overcome this, we introduce structural certification, a transition-local framework that maps bounded goal-conditioned performance to entry-wise guarantees on the agent's internal world model. Our main contribution is constructive. We provide algorithms that filter specific transitions using deep compositional goals and prove that a general agent on these goals has a structural world model with a $\mathcal{O}(1/n) + \mathcal{O}(\delta)$ error bound. Conversely, this bound is tight in the small-$\delta$ regime, whose existence is explicitly guaranteed by our certification. These results enable the certifiable deployment of general agents by localizing the specific transitions where long-horizon planning is reliable.
Science Journals
arXiv:2606.21349v2 Announce Type: replace Abstract: Most IoT malware operates as botnets dependent on Command and Control (C2) servers, but the short-lived nature of attack infrastructure often leaves samples dormant without C2 communication, hindering dynamic analysis. This paper proposes a system that combines Ghidra with a Large Language Model (LLM) to extract communication specifications from a malware binary and automatically generate a pseudo-C2 server. Experiments using Mirai demonstrate that the proposed system semantically interprets binary control structures and extracts all 20 core protocol elements in agreement with the ground truth (100\% specification extraction accuracy). The generated pseudo-C2 server fully reproduces seven of ten DDoS attack vectors with attack behavior consistent with the original C2. When applied to a customized variant created by modifying the publicly available Mirai source code, the method succeeds end-to-end -- from specification extraction through pseudo-C2 generation to attack reproduction -- demonstrating that the LLM infers specifications from binary structures without relying on pre-trained knowledge. This approach extends the applicability of LLMs from analysis assistance to the automated construction of dynamic analysis environments.
arXiv:2606.24598v1 Announce Type: new Abstract: While expert-validated "LLM + script" workflows deliver significant value, they remain static: they encode hard-won domain knowledge yet fail to adapt execution based on feedback. Existing agent research predominantly targets greenfield agents and synthetic benchmarks, leaving the migration of active legacy workflows unresolved. To bridge this gap, we present a reversible, Strangler-Fig migration path that refactors legacy workflows into composable, typed, and auditable stages. Central to this framework is a three-tier convertibility taxonomy (A/B/C), implemented as a routing stage within the system harness, which diagnoses a workflow's readiness and routes it accordingly.
arXiv:2606.24616v1 Announce Type: new Abstract: Tokens have become the practical accounting unit for modern foundation model services, linking information processing, computation, memory use, energy expenditure, pricing, and economic value. This paper develops a framework for AI tokenomics: the study of how tokens are generated, consumed, priced, allocated, and optimized across AI systems. We connect token-level technical costs to workflow-level production functions, enterprise resource allocation, measurement and instrumentation methods, and emerging market-design questions. The framework shows that token expenditure and economic value are distinct: value depends on marginal productivity, workflow position, hidden reasoning activity, risk, and downstream propagation effects. The paper concludes by identifying open research directions in hidden-token measurement, empirical calibration, token productivity, dynamic allocation, and token-based markets.
arXiv:2606.24784v1 Announce Type: new Abstract: High-resolution aerial imagery has recently emerged as a complementary modality for automated driving perception and has shown potential to improve birds-eye-view (BEV) scene understanding when fused with onboard sensors. Prior work demonstrated performance gains for online high-definition (HD) map construction through aerial-onboard fusion; however, conventional end-to-end fusion does not fully exploit the structural information contained in aerial representations. In this work, we introduce AerialFusionMapNet, a fusion-based mapping framework with a structured two-stage training strategy that explicitly enhances the contribution of aerial features within a unified pipeline. The proposed training scheme enables more effective integration of structural aerial priors. On the nuScenes geographic split, AerialFusionMapNet achieves up to 54.7 mAP, improving over prior aerial-onboard fusion baselines from 48.8 mAP by +5.9 absolute and +12.1% relative. The results suggest that structured training design, rather than increased architectural complexity, plays a more decisive role in unlocking the full potential of aerial imagery for online HD map construction. Code and trained models are available at https://github.com/DriverlessMobility/AerialFusionMapNet.
arXiv:2606.24498v1 Announce Type: new Abstract: Grounding deictic gestures in natural images is fundamental to AR and human-robot collaboration, providing a basis for seamless spatial interaction. While Transformer-based visual models have achieved significant progress in general object detection, their global attention mechanisms often neglect micro-geometric relationships, degrading orientation accuracy. In pointing tasks, this deficiency manifests as an inability to accurately capture the pointing ray implied by finger poses, which results in pointing drift and localization ambiguity when dealing with distant or densely packed objects. To address this, we propose VistaRef, a framework designed to explicitly enhance spatial orientation awareness. First, we develop the Local Hand Entity Modeling (LHEM) module, which incorporates hand-pose embeddings to strengthen the model's capability to capture subtle finger deviations. Second, drawing inspiration from multi-view geometry, we construct the Geometric Ray Modeling (GRM) module to transform implicit orientation information into explicit spatial geometric features, guiding feature aggregation and deep fusion via attention mechanisms. Furthermore, we introduce a novel Orientation-Consistent Alignment Loss (OCAL) to synergistically supervise hand presence and pointing consistency, ensuring that all architectural improvements collectively serve the core objective of spatial localization. Experimental results demonstrate that VistaRef significantly outperforms the baseline, achieving a 14-point absolute gain in grounding accuracy. Qualitative analysis further confirms that VistaRef effectively models the geometric correlation from hand to target, bridging the spatial perception gap inherent in traditional Transformers for complex scenarios. Code: https://github.com/lingli1724/VistaRef.
arXiv:2504.05336v4 Announce Type: replace-cross Abstract: A recurring weakness in quantum machine learning (QML) is that reported ``quantum advantages'' are seldom tested against a \emph{capacity-matched} classical control, leaving it unclear whether a gain comes from the quantum substrate or from the architectural change that accompanies it. Our primary contribution is methodological: a protocol for attributing such gains honestly -- a capacity-matched classical bottleneck of identical parameter budget, transparent reporting of where quantum does \emph{not} help, and validation on real quantum hardware -- which we develop and apply through a concrete case study. That case study is Quantum Adaptive Self-Attention (QASA), a hybrid Transformer that replaces the value projection of a \emph{single} encoder layer with a 36-parameter parameterized quantum circuit (PQC), keeping all other layers classical. Across nine synthetic benchmarks and the real-world ETTh1 dataset, QASA improves on a full-capacity classical Transformer for chaotic and trend-dominated signals. To ask whether this is a genuinely \emph{quantum} effect, we introduce a control rarely applied in quantum machine learning -- a capacity-matched classical bottleneck with the same parameter budget -- and find that it matches the PQC on the error metrics. The gain is therefore attributable to the low-rank value-projection \emph{bottleneck} (an \emph{architectural parsimony} principle), not to quantumness; adding further quantum layers only degrades performance and trainability. We accordingly position the quantum layer not as a source of accuracy advantage but as a \emph{competitive} instantiation of this principle: its low-rank compression onto the signal's intrinsic dimensionality is matched by a classical bottleneck, so the gain is architectural rather than quantum.
arXiv:2606.24284v1 Announce Type: new Abstract: This paper focuses on a control-based approach for enhancing the regenerative capabilities of hydraulic multi-actuator systems using individual metering valves. Thanks to this architecture, pressure and displacement of each actuator can be controlled nearly independently. By determining online, the right pressure to be driven, it enables the optimization of regenerative control strategies for resistive or driving forces. Globally, this control strategy behaves such as a load sensing approach but each metering valve is piloted in order to activate regenerative mode when it is allowable. The main contribution relies on optimizing the pressure to be controlled in each actuator and the main pump in order to maximize the regenerative capacity of a hydraulic machine while following a displacement. The effectiveness of the proposed approach is proved in simulation. Only a single pump line regeneration is explored here but extensions to multi-pump or direct regeneration are also possible.
arXiv:2606.24292v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have demonstrated impressive vision-language understanding, yet still struggle with fine-grained perception in high-resolution images. While existing training-free methods typically rely on attention-based localization or coarse-to-fine search, they are often misled by distractors and fail to locate multiple targets. Our investigation attributes these failures to Contextual Dominance, where salient distractors overwhelm target attention and cause inaccurate localization, and Semantic Bias, where global semantics cause the model to fixate on the most salient concept, resulting in incomplete localization in multi-object scenarios. Built on these insights, we propose ActiveScope, a training-free framework that enhances MLLMs by actively seeking and correcting perception. ActiveScope features two modules. The Semantic Anchor Localization (SAL) utilizes fine-grained semantic anchors to independently localize key targets, thereby mitigating semantic bias. The Interference-Suppressed Refinement (ISR) refines localization by suppressing attention on salient distractions to overcome contextual dominance. Extensive experiments on high-resolution image understanding benchmarks demonstrate that ActiveScope outperforms existing training-free methods (e.g., 96.34 percent accuracy on $V^{*}$ Bench), validating the superiority of the active search and self-correction paradigm. Our code is available at https://github.com/jasmine-ww/ActiveScope.
arXiv:2606.24866v1 Announce Type: new Abstract: Blackbody infrared radiative dissociation (BIRD) activates molecules through successive absorption of ambient thermal photons until the internal energy reaches a dissociation threshold. Because these radiative transition rates depend on the electromagnetic density of states (DOS), structured infrared environments provide a route to control thermal unimolecular dissociation. Here we develop a state-resolved master-equation framework for polyatomic BIRD in a planar Au/MgO multilayer cavity, where the reactive cluster $\mathrm{(H_2O)_2Cl^-}$ is studied. The cavity modifies the kinetics through the DOS sampled by anharmonic fundamental, overtone, and combination transitions. We show that MgO surface phonon polaritons produce strong near-field enhancements in the central vacuum reaction region of a microcavity. We find that short cavities with thick polar crystal layers yield the largest BIRD enhancements due to enhanced evanescent surface phonon polariton contributions. We further include collisions with a methane bath gas and show that cavity DOS engineering shifts the crossover between BIRD and collisional activation. These results establish Reststrahlen-band DOS engineering as a practical strategy for controlling polyatomic BIRD in infrared microcavities.
arXiv:2606.24547v1 Announce Type: new Abstract: Intelligent active particles are characterized by self-propulsion, directional sensing of their environment, information processing, decision making and goal-oriented self-steering. This implies, in particular, the prevalence of non-reciprocal interactions, and the importance of information propagation through agent groups. Examples include biological systems (cells, insects, birds, fish, pedestrians) as well as engineered systems (nano- and microbots). As many agents move in an aqueous medium, hydrodynamic interactions strongly affect the dynamics. The emergent dynamics includes the formation of swarms and flocks, predator-prey behavior, and the navigation in complex environments.
arXiv:2606.24301v1 Announce Type: new Abstract: Recovering realistic 3D vehicle models from autonomous driving scenes is crucial for synthesizing training data and building simulation environment. However, most existing vehicle generation methods fail to fully exploit multimodal sensors i.e. multi-view images and LiDAR point clouds) and rely on neural rendering based reconstruction, leading to low-quality mesh. Recently, native 3D generative models have made significant progress, yet they are not built for arbitrary multi-view inputs and often struggle with in-the-wild driving images. In this work, we present MM-TRELLIS, a multi-modal version of TRELLIS for in-the-wild 3D vehicle generation that integrates LiDAR and image sensors from autonomous driving datasets into native 3D generative models. Specifically, multi-view images are cycled as conditioning inputs, while LiDAR point clouds provide test-time guidance to ensure geometric accuracy and cross-view consistency. During denoising, we first align the guidance point cloud with the model priors, then enforce consistency between the generated geometry and the guidance point cloud. Finally, we introduce a voxel filtering strategy based on the opacity of 3D Gaussian Splatting to suppress floaters and produce clean meshes. Comprehensive experiments on Waymo dataset demonstrate our method outperforms existing methods in high-fidelity 3D vehicle generation. Code is available at https://github.com/HongliXiao/MM-TRELLIS.
arXiv:2606.24306v1 Announce Type: new Abstract: In this paper, we describe the 3D Nonconforming Virtual Element Method, where the local spaces are enhanced in order to build an $L^2$ projection of virtual functions onto polynomials of degree $k$. This higher order projection provides a more accurate representation of the load term and, more importantly, it enables the construction of the reaction term. We conduct a convergence analysis of the $L^2$ norm and $H^1$ seminorm errors with respect to mesh size and approximation degree. We also compare our results with those obtained using the conforming Virtual Element Method.
arXiv:2505.02281v3 Announce Type: replace-cross Abstract: This paper explores the performance of a random Gaussian smoothing zeroth-order (ZO) scheme for minimising quasar-convex (QC) and strongly quasar-convex (SQC) functions in both unconstrained and constrained settings. For the unconstrained problem, we establish the ZO algorithm's convergence to a global minimum along with its complexity when applied to both QC and SQC functions. For the constrained problem, we introduce the new notion of proximal-quasar-convexity and prove analogous results to the unconstrained case. Specifically, we derive complexity bounds and prove convergence of the algorithm to a neighbourhood of a global minimum whose size can be controlled under a variance reduction scheme. Beyond the theoretical guarantees, we demonstrate the practical implications of our results on several machine learning problems where quasar-convexity naturally arises, including linear dynamical system identification and generalised linear models.
arXiv:2510.23586v2 Announce Type: replace-cross Abstract: Solving power system capacity expansion planning (CEP) problems at realistic spatial resolutions is computationally challenging. Thus, a common practice is to solve CEP over zonal models with low spatial resolution rather than over full-scale nodal power networks. Due to improvements in solving large-scale stochastic mixed integer programs, these computational limitations are becoming less relevant, and the assumption that zonal models are realistic and useful approximations of nodal CEP is worth revisiting. This work is the first to conduct a systematic computational study on the assumption that spatial aggregation can reasonably be used for ISO-scale CEP. By considering a realistic, large-scale test network based on the state of California with over 8,000 buses, we find that well-designed small spatial aggregations can yield good approximations but that coarser zonal models may result in large distortions of investment decisions, e.g., capacity under-investment of up to 41% for the lowest resolution model considered.
arXiv:2604.03413v2 Announce Type: replace-cross Abstract: Using molecular dynamics simulations, we investigate the crystallization pathways of two exemplary systems that form the same complex crystal structure but differ fundamentally in the nature of their particle interactions. One system is composed of point particles interacting via an isotropic pair potential characteristic of metallic compounds, while the other system contains hard polyhedra whose interactions arise from emergent entropic forces. Despite the stark difference in the origins of the particle interactions, we find that both systems are polymorphic and share the same crystal polymorphs. Moreover, the two systems follow the same multistep crystallization pathways, and by examining the complex crystallization pathways on the single particle level, we find that the local structure evolution of the two systems is also similar. By mapping the hard particle system's interaction to an effective pairwise potential, we find that such resemblance arises from the particle interactions being effectively similar.
arXiv:2606.24848v1 Announce Type: new Abstract: We introduce clique-guarded first-order logic with counting (cgFOC), a fragment of the first-order logic with counting FOC [Kuske and Schweikardt, LICS 2017], and we study the complexity of this fragment. In particular, we prove computable upper bounds on the Vapnik-Chervonenkis (VC) dimension of cgFOC formulas and on the graph dimension of cgFOC counting terms on nowhere dense classes of relational structures. Furthermore, we show algorithmic metatheorems for cgFOC for query answering, enumeration, and probably approximately correct (PAC) learning for Boolean and multiclass classification problems on classes of locally bounded expansion. On the other hand, we show that a slight extension of cgFOC is already intractable on trees.
arXiv:2606.24316v1 Announce Type: new Abstract: Data-driven model predictive control (MPC) has become an attractive approach for controlling unknown systems, especially when data are corrupted by noise. However, most existing data-driven MPC methods focus on linear systems, and little attention has been given to nonlinear dynamics under disturbances. To fill this gap, we propose a robust data-driven min-max MPC scheme for unknown nonlinear systems with process disturbances. We represent the unknown nonlinear dynamics using vector fields built from a dictionary of basis functions, yielding an equivalent linear form with unknown matrices. These unknown matrices are characterized by a set-membership representation derived from noisy input-state data. Using this uncertainty description, we formulate a min-max MPC problem. Two online scenarios are studied: i) when state measurements are noise-free, and, ii) when they are corrupted by process disturbance. For each case, we derive a Lyapunov-based semidefinite program (SDP) to compute a stabilizing state-feedback controller. The resulting schemes are shown to guarantee recursive feasibility and either exponential or robust stability of the closed-loop system depending on whether there is process disturbance. Simulation studies on benchmark examples illustrate the effectiveness and competitive performance of the proposed approach compared to existing data-driven and model-based controllers.
arXiv:2606.20344v2 Announce Type: replace-cross Abstract: Machine learning models have scaled to unprecedented sizes, making training across distributed devices the de facto standard in the field. In this work, we explore how quantum communications can make distributed training both more communication-efficient and information-theoretically private, for both classical and quantum learning models. Ring all-reduce is the foundational communication primitive for large-scale distributed training. We present a quantum version that reduces per-link online communication by a provably optimal factor of two using pre-shared entanglement and superdense coding, without requiring the learning model or gradient computation to change. Beyond bandwidth, the primitive enables privacy guarantees that are information-theoretically impossible for any classical protocol, achieving composable {\epsilon}-secure aggregation, via verified entanglement, at a 2x overhead in GHZ copies. Our hybrid quantum-classical communication architecture yields simultaneous communication and security advantages for large scale distributed training, regardless of whether the learning itself is quantum or classical. Finally, we characterise quantum advantages in gradient conflict detection for server-to-client communication under bandwidth constraints, a setting that arises after ring all-reduce is completed, when full gradient broadcast to external clients is infeasible. Two variants of the problem admit different separations. For margin-based alignment testing (\textsc{GapIP}_{\tau}), the quantum advantage is quadratic in the margin parameter: \widetilde{O}({\tau}^{-1}\log P) qubits versus \widetilde{O}(\min(\{\tau}^{-2},P)) bits. For sign-consistency auditing against a private parameter matching (\textsc{TieAudit}_{\epsilon}), the advantage represents an exponential separation in communication complexity: \Omega(\sqrt{P}) bits whereas O({\epsilon}^{-2}\log P) qubits suffice.
arXiv:2606.23835v1 Announce Type: new Abstract: ABACUS is a unified vision-language model that handles object counting, crowd counting, referring-expression counting, and count-faithful image generation without any benchmark-specific training required. Our model is built on existing 3B-parameter unified foundation model and is adapted for object localization tasks using three key innovations: density-aware adaptive zooming with objectness maps for spatial grounding; a boundary-aware count policy via GRPO to eliminate crop-boundary errors; and a cycle-consistent GRPO strategy where the understanding branch self-critiques generated outputs, closing the understanding-generation gap without any external annotations. ABACUS achieves state-of-the-art results across seven benchmarks, outperforming both task-specific specialists and larger generalist models.
arXiv:2606.24828v1 Announce Type: new Abstract: Scientific long-document summarization datasets commonly treat author-written abstracts as gold reference summaries, although their quality and alignment with the source article vary. At the same time, publicly available scientific summarization datasets remain limited in scale and structure for modern long-context models. In this work, we address both challenges by a) constructing and releasing one of the largest biomedical and life science datasets for long-document summarization, containing 1.88 million PMC articles, and b) analyzing the reference quality of author-written abstracts with source-grounded and model-based metrics. We show that author-written abstracts vary in their alignment with the full article and that these quality signals can guide training-data selection. Training on selected high-quality subsets outperforms random sampling at matched training sizes and can match or exceed larger random subsets on factuality-oriented metrics. Our findings suggest that reference quality is an important factor in scientific summarization and that quality-aware data selection can improve training efficiency.
arXiv:2606.23866v1 Announce Type: new Abstract: Autonomous unmanned aerial vehicles (UAVs) must operate safely in dynamic environments and adapt to changing mission conditions. Although deep learning approaches have shown strong performance for navigation and perception, they are often difficult to explain, verify, and modify for safety-critical tasks. We propose a symbolic state-centered UAV agent using the s(CASP) answer set programming system, enabling autonomous task execution with constraint-based commonsense reasoning in a high-fidelity Unreal Engine 5 environment. We fully implement prior work on the VECSR-A system to support multi-step autonomous behaviors including navigation, search, debris detection, precision spraying, object transport, and inspection. The UAV reasons over environmental and spatial constraints, dynamically revising plans when tasks fail or data is insufficient. Because decisions are based on commonsense reasoning, they are guaranteed to be correct and explainable. We evaluate the feasibility of s(CASP) for UAV control in realistic simulated missions. Results show that our framework enables explainable, adaptive autonomy without retraining, handling complex constraint-aware decisions and dynamic task reevaluation.
arXiv:2606.23741v1 Announce Type: new Abstract: Causal reasoning, which encompasses the discovery of causal structures and the inference of causal effects, is fundamental to data-driven decision making. In practice, data for reliable causal analysis are often distributed across institutions and cannot be centralized due to privacy regulations or communication constraints. Federated learning (FL) addresses this by enabling collaborative analysis without raw data sharing, giving rise to the rapidly growing field of federated causal discovery (FCD) and inference (FCI). However, the interdisciplinary nature of this field and the absence of a comprehensive survey present barriers to entry for researchers. This paper bridges that gap by providing a systematic review through multi-dimensional taxonomies. Grounded in the three core design decisions underlying any FCD solution, namely how structures are learned, how data are partitioned, and what structural knowledge each party obtains, we organize FCD along three axes: methodological paradigm, federation topology, and structural scope. We further examine key practical dimensions, including temporal dynamics, data heterogeneity, missing data, and non-identical variable sets. For FCI, we categorize methods by target estimand (average versus individualized/conditional treatment effects) and by estimation strategy, from classical weighting methods to modern deep generative architectures. Unlike prior works that treat FCD and FCI separately, we formalize their connection as complementary stages of a unified federated causal reasoning pipeline, where FCD supplies the structural knowledge required for valid effect estimation in FCI. Finally, we highlight their shared concerns regarding privacy, communication efficiency, theoretical guarantees, and application domains, and conclude by identifying open challenges for future research.
arXiv:2606.24362v1 Announce Type: new Abstract: Tidal modulation of tectonic tremors provides a sensitive measure of fault response to small stress perturbations, yet how this response varies in a mixed fast and slow earthquake system remains unclear. Here we present the first systematic investigation of tremor tidal sensitivity in such a system, focusing on tectonic tremors along the northeastern Japan subduction zone. Using a tremor catalog from 2016 to 2024, we show that the southern end of the Kuril Trench, characterized by tremor migration and relatively weak seismicity, exhibits the strongest tidal sensitivity, whereas the northern Japan Trench shows the weakest response. Spatial analysis further reveals that areas with weaker tidal sensitivity tend to coincide with more earthquakes ($M_j \geq 4$) and denser tremor activity. In addition, tidal sensitivity at the southern end of the Kuril Trench increases from the early to later stages in tremor migration, potentially reflecting changes associated with underlying slow slip processes. Together, these spatial and temporal patterns suggest that tremor tidal sensitivity may be influenced by the relative contribution of other ongoing perturbations. These results highlight tidal sensitivity as a useful probe of the underlying perturbation environment and provide insight into the possible influence of slow slip processes, earthquakes, and other stress changes on tremor-generating regions.
arXiv:2606.24379v1 Announce Type: new Abstract: Fully Homomorphic Encryption (FHE) enables computations to be performed directly on encrypted data while preserving data confidentiality. However, its practical applications remain limited by high computational costs and development complexity. This paper presents ComputeFHE, an open-source C++ library that facilitates the development of privacy-preserving applications based on the TFHE cryptosystem. The library provides encrypted integer and fixed-point data types together with arithmetic, logical, comparison, conditional, and oblivious array-access operations which allow developers to implement algorithms using a familiar imperative programming paradigm. ComputeFHE supports both conventional TFHE arithmetic based on standard two-input logic gates and an optimized Arithmetic Logic Unit (ALU) architecture utilizing FHE-friendly logic primitives. Experimental results demonstrate significant reductions in the number of required bootstrapping operations, achieving performance improvements of up to 3.9x for selected operations. In addition, the library includes a simulation mode that enables testing, debugging, and complexity analysis without performing actual cryptographic computations while providing circuit complexity and bootstrapping costs. Built on top of OpenFHE, ComputeFHE offers a practical and accessible framework for developing and evaluating privacy-preserving algorithms and applications.