Papers

Anomalies are rare, and anomalous data are often unavailable during development, making it difficult to determine which anomaly detection models and configurations will generalize to unseen anomalies. Recent approaches address this challenge by generating pseudo-anomalies and using bounds on the achievable area under the ROC curve (AUC) to select the optimal configuration from a finite set of candidates. Instead, we use the AUC bound as a differentiable, anomaly-free objective for directly optimizing continuous parameters of anomaly detection systems. We demonstrate this framework by optimizing ensemble weights and introducing a learnable score-rescaling mechanism that adapts pseudo-anomaly scores, enabling optimization beyond a predefined candidate set. Experiments across multiple datasets and embedding models show that AUC-bound optimization achieves significant performance gains over conventional model selection and prior development-set-based parameter selection. The results further show that direct optimization is less sensitive to the choice of pseudo-anomaly construction.

Vietnam's Personal Data Protection Law (Law No. 91/2025/QH15) and Decree No. 356/2025/ND-CP, effective January 1, 2026, require organizations to establish and maintain Records of Processing Activities (RoPA). Manual RoPA preparation is labor-intensive, while cloud-hosted large language models (LLMs) may conflict with data-sovereignty requirements. We propose RoPA Manager, a system for automated RoPA information extraction using hybrid retrieval that combines lexical ranking over tsvector, dense-vector search, Reciprocal Rank Fusion (RRF), and locally deployed LLMs. We introduce a Vietnamese RoPA benchmark with 32 organizations, 77 processing activities, 12 field groups, and 4,338 reference values. Evaluation is reported at three distinct levels. The automated scorer, tested on perturbed data without invoking an LLM, achieved F1 = 0.9493 [0.9436, 0.9548]; this measures scorer robustness rather than end-to-end extraction accuracy. End-to-end extraction achieved token coverage of 50.04-55.25% against the reference labels. Two independent experts reviewed 1,558 reference values (35.9% of the benchmark), found no incorrect values, and achieved 99.68% agreement with PABAK = 0.9936. Value-level precision was not measured. Across 32 paired scenarios on a 24 GB GPU, locally deployed Qwen3.5-27B-GPTQ-Int4 showed no statistically significant difference from cloud-based DeepSeek-V4-Flash (difference 0.20 percentage points in favor of DeepSeek, 95% CI [-0.93, 1.32], p = 0.72), while Gemma-4-31B performed significantly worse (p < 0.01).

Field robots must traverse varied terrain and obstacles while remaining robust to water, debris, vegetation, and physical contact. We present MARBLE, a fully enclosed omnidirectional amphibious rolling robot driven entirely by internal mass redistribution. Three mutually orthogonal linear sliders shift internal masses to generate body rotation, while an orientation-aware controller maps planar velocity commands into slider positions. A rigid spherical shell encloses all active mechanisms and simultaneously serves as the terrestrial contact surface, buoyant enclosure, and mounting structure for passive fins that enable water-surface propulsion. Rotation of the same shell architecture hence produces rolling on land and surface propulsion in water without mechanical reconfiguration or separate locomotion actuators. The spherical morphology further allows the robot to accommodate changes in body orientation and contact location during direct interactions with terrain and obstacles. We evaluate MARBLE through omnidirectional locomotion characterization, traversal across heterogeneous terrestrial environments, aquatic surface locomotion, land-water transitions, and deliberate obstacle interactions. These experiments demonstrate how a single enclosed mechanical architecture can combine omnidirectional mobility, cross-medium locomotion, and tolerance to environmental contact. MARBLE provides a compact design for field mobility across heterogeneous terrain, obstacles, and land-water transitions. We will open-source all software and hardware design. Our website is https://generalroboticslab.com/MARBLE

Malware evolves faster than rule-based and signature-driven detection pipelines. This paper presents SAGEGAN, a benign-only trained malware anomaly detection framework that converts portable executable files into compact three-channel images and models benign structure through style-conditioned adversarial reconstruction. The representation combines Hilbert-mapped byte values, benign-referenced byte-transition surprise, and entropy deviation from benign software. The model encodes each image into a layer-wise style tensor aligned with a seven-stage modulated generator, rather than a single latent bottleneck. A Gaussian style prior, moment-based prior alignment, and latent consistency are used to reduce mismatch between encoded benign styles and the generator's sampled manifold. For interpretation, a deterministic encoder pathway maps each executable to a fixed style tensor, enabling repeatable layer-wise family distance, gradient sensitivity, principal component, and class-behaviour analyses. On a self-collected portable executable corpus containing malware from 214 families, the Gaussian variant achieves 89.76% area under the receiver operating characteristic curve and 88.19% balanced accuracy, while the genome-style variant reaches 88.03% and 84.09%, respectively. Without refitting model weights, benign reference statistics, or decision thresholds, the same checkpoints are evaluated on DIKE, Microsoft BIG 2015, and Lester malware subsets. The results suggest that layer-wise style modelling supports both anomaly ranking and structured post hoc analysis of how malware families depart from the benign manifold.

A region can overlap a target object yet contribute more to another class. We test regions selected by Cluster-based Concept Importance (CCI) in frozen CLIP. Across COCO and VOC with two checkpoints, 41.08-64.78% of regions that pass overlap and mean-contrast checks fail against the strongest competing class. Removing competitors annotated in the image leaves 39.69-63.64% failing. We then test all eight candidate regions per image. An alternative passes the test for 6.25-7.84% of failures on COCO and 27.40-31.15% on VOC. Requiring it to preserve the original target-score drop within $\epsilon = 0.02$ reduces these rates to 0.16-0.98%. Available regions and target-drop tolerance constrain repair; relaxing the tolerance increases repair opportunities.

Unlearning ensures LLM compliance by removing the influence of private or copyrighted training data. However, since LLM models typically undergo post-training compression, like quantization, in practical deployment, it has been observed that the unlearning effect can be substantially weakened, with the forgetting behavior degrading more severely than that of model utility. This paper proposes a quantization-robust unlearning framework that makes forgetting robust to quantization while maintaining overall model utility. We analyze this gap through the lens of loss landscape. Specifically, our analysis reveals a curvature-based criteria that pinpoints sensitive weights in the unlearned model that leads to both non-robust forgetting and reduced utility. We therefore propose sensitivity-guided noisy regularization, which is applied on the sensitive parameters to steer the model convergence towards a smoother minima of uniformly low forget and retain losses. Balancing unlearning and utility, we further propose forget-critical optimization, which updates only forget-critical layers, preserving most of the network to retain useful knowledge. Extensive experiments on the MUSE and TOFU benchmarks across multiple LLM unlearning algorithms show that our approach achieves substantially more quantization-resilient forgetting while maintaining utility. Code is available at https://github.com/jialuw0830/QRUnlearn.

Generative AI systems are increasingly deployed to address domain problems. These systems operate under technical, regulatory, institutional, and normative constraints that define acceptable AI behaviour and outcomes within their domains. We observe a recurring pattern in our industry engagement: partners often arrive with a functioning but relatively generic AI solution. The challenge is no longer to build an AI system from scratch, but to improve the quality and domain appropriateness of an AI-generated solution. In these settings, the limiting factor is often the quality, scope, and structure of the context available to the system. Yet, existing context engineering approaches primarily focus on supplying domain knowledge through retrieval, memory, and tools, with limited support for systematically identifying and operationalising the constraints that govern AI systems in their operational environments. This paper proposes Constraint-Driven Context Engineering (CDCE), a design approach for engineering domain interfaces for AI systems. Drawing on software architecture design and Domain-Driven Design (DDD), CDCE treats domain constraints as first-class design drivers. It identifies and characterises constraints, determines the required context assets, and designs representations through which these assets are made available to AI systems. We conducted a comparative multiple-case study with industry and public-sector partners across educational assessment, healthcare decision support, and financial-distress prediction. Depending on their characteristics, constraints can guide AI behaviour, enforce permissible boundaries, or support verification of AI-generated outcomes. The cases demonstrate CDCE's applicability across contrasting domains and show how constraint characteristics shape the resulting domain interfaces.

Web agents are usually evaluated in live environments, where environment state and judge models drift between runs, so the same checkpoint rarely reproduces the same score, making controlled studies of training phenomena impractical. We present WebMRE, an offline benchmark of 541 tasks and 5,293 steps derived from successful WebArena trajectories, with fully audited test labels and a deterministic protocol that scores a checkpoint identically on every run without any environment. Each step pairs a human oriented guide sentence with a grounded action, enabling the first study of the mutual reinforcement effect between them in web agents. Averaged over three seeds the effect holds for both models in both decoding orders and grows with scale: jointly decoding a guide lifts element selection over an action only reference by 0.9 and 0.2 points for Qwen3.5-4B and by 1.7 and 2.2 points for Qwen3.5-9B. A mediation analysis shows that the guide is a causal channel rather than commentary: forcing the gold guide as a decoding prefix lifts action accuracy from .422 to .684, another step's guide collapses it to .055, and a paraphrase that renames the target still recovers half of the gain, so the channel carries instruction meaning and not only the label string. The same channel yields an offline reward that only a replayable protocol makes computable, though optimizing it from a strong checkpoint brings no gain yet. Our fine tuned models outperform GPT-5.5, Claude Opus 4.8, and Gemini 3.5 Flash, run zero shot, on every offline metric.

Real-world molecular design remains challenging for large language model (LLM)-based agents. It requires them to interpret design contexts, satisfy multiple constraints, identify infeasible specifications, and reason over multi-step tool outputs. Existing benchmarks do not capture this complexity, focusing instead on explicit and narrow constraints, only feasible problems, and single-path solutions. To address this gap, we propose MolDesignBench, a scenario-grounded benchmark that more closely reflects real-world molecular design for evaluating tool-augmented LLM agents. MolDesignBench comprises 2K generation and optimization instances that combine implicit requirements embedded in design narratives with explicit property and functional-group constraints, including infeasible cases, and require the effective use of 17 specialized chemistry tools. Experiments across diverse frontier LLMs reveal low success rates–with the best achieving only $\sim43$%–and frequent failures in implicit-constraint reasoning, infeasibility detection, and tool reasoning. The corresponding fine-grained failure-mode analysis identifies implicit constraint interpretation and infeasibility detection as the primary bottlenecks, establishing MolDesignBench as a rigorous testbed to guide future research on chemical agents. The benchmark, tool interface, and evaluation code are publicly available.

We study trust-region minimization of a smooth, possibly nonconvex functional plus a convex functional composed with a bounded linear operator. A relative primal–dual gap condition controls both the error in an approximate proximal-gradient step and its linear-model decrease. Together with a computable absolute stationarity test, it yields a finite Cauchy search, convergence of the proximal stationarity measure to zero, and an $O(\varepsilon^{-2})$ bound on outer trials. The outer analysis allows the linear operator to take values in a Banach space and does not require dual attainment. When the operator takes values in a Hilbert space and the regularizer is finite and Lipschitz, the dual proximal-gradient method produces finite gaps tending to zero, provided the required proximal maps and functional values can be evaluated. We prove $O(j^{-1})$ gap bounds for both recovered and averaged primal candidates and give a sharper bound on the primal error for exactly recovered points. A semilinear elliptic control problem with unsmoothed total-variation regularization and an $L^2$ control cost illustrates the method in the full $H^1$ metric. Across five meshes, outer and state Newton counts remain constant, while interior-point iteration counts vary mildly.

Reliable localization of non-line-of-sight (NLOS) pedestrians is critical for safe urban autonomous driving, yet it remains highly challenging in ego-dynamic outdoor environments, where ego-vehicle motion makes radar multipath propagation complex and noisy. In this paper, we present a reflection-aware framework for NLOS pedestrian localization with a moving ego-vehicle in outdoor testbed scenarios. Our framework fuses front-view camera images and 2D radar point clouds to infer reflection orders and reflective surface distributions in bird's-eye-view space. It then uses physics-guided ray tracing to reconstruct distorted reflection paths and localize the hidden pedestrian. We validate the framework in outdoor testbed scenarios under ego-dynamic conditions. The results demonstrate the effectiveness of the proposed framework for NLOS pedestrian localization with a moving ego-vehicle.

Cutting changes both the shape and topology of deformable objects, making accurate simulation challenging for robotic manipulation. A simulator must track the cutting tool as a cut develops, preserve the resulting discontinuities after tool withdrawal, and enable newly exposed surfaces to interact with the tool and with each other. Existing formulations often prescribe cut surfaces in advance or couple material separation to auxiliary geometric fields. We introduce BladeMaster, a GPU-accelerated cutting framework based on the total Lagrangian material point method (TLMPM). Our key idea is to encode the cutting history directly on material points through persistent side labels generated online from the blade geometry. These labels govern particle-grid coupling, preserving connectivity within intact material while preventing spurious coupling across cut faces after tool withdrawal. Our formulation supports progressive and intersecting cuts without predefined cut surfaces or particle duplication. Material-material contact enables cut surfaces to recontact and slide against each other without reconnecting, while two-way tool-material coupling allows material reaction forces to influence tool motion. Experiments demonstrate tool-driven cutting followed by manipulation, with faster-than-real-time performance on representative tasks. Project page: https://jango6324.github.io/blademaster/.

Large language models (LLMs) are increasingly used as natural-language interfaces for robotic systems, yet their integration with Robot Operating System (ROS)-based navigation remains limited by two gaps. First, navigation data such as occupancy grids are represented as raw geometric messages that are difficult for LLMs to use directly as spatial or semantic context. Second, adding LLM-driven capabilities often requires custom wrappers or robot-specific interfaces, limiting reuse across systems. To address these challenges, we propose a non-invasive framework that connects LLM reasoning with ROS-based navigation through a navigation-oriented representation layer, exposed through the Model Context Protocol (MCP) as standardized, reusable tools so that any MCP-compatible LLM can access them without robot-specific wrappers. The visual map modules transform occupancy grids into metric, pose-aware images for goal reasoning, while the semantic annotation modules record waypoint-level observations with robot poses. We evaluate the framework on three tasks: autonomous mapping, spatial reasoning-based navigation, and semantic reasoning-based navigation. The results show that the evaluated LLM backends use these representations to achieve over 97% map coverage and select spatial or semantic navigation targets from natural-language instructions in a simulated indoor environment. This demonstrates representation-mediated LLM navigation without modifying the existing ROS navigation stack.

Learned local navigation in crowded indoor environments is sensitive to how dynamic obstacle motion is represented, while collision-prone behaviour may persist after nominal policy training. We present a risk-aware reinforcement-learning framework that addresses these two issues through an uncertainty-aware Dynamic Uncertainty Grid Map (DUGM) and a modular post-training recovery mechanism. DUGM combines local occupancy, estimated obstacle motion, and motion-estimation uncertainty in a robot-centric representation. After the nominal policy is frozen, a finite-horizon Risk Value Function (RVF) is trained from nominal rollouts and used to trigger a dedicated recovery policy when continued nominal execution is predicted to be collision-prone. Experiments in a held-out NVIDIA Isaac Sim clinical-logistics benchmark show that uncertainty-aware dynamic representation improves nominal navigation over static and deterministic alternatives, while the recovery mechanism further mitigates residual collision-prone behaviour. The complete framework is also deployed directly on a TurtleBot3 without policy fine-tuning, retraining, or site-specific adaptation, retaining the performance trend observed in simulation. These results indicate that uncertainty-aware dynamic representation and post-training recovery provide complementary mechanisms for improving learned local navigation.

Open RAN (O-RAN) slicing xApps must adapt resource allocations to changing channel conditions and traffic demands while meeting service-level agreements (SLAs). Deep reinforcement learning can produce adaptive policies, but their allocation rules remain encoded in neural-network parameters. Our goal is to retain this adaptability while making the controller's decision logic directly inspectable and editable by operators. We use a large language model (LLM) to evolve slicing controllers as compact Python programs whose decision logic remains readable and editable after optimization. The LLM proposes and revises candidates offline, while a calibrated simulator scores them, and the selected decision module runs unchanged in the O-RAN control path. On the NSF POWDER 5G testbed, the evolved controller releases resources from a guaranteed slice whose throughput target becomes unattainable under a sustained channel fade, improving best-effort throughput from 158.2 to 228.6 Mbps, a 44.5% gain over the best static allocation. Since the controllers are readable source code, their behavior can be predicted from their equations, defects can be diagnosed by reading the code, and calibration errors can be corrected with one-line edits, reducing SLA misses from 79.9% to 2.2% in one case and more than doubling fitness in another. In a four-slice trace-driven simulation calibrated to the same testbed, evolutionary search achieves higher average evaluation scores than independent prompting at a matched proposal budget, with mean normalized gains on held-out traces of 16.3% for prompting alone, 32.1% for evolution from scratch, and 51.0% for evolution from a starting program.

Automated red teaming often replays a fixed set of prompts, which measures known risks but cannot learn from failures found during testing. We present CART (Closed-Loop Adaptive Red Teaming), a framework that uses each result to guide what it tests next. CART begins with broad risk coverage, follows weaknesses that emerge, keeps new probes diverse, and records the evidence and source of every finding. It separates the Challenger that creates tests, the Target being tested, which may be a text-only model or a bounded tool-using agent, and the Judge that evaluates the results, allowing these roles to be studied independently. Across three evaluation families (Frontier, JAH, and Agentic), CART discovers more failures and higher average risk than static seed replay for every Target with an available baseline. The gains extend to tool-mediated agent tests, suggesting that contextual adaptation can reveal weaknesses that direct prompt replay does not exercise. These results describe what the test policies discover, not how often failures occur in real deployments. We also find that Challenger-Judge choices affect the evidence uncovered, highlighting the need for role separation and independent review. Overall, CART turns red teaming from a one-time checklist into a continuous, adaptive, and auditable search for model and agent weaknesses.

Agentic memory systems reuse past experience to improve future performance, yet most existing designs curate memory at write time: once a task is completed, its trajectory is distilled into a fixed artifact, such as a reflection, workflow, skill, or reasoning strategy, that is later retrieved by similarity. This forces the system to decide what is worth remembering before the future query is known, irreversibly discarding information and producing a query-independent summary that must serve many possible downstream tasks. Learning such a write-time curator is also difficult because the value of a storage decision may only become apparent when a relevant query arrives, potentially many tasks later, creating a long-horizon credit-assignment problem. We instead retain raw trajectories and defer curation until read time, when the current task is known. Given the retrieved traces and the new task, a memory curator synthesizes a compact, task-adaptive payload tailored to the immediate need. Because this payload is consumed on the same task, the curator can be trained directly from immediate task success, avoiding delayed utility signals and the need to artificially group related tasks. Across ALFWorld, WebShop, and $\tau^2$-bench, our Just-in-Time Memory (JitMem) consistently outperforms no-memory agents as well as heuristic and learned write-time memory methods, improving over the strongest baseline by 16.2, 16.3, and 3.9 absolute success-rate points, respectively. Notably, even an untrained curator is already competitive with or surpasses these baselines, showing that task-adaptive read-time curation itself is a major source of the gain; training the curator further compounds the improvement.

Platform operators increasingly rely on system prompts and fine-tuning to govern model behavior, yet it remains unclear how reliably these interventions override behavior inherited from prior training. We propose Override Success Rate (OSR) and alignment inertia to measure when operator interventions succeed or fail to change prior behavior. We evaluate zero-shot prompting and LoRA fine-tuning across Llama and Mistral in medical misinformation and hate speech. Alignment inertia persists across both models but varies by model, domain, and policy direction. Notably, in Mistral's restrictive hate-speech condition, LoRA increased inertia by 46.5 percentage points, showing that fine-tuning can reinforce rather than override prior behavior. We also use TRAK to test whether inertia is associated with weaker adaptation signals. TRAK achieves AUC of at least 0.85 in 7 of 8 conditions and outperforms model confidence, TF-IDF similarity, and embedding similarity as a predictor of inertia. These results provide an operator-facing audit of where prior training constrains downstream model governance.

Long horizon agents accumulate growing interaction histories that increase context and inference costs. We find that geometric redundancy alone is an insufficient criterion for safe compression. Although agent histories exhibit strong low dimensional structure, similar global geometry can preserve very different amounts of task evidence. At identical retained block counts, evidence aware selection raises next action Top 3 retention from 0.31 to 0.69, while centroid similarity remains 0.98. Controlled replacement further shows that action related information can be substantially altered while global geometric measures remain nearly unchanged. Motivated by this gap between geometry and evidence, we introduce Geometry Guided Evidence Preserving Memory (GEM), a training free compressor that protects task and execution evidence before using geometric residuals to complete coverage. GEM reduces mean combined token usage from 2.69M to 2.11M per task, a 21.4% reduction, while maintaining comparable task reward. Our results show that efficient agent history compression should optimize for preserved task evidence rather than geometric coverage alone.

We present JEV-Star, a StarCraft II controller that defeats the strongest non-cheating built-in AI, Lv7, by combining fast JEV action selection with persistent GPT-6 planning. The combined system wins four full games at Lv5–Lv7, including two Lv7 victories with different seeds, while retaining a median JEV response time of 0.422 seconds. Mean estimated model cost across these games is USD~3.71 per game: USD~0.15 for JEV and USD~3.56 for GPT-6. We compare this system with an initial JEV-only controller in full-game macro control and multi-unit micromanagement. The standalone controller reaches a 20-minute limit against Lv2 without expanding. Across 35 battle maps with three episodes per map and controller, the combined system raises mean enemy elimination from 16.50% to 37.69% and wins from 3 to 7 out of 105. Replay frames and decision logs document resource reservation, persistent economic goals, and stable army objectives in successful full games. The results demonstrate a practical division between inexpensive, subsecond decisions and longer-horizon planning. The comparison evaluates complete systems; concurrent interface improvements mean that planning's contribution is not isolated by a controlled ablation. Code is available at https://github.com/sc2musa/Jev_Star

We develop a sample-based framework for constructing information-driven hierarchical multi-resolution representations of probabilistic occupancy grids. Recent methods compute information-optimal abstractions via dynamic-programming-based exhaustive recursions, which become computationally prohibitive for large-scale grids and are ill-suited to robotics applications. To address this limitation, we introduce a sample-based strategy inspired by Monte Carlo Tree Search (MCTS) that incrementally constructs hierarchical abstractions through statistical estimation rather than exhaustive enumeration. The proposed method is anytime in nature, allowing computation to be terminated at any stage to produce a valid compressed representation. We compare our approach with the information-optimal Q-tree search algorithm and demonstrate its effectiveness in rapidly generating abstractions of large real-world probabilistic occupancy grids.

Multi-Fidelity optimization (MFO) has received extensive attentions in engineering design, which resorts to augmenting the small number of expensive high-fidelity (HF) samples by a large number of low-fidelity (LF) but cheap samples to improve the optimization performance. A key factor that influences the effectiveness of MFO is how to adaptively assign samples for HF and LF simulations in the iteration process. To address such sample assignment issue in MFO, we propose a new infill criterion named as Filter-GEI, which imposes an adaptive filter function on top of the generalized expected improvement (GEI) acquisition function. In particular, by taking the correlations between HF and LF models into account, the Filter-GEI can efficiently allocate HF and LF samples to achieve a good balance in between the local and global search. Furthermore, considering parallel computing, the Filter-GEI infills multiple HF and LF samples in each iteration, which can further improve its efficiency as computing power increases. Through tests on five mathematical toy problems and one engineering problem for the turbine blade design, the effectiveness of the proposed algorithm has been well demonstrated.

Video provides a rich record of human behavior, interaction, and situated contexts, offering important evidence for understanding people and conducting human-centered research. As vision-language models (VLMs) become increasingly capable of analyzing video, they offer opportunities to automate this traditionally human-intensive process. Yet a central question remains: when can VLMs analyze human-centered video independently, and when does reliable analysis still require human involvement? To address this question, we first characterize video analysis practices in human-centered research. We systematically analyze all 1,702 CHI 2026 full papers and identify 125 that annotate videos. Through iterative coding, we derive a five-dimensional taxonomy spanning analytic purpose, viewpoint, phenomenon, reasoning requirement, and annotation authority. Grounded in recurring annotation tasks captured by this taxonomy, we construct a benchmark of 15 representative tasks from open datasets to map the capabilities and limitations of a general-purpose VLM. We examine the division of labor between humans and VLMs by comparing three annotation workflows: VLM alone, human alone, and human verification of VLM outputs. Across tasks, VLM-alone annotation approaches human accuracy on average (HNS = 97.0, where 100 denotes human-alone performance), demonstrating substantial potential to automate human-centered video analysis. Human verification achieves the highest accuracy (HNS = 121.5) while reducing human annotation time by 48.9% and monetary cost by 31.3%-44.5% relative to human-alone annotation. Our findings connect real-world human-centered video analysis tasks and current VLM capabilities, and clarify how human-AI collaboration can make VLM-assisted analysis reliable and efficient.

In this work, we propose a hybrid iterative deep Ritz method (H-IDRM) for a class of interface problems for second-order elliptic operators. It is based on a new mixed formulation of the problem and involves solving a sequence of convex minimization problems. We employ a level-set neural network architecture, featuring a level-set representation of the interface, to accommodate the piecewise smoothness of the solution and the flux. The approach involves only volumetric representations instead of duality pairing on the interface and avoids explicit interface sampling that is inconvenient for complex interface geometries. Further, we present an analysis of the method, including the errors arising from the neural network approximation, Monte Carlo approximation, iterative scheme, and penalty parameters. Numerical experiments indicate that the H-IDRM outperforms existing neural solvers on problems with high-dimensional domains, intricate interface geometries, and lower subdomain regularity.

Language-model agents increasingly face long-horizon tasks with evolving state, interdependent decisions, and delayed outcomes. Scaling their training requires diverse agentic environments, dependable outcome signals, and low extension cost. Existing generation pipelines commonly construct an environment before defining its outcome rule or annotating its trajectories, leaving dynamics and evaluation to be aligned post hoc. VHD-Play reverses this dependency by sampling and solving a mathematical model before a corpus-grounded setter renders its decision process as stateful tools. The executable dynamics and trajectory-scoring reference are inherited from the same solved model. The pipeline produces 3,300 diverse agentic environments at a cost of a few cents each. Training Qwen3.6-35B-A3B on three families raises its mean agentic score from 0.204 to 0.815 in a five-family diagnostic. Gains also appear on held-out instances from all three training families and eight unseen mechanism families, then extend beyond the generated substrate to external benchmarks for general function calling, travel planning, and 365-day e-commerce. On E-Commerce Bench, the trained checkpoint completes every run without bankruptcy and exceeds Qwen3.7-Max. We compare written-out problems with stateful versions that reveal or hide their parameters. The comparison shows that most of the learnable gap lies in stateful interaction rather than underlying problem solving. A frozen 35B setter realizes larger environments, and scale-matched training retains gains as mechanism size and horizon grow, indicating the potential for an evolving training substrate.

The recently proposed XOR cache is an inter-line-compressed last-level cache (LLC) that leverages the data-inclusion relationship between the private caches and the LLC, compressing two cache lines into one by XORing them. The architecture relies on the cache coherence protocol for data decompression. In this paper, we demonstrate that this mechanism - specifically the latency asymmetry between a cache hit on an uncompressed vs. compressed line - introduces microarchitectural vulnerabilities. Based on this observation, we propose a covert channel attack targeting the XOR cache. A colluding sender controls the receiver's access latency by triggering decompression through targeted write requests to partner cache lines. By exploiting the data-dependent compression behavior of the XOR cache, the sender and receiver establish the channel using pre-agreed data values. The channel achieves higher bandwidth than the Prime+Probe baseline for two reasons: first, each bit is encoded in the compression state of an individual line rather than the occupancy of a cache set, so a single set carries multiple bits; second, each bit is resolved by manipulating coherence-protocol state rather than forcing shared-cache evictions, so it costs fewer LLC accesses and demand misses than Prime+Probe. Full-system simulations show a bandwidth of 2.9 Mbps at an observed 0.98% bit-error rate (BER) over 50,000 transmitted bits, 13.1 times the bandwidth of Prime+Probe under the same sub-1%-BER selection rule.

Infrared (IR) imaging is crucial for autonomous driving, remote sensing, and other perception tasks. However, adverse weather may introduce fake structural responses that are entangled with real thermal structures. Existing IR restoration methods are typically designed for a single degradation type or directly reconstruct from degradation-entangled representations. Consequently, they struggle to distinguish intrinsic thermal structures from weather-induced fake responses and to accommodate spatially varying degradation severity, leading to artifacts or the over-suppression of weak but meaningful thermal responses. To address these issues, we propose TSGPD-IR, a type-severity guided progressive disentanglement network for all-in-one infrared restoration that factorizes restoration guidance into task-level weather semantics and region-level degradation severity. Specifically, a Weather and Semantic Co-Guided Multi-Level Prompt Generation Module combines global weather semantics with stage-wise local features to generate adaptive prompts that progressively suppress degradation-induced responses while preserving intrinsic thermal structures. To complement global weather semantics with spatial restoration control, a Proxy-Supervised Regional Degradation Estimator derives severity supervision without manual annotations and predicts spatially varying degradation priors. Guided by these cues, a Multi-Source Collaborative Expert Selection Strategy uses a shared branch to preserve weather-invariant thermal structures and hierarchical routing to select weather-specific expert pools and severity-compatible regional experts. This design progressively separates degradation interference from genuine thermal content and enables region-adaptive restoration, reducing both residual artifacts and over-suppression.

Comparing current and past observations helps robots understand scene changes and select subsequent actions during manipulation. However, comparing visual features at the same image location can mix different scene content when objects or the camera move. We introduce CoRe-WAM, a world-action model that incorporates correspondence-aligned visual changes through a parameter-efficient temporal interface. Its TraceDelta module uses correspondences from a frozen tracking model to transport historical visual features to current locations before computing signed differences in a shared pretrained feature space. Correspondence thus determines which historical content is compared with the present, rather than entering the policy as a separate trajectory representation. A lightweight adapter converts these differences into validity-gated residuals that supplement current visual conditioning, allowing the policy to use recent changes alongside current-scene information. Built on Motus, CoRe-WAM keeps the pretrained backbone weights frozen and optimizes 1.59 million parameters. With a 5,000-update adaptation budget, CoRe-WAM achieves 92.22% clean success across 50 RoboTwin 2.0 tasks, 3.56 percentage points above Motus; on randomized evaluation, it achieves 89.60% success, a 2.58-point gain. Integrating TraceDelta into a StarVLA-based policy improves clean success from 58.10% to 67.62%, supporting transfer of the temporal interface beyond Motus.

Robots deployed for competitive tasks must outmaneuver their opponents without sacrificing safety. Existing approaches, including safe reinforcement learning (RL), train a single policy to achieve task success and avoid failures simultaneously. This coupling can complicate training and leave the learned policy exploitable by deliberate attacks. We propose Safety to Competence (S2C), a two-stage RL framework that separates safety synthesis from competitive task learning. We formulate competitive interactions as safety-critical Markov games and prove that perfect filtering preserves policy non-exploitability when all players commit to safe maneuvers. S2C learns a robust safety filter via adversarial RL, embeds it in the environment during task policy training, and retains the same filter at deployment. In simulated touchdown games, S2C outperforms eight safe RL baselines, achieving the highest win rate and Elo rating, and the lowest exploitability. Hardware stress tests against a human opponent confirm S2C's competence.

Command-and-control (C2) traffic increasingly hides within TLS, so defenders now apply machine learning to traffic metadata. Many studies assume that combining two metadata views, namely flow statistics and TLS handshake fingerprints, improves both accuracy and robustness. We tested this assumption on 17,577 TLS flows from 62 real Cobalt Strike captures. Our evaluation removes the data leakage that leads to overly optimistic reported scores. We report three findings that matter more than the fusion result itself. First, an incorrect preprocessing step increases the F1 score by 0.28. This step computes the frequency encoding across the entire dataset rather than within each cross-validation fold. The increase is about ten times larger than any real effect we measured. Second, both the labels and the behavioral features depend on the destination address. Because of this, the 17,577 flows form only 2,132 independent groups, and the positive rate of 55.1%, which looks balanced, drops to 4.2%. Therefore, class balance is just a result of how we analyze the data, specifically whether we count flows or endpoints, and not a real feature of the task. Third, 20 of the 62 captures (32%) have no TLS flows to any known C2 address, so they contain only benign samples. We checked these captures directly and confirmed that this is a gap in the ground truth, not a labeling error. In this context, fusion beats the best single view by only 0.022 in F1. When an attacker forges both feature surfaces simultaneously, every model performs worse than a simple baseline that always predicts positive (F1 = 0.711). For encrypted C2 detection, the evaluation design is not a preliminary step. It is the main result.