Science & Mathematics

Constrained-random verification tests hardware designs by sampling solutions of Boolean formulas. Existing samplers largely follow one of two approaches: CPU-based SAT samplers (e.g., CMSGen), which remain effective even at rare targets but offer limited throughput, and GPU-based gradient samplers, which achieve high throughput at high-probability targets but collapse at rare ones. We introduce Gpap-S, the first GPU sampler built on Boolean constraint propagation (BCP). At its core is Gpap, a new GPU engine for BCP that efficiently propagates many independent partial assignments concurrently on a shared Boolean formula. This capability allows Gpap-S to combine the rare-target robustness of SAT-based sampling with the throughput of GPU parallelism by exploring thousands of decide-and-propagate trajectories simultaneously. On the benchmark families used to evaluate prior GPU gradient samplers, Gpap-S is 2.2-18.7x faster at the high-probability targets considered in those evaluations. At rare targets, where those samplers often fail to generate any samples, Gpap-S achieves a geometric-mean speedup of roughly two orders of magnitude. Across all benchmarks, Gpap-S is also 83.4-1,571x faster than the CPU-based CMSGen. This throughput does not come at the expense of sample quality: under Barbarik, the statistical tester used to evaluate CMSGen, Gpap-S is accepted on every instance we test, as is CMSGen. We release Gpap as a reusable GPU library for high-throughput BCP.

In vision and scientific domains involving spatial data, a key question is whether node feature values vary independently across space or form coherent spatial structures. Classical spatial autocorrelation statistics such as Moran's $I$ and Geary's $C$ quantify this phenomenon but operate on a single scalar variable with fixed neighbourhood weights, limiting their use for high-dimensional graph-like data. We introduce SAIL (Spatial Autocorrelation Index Learner), a differentiable framework that generalises spatial autocorrelation into a learnable operator by jointly learning feature projections defining candidate scalar fields and inter-node neighbourhood affinities. SAIL produces interpretable local and global spatial coherence descriptors, enabling supervised prediction and unsupervised discovery of spatial autocorrelation modes and hotspots. We show that SAIL exactly recovers classical spatial statistics under appropriate parametrisations and, for tasks driven by latent spatial autocorrelation, can approximate the Bayes-optimal predictor with an inductive bias matched to such tasks. Experiments on synthetic data and large-scale biomedical graphs demonstrate statistical correctness and discovery of spatial coherence with applications across vision, geospatial analysis, and biomedicine. The code repository is available at: https://github.com/pkeller00/SAIL.

This paper investigates a multi-objective optimiza tion framework for UAV-mounted reconfigurable intelligent sur face (STAR-RIS)-assisted Sixth Generation (6G)-enabled Internet of Things (IoT) systems. The considered framework jointly supports communication enhancement and radio-frequency (RF) energy harvesting (EH) through adaptive STAR-RIS transmission and reflection control. A weighted multi-objective optimization problem is formulated to jointly optimize the base station (BS) beamforming vectors, STAR-RIS transmission and reflection coefficients, RF power-splitting factor, and UAV trajectory under quality-of-service (QoS), power, and mobility constraints. Due to the strong coupling among the communications, EH, and UAV mobility variables, the resulting problem is highly non convex and difficult to solve using conventional optimization methods. To address this challenge, a hybrid proximal policy optimization (H-PPO) framework is proposed. The developed hybrid actor-critic architecture combines a categorical policy for discrete UAV movement decisions with Beta-distributed policies for continuous communication-control variables. In addition, a two-stage reward mechanism is introduced to enforce QoS feasibility while balancing system sum rate, energy efficiency, and harvested energy. Simulation results demonstrate that the proposed H-PPO framework achieves faster convergence and more stable training behavior than proximal policy optimization (PPO) and twin delayed deep deterministic policy gradient (TD3) baselines. In particular, H-PPO achieves approximately 66% and 60% faster convergence than PPO and TD3, respectively, while maintaining more stable training behavior. Furthermore, the proposed framework effectively balances communication performance and EH under different QoS requirements, STAR RIS sizes, transmit power levels, and user densities.

The widespread adoption of generative artificial intelligence in recent years has brought a parallel rise in prompt injection and jailbreak attacks. By embedding malicious instructions, these exploits manipulate language models to bypass intended behavior and operational boundaries. In traditional chatbots, such attacks are confined to a single session because state memory is not preserved across conversations. However, persistent memory features in modern conversational interfaces enable cross-session attack propagation, establishing a threat vector known as memory injection. This work explores this emerging vulnerability in leading LLM-as-a-service platforms, highlighting its real-world severity and the need for countermeasures. While existing literature focuses on violations tailored to specific platform features, we present a general threat model. We demonstrate that all NIST attack goals are collectively achievable across ChatGPT Instant 5.3, Claude Sonnet 4.6, and Gemini 3 Flash under this model. We also show that memory injections can exploit multiple attack vectors, including personal memories, projects, skills, organization instructions, and Model Context Protocol (MCP)-connected resources. Our findings reveal that all evaluated services can be manipulated to persistently spread disinformation, generate harmful content, or simulate offline unresponsiveness, even in fresh sessions. In the particular case of Claude, shared-memory features allow an attacker to compromise multiple users within an organization and also exfiltrate user prompts. To address these vulnerabilities, we derive defense mechanisms from our proposed threat model and foundational cybersecurity principles. Our analysis indicates that nearly half of these mitigations remain unimplemented across the target services. Datasets and attack artifacts are available at https://github.com/PedroBarcha/Memory-injection.

Agentic systems have demonstrated strong capabilities on complex tasks by enabling multiple interacting large language model (LLM)-based agents to collaborate and execute subtasks in parallel. However, this collaborative structure introduces new challenges for reliably evaluating and diagnosing system failures, particularly failure attribution: tracing a system's execution to identify the agent step responsible for an incorrect final outcome. Because manually performing this task is time- and labor-intensive, recent work has turned to LLMs as judges to automate the process, referred to as automated failure attribution (AFA). These approaches assume that a sufficiently capable model can infer the failing step directly from an interaction transcript. In practice, however, direct prompting often performs poorly, which motivates follow-up work that explores various ways to improve AFA performance, including alternative prompting strategies and inference-time interventions. In this paper, we examine a more fundamental question: whether this weak behavioral performance reflects a genuine capability limitation. Using linear probes applied to internal activations, we uncover a significant \emph{knows-but-won't-say gap:} models encode sufficient information to attribute a failure correctly but fail to explicitly output the responsible step. Building on this finding, we propose \emph{grafting}, a simple manual intervention that exploits knowledge already encoded in the model to improve AFA performance without requiring ground-truth labels at inference time. Extensive evaluations across multiple LLMs and benchmark settings demonstrate that grafting consistently outperforms direct-prompting approaches, underscoring the potential of leveraging latent model knowledge for more reliable failure attribution. Our code is available at https://github.com/wendyzheng22/mi4afa.

The initiation of detonation in high explosives (HEs) under shock loading is strongly influenced by the collapse of porosity and the subsequent formation of hotspots. As these materials are widely integrated with polymer binders to form polymer-bonded explosives, the collapse of pores at explosive-binder interfaces has received significant interest. Recent molecular dynamics (MD) simulations revealed that the shock-driven expansion of soft polymers (e.g. polystyrene) into a void produces greater pressure-volume (PV) work during recompression and higher local temperatures compared to pore collapse in neat HEs, ultimately expediting the onset of chemical reactions. Yet how pore collapse, interfacial phenomena, and binder chemistry jointly affect initiation remains insufficiently understood. We provide deeper insights into the role of polymer in hotspot evolution through reactive MD simulations, specifically characterizing its chemical behavior during the shock-induced collapse of voids at sizes approaching experimentally relevant length scales. We consider the HE 1,3,5,7-Tetranitro-1,3,5,7-tetrazocane (HMX) in a crystal-gap-crystal stack with polystyrene along the upstream surface. In the hydrodynamic collapse regime, hotspots nucleate at the material interfaces, with void length-dependence in PV work, shock temperature, and the onset of self-sustaining chemistry. Driven by shock-induced deformation and HMX decomposition, the binder undergoes backbone chain scission followed by extensive aromatic-backbone cross-bonding, akin to a graphite to diamond transformation. These reactions act as heat sinks, suppressing deflagration and ultimately inhibiting hotspot growth and coalescence between crystals. The simulations further link the polymer's chemical response to the shock-to-detonation transition in polymer-bonded explosives.

In contact-rich manipulation, contact conditions can change before a visual policy updates its action plan. Recomputing the plan through the visual policy adds latency, while tactile feedback alone does not specify how the ongoing motion should be revised for the task. We introduce T-WLA, a world-language-action architecture that uses shared visual memory to guide both action generation and tactile correction. At each visual update, the visual branch integrates the current image, task instruction, and retrieved history into a visual memory that conditions base-action generation. The same memory is cached for a lightweight tactile editor, which combines it with the active action chunk and incoming touch to predict residual corrections to unexecuted actions without rerunning the visual branch. Future-image and future-pressure prediction provide training-only supervision, encouraging the visual memory and tactile features to capture interaction dynamics relevant to action generation and correction. We evaluate T-WLA on tactile-conditioned hand-motion prediction using EgoTouch and Tachin, and on four real-world dexterous manipulation tasks. T-WLA reduces finger-joint MPJPE by 30% relative to the strongest reported baseline and achieves an average success rate of approximately 50.0% across real-robot manipulation tasks. Videos and code are available at https://anonymous-submission-20.github.io/T-WLA.github.io/.

Agentic harness evolution optimizes an agent's instructions, tools, and control flow by using its task execution experiences as evolving evidence. While more evolution tasks should intuitively improve performance, our preliminary study reveals a paradoxical failure mode: the performance of existing methods degrades as the task pool expands. Our further analysis demonstrates that this degradation stems from suboptimal evidence selection during optimization, as different evidence combinations lead to substantially different evolution outcomes. To address this issue, we propose Successive Scale Racing (SSR), a framework that adaptively searches for optimal evidence combinations in each evolution round. SSR constructs diverse evidence combinations based on the corresponding task quality and coverage, and efficiently races the resulting candidate harnesses under a small compute budget. Empirically, across three distinct benchmarks, SSR achieves state-of-the-art results in all six configurations. Notably, as the task pool grows from 40 to 200, SSR steadily improves the pass rate from 64.3% to 68.3%, while the baselines' performances all drop. Furthermore, its racing mechanism identifies near-optimal candidates while reducing validation executions by 60%.

We investigate phase-coherent multistatic sensing in distributed ISAC systems where spatially separated access points (APs) share a common, non-ideal local oscillator (LO) impaired by phase noise (PN). In the presence of a single transmit AP and multiple receive APs with OFDM transmission, we derive the joint statistics of PN across receive APs and reveal its unique covariance structure that depends on target location through propagation delays of individual bistatic links. Based on this statistical characterization, we carry out a Cramér-Rao bound analysis of three receiver models, namely PN-free, PN-aware and PN-ignorant, for both isotropic and non-isotropic targets. Simulation results verify the location-dependent statistics of PN in shared-LO distributed architectures while demonstrating how transmit power, LO quality, AP deployment and scattering characteristics affect multistatic localization.

As systems engineering (SE) shifts from monolithic systems to Systems of Systems, Mission Engineering (ME) has emerged as a new line of thinking in SE. Mission environments are uncertain and dynamic, and outcomes depend on how mission assets interact with them, rendering static architectures brittle and calling for flexible yet analytically rigorous ME approaches. We propose an intelligent mission coordination methodology integrating high-fidelity digital mission models with Reinforcement Learning (RL) for adaptive task allocation. Using a Digital Engineering (DE) infrastructure comprising a digital mission model and agent-based simulation, we formulate mission tactics management as a Markov Decision Process and train an RL coordinator with Proximal Policy Optimization to map mission states to tactical actions from realized outcomes. An aerial firefighting mission with three scenarios demonstrates the approach. First, all trained coordinators significantly outperform random and heuristic baselines. Second, single-scenario coordinators can be brittle: the coordinator trained on the more forgiving scenario degrades to random tactics performance in more demanding scenarios, whereas the demanding-scenario coordinator retains most of its advantage when transferred. Third, within the same training budget, a coordinator trained on both scenarios concurrently matches the scenario-specific coordinators in their own scenarios and significantly outperforms every other policy. Sensitivity analyses on ignition location, weather, scenario, and training curricula show that the variability represented during training shapes post-deployment performance. This proof of concept shows DE-enabled mission simulations combined with RL offer a reusable methodology for improving ME practice, which can be extended to fleet design and selection from a mission-first perspective.

Common instruction following (IF) benchmarks, such as IFEval, evaluate the ability of language models (LMs) to follow simple constraints that can easily be verified programmatically. These same constraints, in turn, are used as reward signals for reinforcement learning. This design may lead to reward hacking of constraints, pushing models away from the main task. Recent attempts to address this resulted in benchmarks with richer constraints, but required expert labor, remained closed-access, and cannot be easily extended. We present Lift-Bench, an IF benchmark that asks a model to draft a document given a topic and a structured outline. Constraints are scoped to certain document sections, providing finer control over constraint density. Lift-Bench includes both verifiable and judged constraints, as well as a training counterpart, Lift-Train. Importantly, Lift-Bench provides both main task and instruction-following constraints, mitigating shallow reward hacking. On experiments with multiple models, we find that frontier models struggle with the harder subsets of our benchmark and analyze model performance along multiple axes. We also find that training on Lift-Train is much less susceptible to reward hacking than training on the common IFTrain. Lift-Bench and Lift-Train are publicly available, as well as the data generating pipeline.

We study random access in linear codes when codeword symbols are sampled independently and uniformly with replacement. A code is recovery balanced if the expected number of samples required to recover a codeword coordinate is the same for every coordinate. We express the recovery expectations in terms of the associated matroid and derive a coordinatewise identity relating these values for a code and its dual. This proves that a linear code is recovery balanced if and only if its dual is recovery balanced, establishing a conjecture of Gruica et al. We further define the recovery balance for arbitrary matroids and show that asymptotically almost every sparse paving matroid is not recovery balanced.

We present Register Anything Model (RAM), a unified model for generalizable 3D point cloud registration that robustly estimates 6-DoF transformations across diverse domains spanning sensors and environments. Existing registration methods are typically tailored to a single domain, with carefully tuned architectures and hyperparameters, and thus degrade markedly when applied to novel domains. We bridge this gap by training a single model on diverse data from indoor, outdoor, and object domains. A central challenge is the severe geometric discrepancy across domains, i.e., differences in scene scale, sensor noise, point density, and structural patterns, which hinders processing with a unified point-based model. To address this, we introduce a geometry-aware normalization strategy that jointly voxelizes and normalizes point clouds into a standardized geometric space, adaptively based on their geometric properties to preserve salient local structures while standardizing resolution. Built on this normalization, RAM follows a hierarchical coarse-to-fine paradigm, and employs an efficient geometric transformer that models local geometric consistency instead of globally over all points. This design mitigates discretization noise introduced by voxelization, and improves computational efficiency. Extensive experiments on 12 benchmarks demonstrate state-of-the-art performance in terms of inlier ratio and registration recall. Notably, our method outperforms the counterparts trained on each separate domain, showing superior generalization capability.

Dependency parsing resource development for Arabic has largely focused on Modern Standard Arabic, leaving dialectal Arabic comparatively under-resourced. We introduce MADARTB, the first large-scale parallel dependency treebank spanning five city dialects: Beirut, Cairo, Doha, Rabat and Tunis, and extending an existing parallel MSA version. In total MADARTB consists of 5,000 sentences annotated using the CATiB dependency scheme. We analyze syntactic similarity across all city dialects and MSA. Although global construction profiles exhibit high similarity across varieties, local syntactic structures frequently diverge, revealing substantial structural variation across dialects. We further establish a cross-dialect dependency parsing comparison using a supervised parser and four LLMs. The supervised parser achieves 87.5% LAS, while the best LLM setup trails at 85.7% LAS across all six varieties on the test set. We release our annotations and models publicly.

Automated judges of clinical artificial intelligence (AI) outputs are necessary, but large language model (LLM) judges are costly and slow at scale. We compared Jev, a decision language model that returns structured judgments, with frontier and open-weight LLM judges in four use cases. On 118 pediatric answers, Jev's agreement with physicians' ordinal quality ratings (Gwet's AC2 0.50; expected score 0.53) matched or exceeded that of seven of 12 LLM judges, including frontier models, Claude Opus 5.5 (0.49) and GPT-6 (0.38). On physician-graded binary HealthBench criteria, Jev matched or exceeded four of six LLM judges. On MedHELM's open-ended tasks, Jev's agreement with a three-LLM jury approached or exceeded the jurors' internal agreement. On HealthBench Professional, Jev matched the benchmark's reference judge, GPT-5.4, on 91% of criteria (other judges, 91-93%). Frontier judges cost 73-294 times more than Jev and took 7-57 times longer. Jev may enable efficient clinical AI evaluation at scale.