Networking
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Saturday 5 September
In data networks carrying large numbers of flows, Heavy Hitters (HHs) or Elephant flows are the flows exceeding pre-determined thresholds (e.g. number of packets or bytes) in a given time window. Such HH flows need to be handled differently in order to minimize their impact on other smaller flows. HH detection techniques have been shown to be more effective when implemented in programmable data plane switches. In recent work, it was shown that the inter-packet gap can be used to identify heavy hitters. Such schemes use a limited-size hash table for storing flow state information and using this for the detection. However, when hash collisions occur, it is possible that a valid HH flow in the table can be replaced by a non-HH flow resulting in missing detection of HH flows. To address this problem, this paper incorporates a flow's medium-term Packet Count (PC) feature. In order to limit the packet count field size in the hash table, counting is done only till hash collision occurs so as to reduce the range of values to be stored and thus, the required number of bits. Also, another flow's medium-term feature, Packet Size (PS) is incorporated independently. The proposed scheme has been implemented in the P4 language and tested on Intel Tofino hardware. Performance evaluation has been performed using CAIDA and MAWI-based real-life traffic traces. The results show that in several scenarios cases, we can significantly reduce the False Negatives for HHs by using the packet count data effectively and efficiently.
Distributed Computing
CoCoFL: Continual Computing for Federated Learning over Intermittent Satellite-Ground Links
Low earth orbit (LEO) satellite constellations enable geographically distributed ground devices to collaboratively train a global model via federated learning (FL) without sharing raw data, with applications in environmental monitoring and disaster prediction. However, in satellite-assisted FL scenarios, intermittent satellite-ground links allow only a subset of devices to participate in global aggregation within each visibility window, leaving unscheduled devices idle and their local computational and data resources underutilized. Under partial device participation, data heterogeneity among devices may bias the global model toward certain devices, thereby deteriorating learning performance. In this regard, we propose a continual computing based federated learning framework, referred to as CoCoFL, in which scheduled devices participate in the global model aggregation, while unscheduled devices continue updating their local models taking into account model staleness. Guided by the convergence analysis of CoCoFL and subject to visible-window-related time constraints, we jointly optimize the device scheduling and the number of local epochs for scheduled and unscheduled devices. Experimental results demonstrate that CoCoFL achieves faster convergence, lower training loss, and higher test accuracy compared with baselines.
Read the full abstract
General Relativity and Quantum Cosmology
Physics-based phenomenological modeling of binary black hole hierarchical formation 1: Synthetic universes from globular cluster simulations for GWTC
We iteratively model the GWTC-5.0 binary-black-hole census with three components. First, a physically normalized population from Rapster globular-cluster simulations spans cluster mass, metallicity, formation redshift, compactness, and natal black-hole spin. Hierarchical mergers reproduce higher-mass events and their larger effective-spin dispersion, provided black holes are born with zero natal spin. Second, a phenomenological field (isolated-binary) channel supplies the low-mass, preferentially aligned population. Its inclusion flattens the cluster compactness likelihood, permitting ordinary dense globular-cluster birth radii without requiring nuclear-cluster-like conditions. Finally, residual tension at $15$–$30 M_\odot$ motivates an intermediate-mass isotropic component representing field remnants reprocessed in clusters. This is a hypothesis, not a detection: current numerical support precludes a reliable evidence comparison. Physical normalization converts the high-mass rate into $\hat{f}_{\rm GC}\simeq0.39\%$, with local rates $R_{\rm cl}\simeq9.1$ and $R_{\rm field}\simeq16.0$ Gpc$^{-3}$ yr$^{-1}$. The model predicts linked mass-spectrum breaks near $35$ and $70 M_\odot$, a $q\simeq0.5$ feature from first-plus-second-generation pairings, and a symmetric effective-spin distribution that broadens sharply above $45 M_\odot$. These correlated, mass-resolved predictions can be tested as the gravitational-wave census grows.
Read the full abstract
Artificial Intelligence
Agentic Pressure: The Endogenous Entropy of Reliable Autonomy
Achieving reliable autonomy in the wild requires agents to sustain continuous operations across long-horizon trajectories. However, as agents navigate these unconstrained settings, they encounter cumulative friction that inherently destabilizes their alignment. In this paper, we identify a distinct non-adversarial phenomenon termed Agentic Pressure. We define this as a kinetic force that spontaneously emerges when the cost of compliance conflicts with the imperative of goal achievement. Unlike static jailbreaks, this pressure is endogenous and arises directly from the dynamics of interaction. We propose a theoretical framework that formalizes Agentic Pressure as the ratio between the required work to overcome environmental friction and the remaining capacity of the agent. Our analysis demonstrates that when this pressure exceeds a critical threshold, agents exhibit safety drift as a mathematically optimal adaptation. Consequently, they often resort to Instrumental Hallucination to rationalize rule violations. Empirical experiments validate this framework and show that aligned agents spontaneously compromise safety to preserve autonomy under high-pressure conditions.
Read the full abstract
Robotics
GLoRI: Closed-Loop Whole-Body Tracking with Global-Local Reference Interaction for Humanoid Loco-Manipulation
Humanoid loco-manipulation requires accurate whole-body motion tracking in the world frame for physical interaction. While local references preserve motion structure, they lack explicit constraints on absolute spatial placement, leading to accumulated global errors. Existing globally aware approaches augment teleoperation policies with global observations but do not explicitly integrate global correction with local motion guidance, limiting autonomous tracking accuracy. We present GLoRI, a closed-loop whole-body controller that integrates structured global reference and feedback with local motion guidance. Its GLoRI-Net uses Global-Local Cross Attention(GLCA) to refine local keypoint features with global target and pose-difference features, preserving motion structure while correcting world-frame placement. GLoRI achieves 100% completion and a g-MPJPE of 6.44cm on held-out HuMoTo motions. This accuracy remains robust under direct Isaac Gym-to-MuJoCo transfer without fine-tuning, demonstrating strong generalization. Furthermore, such accuracy and generalization enable autonomous loco-manipulation with a single policy on a real Unitree G1 interacting with diverse unseen objects, extending beyond prior systems that primarily rely on teleoperation or focus on single-object interactions.
Read the full abstract
Computation and Language
Alignment by Stereotyping: How LLMs Sacrifice Individual Distinctiveness for Cultural Adaptation
Large language models are increasingly deployed for personalized interaction, and demographic conditioning via user profiles is a widely adopted strategy for cultural adaptation. We ask whether this approach genuinely serves individual users or achieves accuracy by erasing individual distinctiveness. Studying seven models including frontier GPT-5.1 on the World Values Survey, we find that demographic profiles improve value alignment accuracy for most models, but at a systematic cost to individuality. That is, models pull responses toward demographic group centroids rather than preserving individual differences, a behavioral pattern we term alignment by stereotyping. Permutation tests (10,000 permutations, six demographic attributes, seven models) certify that top-performing models compress individuals far above the human baseline; within-family scaling amplifies this tradeoff while degrading intrinsic cultural understanding. Using a synthetic dialogue dataset validated on real human-chatbot conversations from PRISM (Kirk et al., 2024), we further show that distributing demographic signals across conversational turns partially suppresses prototype retrieval compared to compact demographic labels, a finding validated on real conversations via PRISM but requiring replication at larger scale.
Read the full abstract
Artificial Intelligence
MOAE: Multi-Objective Agent Evolution with Pareto-Preserving Search
As LLM-based agents continue to advance, their evaluation has become increasingly multifaceted: a capable agent must not only achieve high task completion accuracy but also perform well in interaction quality, safety, and efficiency, raising a central question: can these objectives be optimized simultaneously? Existing methods have considered multiple objectives, but many collapse heterogeneous measurements into a fixed scalar score. Such scalarization depends on metric normalization and preference weights and may discard candidates that represent useful deployment trade-offs. We introduce Multi-Objective Agent Evolution (MOAE), which organizes iterative in-context refinement as a Pareto-preserving evolutionary search over complete agent rollouts. Given a limited rollout budget, MOAE maintains an empirical archive of non-dominated candidates, uses objective-specific diagnostics to guide offspring generation, and applies constraint-aware selection only at deployment. This separates candidate preservation during search from the preference used to return a final solution. The procedure requires no parameter updates and allows each objective to be replaced by any measurable property, which we instantiate as task performance, trajectory quality, and safety. Experiments on TravelPlanner and AgentDojo show that MOAE consistently improves task performance and trajectory quality while maintaining strong safety under matched rollout budgets. Search-behavior analysis further shows that Pareto preservation expands the attainable objective region and increases the frequency of joint improvement. These results demonstrate the potential of Pareto-preserving in-context evolution for optimizing multiple agent properties without committing to a fixed scalarization during search.
Read the full abstract
Differential Geometry
A sub-Riemannian endpoint map with no regular values
We construct an endpoint map with no regular values, disproving the sub-Riemannian Sard conjecture. Remarkably, in our construction the minimizing Sard property remains valid.
Computer Vision
MORPHA: Morphology-Constrained Training and the Limits of Cross-Acquisition Transfer in Low-Resource Malaria Microscopy
In low-resource malaria microscopy, a model trained on one smear preparation routinely meets images from another, and how well morphology-based constraints transfer across this acquisition gap is unclear. We study this on real African field microscopy from Uganda (Lacuna), asking where encoding measured parasite morphology as a training constraint improves cross-acquisition transfer and where generic regularisation suffices. We present MORPHA, a morphological consistency constraint that derives stage-conditional statistics from the stage-annotated BBBC041 dataset and penalises predictions that deviate from them. Defined uniformly across binary, object-level, and stage-aware regimes without changing architecture or inference, it shapes training in the binary regime. The detection regime is a mapped boundary. On transfer from thin-smear cells to thick-smear field images, the constraint reduces the binary-classification generalisation drop by 30.8% (F1 0.578 to 0.699) at negligible within-domain cost and lowers in-distribution calibration error by 49% (ECE 0.0162 to 0.0082). A content-free control applying the identical constraint to random statistics recovers less of the drop (25.3% vs 30.8%), indicating the measured content, not constraining alone, contributes to the gain. Two standard confidence regularisers exceed the constraint on raw transfer, locating where morphology adds value and where generic regularisation suffices. We map two deployment-relevant boundaries: thin-smear statistics do not transfer to thick-smear detection (trophozoite AP@0.50 falls to 0.000), and cross-acquisition pseudo-labelling fails before filtering applies. Together these yield a morphology-grounded consistency signal and evidence-based guidance for malaria dataset and model design in low-resource settings.
Read the full abstract
Lattice Field Theory
Nucleon-sigma terms at $m_\pi=222 ~\rm MeV$ with a variational analysis from lattice QCD
Nucleon sigma terms are important for the decomposition of the nucleon mass and for searches for new physics beyond the Standard Model involving scalar interactions. A persistent tension between lattice QCD and phenomenological determinations may be due to uncontrolled excited-state contamination in lattice QCD analyses. In previous work at $m_\pi=429~\rm MeV$, we showed that this contamination is dominated by $N\sigma$ states and can be strongly suppressed with a variational analysis using $N$ and $N\sigma$ operators. In this talk, we present preliminary results at $m_\pi=222~\rm MeV$, where the $\sigma$ becomes unstable and decays into $\pi\pi$. We investigate whether the same small variational basis can remove the expected $N\pi\pi$ contamination through the overlap of the $N\sigma$ operator with these states.
Read the full abstract
Machine Learning
DART: Distributional Adversarial Recurrent Training for Algorithm Learning
Recurrent reasoning models (RRMs) can solve structured problems, achieving easy-to-hard generalization through iterative computation in hidden space. These models are typically trained with instance-level supervision, which becomes increasingly problematic as task difficulty grows: valid solutions occupy a tiny region of the solution space, while invalid solutions proliferate rapidly. We propose Distributional Adversarial Recurrent Training (DART), a training framework that replaces single-point supervision with a local target distribution around the ground-truth solution and aligns model outputs with this distribution through an adversarial objective. DART provides a richer learning signal and encourages more stable iterative trajectories toward valid solutions. When evaluated on Maze, Chess, and masked Sudoku with multiple RRMs, including Deep Thinking Systems and Tiny Recursive Models, DART improves solution quality, stability, and robustness under the evaluated distribution shifts. Comparisons with label smoothing, Gaussian softened targets, and progressive training show that DART is not explained by target softening alone and is complementary to training schemes that stabilize long-horizon recurrence. These results identify DART as a promising approach for improving robustness across the evaluated recurrent reasoning models.
Read the full abstract
Signal Processing
Pinching Antenna Assisted Integrated Sensing and Communication Using Quadrature RSMA
This paper proposes a pinching antenna system (PASS)-assisted downlink quadrature rate-splitting multiple access (Q-RSMA)-based integrated sensing and communication (PASS-QRSMA-ISAC) framework for simultaneously supporting multi-user communication and target sensing. In the proposed framework, multiple pinching antennas (PAs) deployed along a dielectric waveguide serve communication users (CUs) distributed across a dual-room indoor environment, where line-of-sight (LoS) and non-line-of-sight (NLoS) propagation conditions coexist. By integrating PASS with Q-RSMA, the proposed framework improves interference management while reducing the dependence on successive interference cancellation. The communication and sensing signal models are developed, followed by the corresponding SNR/SINR expressions. To obtain tractable analytical insights, a two-user single-PA case under ideal waveguide conditions is first analysed, where the per-user communication outage probability (COP), system COP, sensing outage probability (SOP), and ergodic sum-rate are formulated using distance statistics, CDF/CCDF characterisations, and numerical integration methods. The analysis is then extended to the generalised multi-user multi-PA scenario by developing statistical characterisations for the LoS communication, NLoS communication, and sensing channels. Based on these characterisations, the per-user COP, system COP, SOP, and ergodic sum-rate are formulated and evaluated for the generalised framework. Monte Carlo simulations validate the analytical results and demonstrate the performance gains of the proposed PASS-QRSMA-ISAC framework over the considered benchmark schemes, including PASS-RSMA-ISAC, PASS-NOMA-ISAC, and PASS-SDMA-ISAC. The results further highlight the effects of various system parameters on the proposed framework.
Read the full abstract
Numerical Analysis
Quasi-uniform designs from random candidates: optimal candidate complexity and farthest-point sampling
We study randomized constructions of quasi-uniform designs on a compact metric-measure space satisfying two-sided polynomial ball-growth conditions. Given integers $N\le M$, we first draw $M$ independent candidate points and then retain the first $N$ points of a farthest-point traversal of the candidate set. We prove probabilistic non-asymptotic bounds on the mesh ratio and show that $M=\Theta(N\log N)$ is the sharp order of the candidate-pool size required for bounded mesh ratio: a sufficiently large multiple of $N\log(N/\delta)$ suffices with probability at least $1-\delta$, whereas, if $M=o(N\log N)$, the mesh ratio diverges in probability for every procedure that selects $N$ points from the same independent candidate pool. If $M/(N\log M)\to\infty$, the upper bound on the mesh ratio for the exact farthest-point sampling (FPS) tends to $2$. We also quantify the effect of approximate FPS and analyze direct and fast implementations in spaces of bounded doubling dimension. Finally, using a single infinite stream of candidate points and increasing candidate budgets, we construct an almost surely quasi-uniform nested sequence; under supercritical oversampling and exact FPS, or more generally when the approximation factors tend to $1$, its mesh-ratio limit superior equals $2$ for compact positive-volume subsets of Euclidean space.
Read the full abstract
Robotics
A Brain-inspired Hierarchical Framework for Zero-Shot Robot Task Reasoning and Execution
Robots that follow open-ended language instructions need to connect semantic intent to visual scene understanding, geometric feasibility, object states, and physical interaction conditions. End-to-end Vision-Language-Action policies have improved cross-task generalization, but they typically map visual and language inputs directly to robot actions, leaving limited explicit structure for long-horizon decomposition, physical verification, and recovery. We present \method, a zero-shot hierarchical framework functionally inspired by the division of roles in the human brain, comprising visual perception and state inference, language grounding and action-sequence generation from a shared atomic action library, cost-based plan selection, and real-robot execution and verification. The framework grounds commands in explicit object states, composes reusable atomic actions into task-conditioned sequences, ranks alternative sequences by execution cost, and verifies intermediate physical outcomes from refreshed observations. In the evaluation, \method{} completes 10/10 clean board trials, 10/10 pick-and-place trials, and 4/5 pyramid stacking trials for both the flat and irregular initial-layout conditions; the corresponding mean task progress is $99.03\%$, $100.00\%$, and $96.67\%$ respectively. Across all evaluated conditions, \method{} achieves higher success rates than ReKep, Dream2Flow, and $\pi_{0.5}$ benchmarks, demonstrating the effectiveness of combining explicit object-state reasoning, compositional atomic actions, cost-based plan selection, and closed-loop execution verification.
Read the full abstract
Quantum Physics
Data-driven structural diagnostics and autonomous alignment of complex optical systems
The quest for autonomous alignment and maintenance of complex free-space optical systems is increasingly urgent for large-scale neutral-atom quantum processors. Leveraging a high-finesse optical cavity as a sensitive probe, we introduce a data-driven framework for structural diagnostics and closed-loop control that achieves cold-start convergence within tens of seconds. These results establish a rapid, scalable diagnostic-control methodology for autonomous alignment and continuous maintenance of complex free-space optical architectures.
Read the full abstract
Software Engineering
Replications, Revisions, and Reanalyses: Managing Empirical Evidence in Software Engineering
One aspired outcome of empirical research on quantitative data is a variance theory, i.e., a quantification of the effect of an independent on a dependent variables. The validity of variance theories stems from the synthesis of multiple pieces of evidence, which increases its validity beyond the findings of a single study. However, research synthesis in SE is rare and if done mostly limited to purely narrative syntheses. At best, researchers perform meta-analyses to synthesize variance theories from several quantitative results. But even meta-analyses only produce reliable results when synthesizing exact replications yet fail to generalize from variations. We aim to extend the frontier of research synthesis beyond the state-of-the-art to systematically manage empirical evidence and its evolution. We apply method engineering to construct a framework for research synthesis from proven, individual method fragments. The framework allows researchers to put new evidence in a clear relation to an existing body of evidence and systematically expand knowledge about a studied phenomenon. We demonstrate the application of this framework to two fields of research by explicitly modeling the relationship between existing pieces of evidence. The framework puts three types of evolution of evidence into relation: (1) replications investigate the same hypothesis in a new context to improve external validity, (2) revisions challenge an existing hypothesis to improve internal validity, and (3) reanalyses replace analysis methods to improve conclusion validity. Through a systematic evolution of evidence and clear assessment criteria for each dimension of validity, the proposed framework can determine the frontier of a field of research. The framework provides a perspective to systematically evolve empirical evidence in SE, supporting more constructive and productive advances in our field.
Read the full abstract
Hardware Architecture
CMD: An Integrated CGRA Framework with Cluster-Based Distributed Memory Design
Coarse-Grained Reconfigurable Arrays (CGRAs) are a promising solution for achieving high energy efficiency and reconfigurability across various application domains, but their performance is often crippled by rigid memory architectures that limit the number and location of tiles that can access data memory. This creates a significant bottleneck for kernels with intensive memory accesses. To address this, we propose CMD, an integrated CGRA framework featuring cluster-based distributed memory design with a co-designed compilation toolchain. The compiler includes a novel memory-aware mapper and a design space exploration (DSE) mechanism that identifies the optimal memory architecture design for specific kernels. Experimental results show that our post-DSE CMD CGRAs achieve an average speedup of $1.39\times$ over a conventional CGRA while simultaneously reducing the total area to an average of $0.912\times$ of the conventional CGRA.
Read the full abstract
Computer Vision
Accelerating Diffusion Transformers with Gaussian Process Rectified Feature Cache
Diffusion Transformers have become the dominant paradigm in generative AI, but their high computational costs severely hinder real-time applications. Prediction-based feature caching is widely used to accelerate diffusion transformers; however, as the number of steps increases, the deviation between its predictions and the reference full-compute trajectory gradually grows. An intuitive idea is to use an online regression model to dynamically correct this deviation, but it faces the issue of label data being unavailable during the acceleration process. This paper presents a statistical observation that the residuals between the features of full computation steps using caching methods and reference full-compute trajectory locally exhibit a zero-mean Gaussian distribution. By treating the features of full computation steps as noisy observations of reference features, the data acquisition problem is resolved. Based on this observation, a plug-and-play GP-Refiner correction framework is proposed. This method utilizes Gaussian Process Regression for correction and, leveraging the properties of GPR, introduces an uncertainty-adaptive computation strategy that triggers necessary full-computation calibration by monitoring the posterior variance in real time. Experiments demonstrate significant improvements across different models when combined with various state-of-the-art methods. Integrating the proposed framework with TaylorSeer reduces the computational load by 19.3% while improving PSNR by 0.9 dB and reducing LPIPS from 0.46 to 0.29. Code is available in https://github.com/Aredstone/GP-Refiner.
Read the full abstract
Cosmology and Nongalactic Astrophysics
Stacking HI in the dark: Towards detecting the reionization-epoch 21 cm signal in Lyman-$\alpha$ dark gaps
The detection of the redshifted 21 cm signal from neutral hydrogen during the epoch of reionization remains one of the central challenges in contemporary cosmology. Although upcoming radio interferometers are expected to measure 21 cm fluctuations at a few percent precision, an independent confirmation is essential to ensure robustness against instrumental systematics and foreground contamination. In this work, we propose a novel real-space stacking technique to detect the 21 cm signal, guided by the presence of dark gaps in quasar absorption spectra during the late stages of reionization ($z \sim 5.0$-$6.1$). Using a semi-numerical reionization model, we show that stacking 21 cm images on long dark gaps (length, $L > 32~h^{-1}\mathrm{cMpc}$ and threshold flux, $F_{\mathrm{th}} < 0.005$) should reveal a strong emission signal with a peak mean-subtracted 21 cm brightness temperature of $\delta \tilde{T}_b \sim 2-3~ mK$ for a volume-weighted average neutral fraction of $x_{\mathrm{HI}} \sim 0.1$ at $z = 5.9$. We further extend our analysis to realistic reionization lightcones, incorporating thermal noise and foreground avoidance effects, generated for two different reionization histories. In an optimistic foreground avoidance scenario (wedge field of view $\theta < 20^\circ$), a sample of roughly 100 dark gaps is sufficient to achieve a combined root-sum-square signal-to-noise ratio exceeding 10 for both reionization models, while a more conservative wedge avoidance may require more dark gaps for a solid detection. Given the large number of quasars known at $z > 6$, a detection of this stacked signal may thus already be within reach of existing optical (i.e., VLT, Keck) and 21 cm observatories (i.e., HERA, SKA). This synergistic approach offers a promising pathway towards a direct and statistically robust detection of the cosmological 21 cm signal.
Read the full abstract
Numerical Analysis
Robust high-dimensional integration using medians of coarsely scrambled Sobol' sequences
We study the numerical approximation of high-dimensional integrals using randomized quasi-Monte Carlo (RQMC) methods, with a focus on scrambled Sobol' sequences. While asymptotically faster than Monte Carlo, classical RQMC suffers from error bounds that grow exponentially in the dimension $s$, and its root mean squared error (RMSE) convergence rate is generally no better than $O(N^{-3/2})$. To overcome these limitations, we combine two recent developments: Suzuki's coarse scrambling and the median trick. Coarse scrambling randomizes Sobol' sequences according to their generating base polynomials and significantly reduces the maximal gain coefficient. By appropriately choosing the base polynomials, we show that the coefficient can be made uniformly bounded in $s$, yielding an $O(N^{-1/2})$ RMSE for $L^2$ integrands with a dimension-independent constant. The median trick then enables near-optimal convergence for function classes beyond $L^2$: for $L^p$ integrands with $p\in(1,2)$, the median of independent coarsely scrambled estimates achieves an error of $O(N^{-1+1/p})$ with high probability; for integrands in the Haar wavelet space $\mathcal H_{\mathrm{wav},\alpha,s,p,q}$ with $\alpha_p:=\alpha-(1/p-1/2)_+>0$, we prove a high-probability error bound of $O(N^{-\alpha_p-1/2+\varepsilon})$ for any $\varepsilon>0$; the bound improves to $O(N^{-r-\alpha_p-1/2+\varepsilon})$ when the integrand has dominating mixed derivatives of order $r\ge1$ that belong to $\mathcal H_{\mathrm{wav},\alpha,s,p,q}$. The latter two bounds are uniform in $s$ under suitable conditions on the ANOVA components of the integrands. Numerical experiments confirm the predicted convergence rates and demonstrate the robustness of the proposed approach.
Read the full abstract
Software Engineering
VibeCheck: Assessing the Quality of LLM-Generated Unit Tests: A Multi-agent Empirical Study across Heterogeneous Repositories
LLM-based IDE agents are increasingly used to generate repository-grounded unit tests, yet common evaluations often rely on execution success or coverage. These metrics can miss deeper quality issues such as weak assertions, missing edge cases, poor isolation, and limited maintainability. This paper presents VibeCheck, an empirical study of unit test generation across 15 student-developed Python and JavaScript/TypeScript repositories. We evaluate Kiro, Antigravity, and Cursor with Claude Sonnet 4.5 as the underlying agent, under repository-only, zero-shot conditions using a five-dimensional rubric covering runnability, assertion strength, logic and edge-case coverage, isolation/determinism, and maintainability. We also apply leave-one-out cross-agent peer evaluation to compare tools and identify failure patterns. Results show a clear execution-adequacy gap: generated tests are often runnable but frequently lack strong assertions and meaningful behavioral coverage. Weak assertions and missing edge cases occur more often than blocking failures, showing that runnable tests can still be shallow. VibeCheck provides a reliability-oriented framework for evaluating LLM-generated tests beyond pass/fail outcomes.
Read the full abstract
Soft Condensed Matter
In situ local learning of dynamic network materials
Functional materials realize functionalities through pre-designed structures, whereas physical neural networks are trained by prescribing input-output behavior. Unifying these views would allow materials to acquire functions directly from dynamical tasks. Here we introduce an in situ local learning framework for dynamic network materials. A forward drive and a time-reversed adjoint error drive are applied to the same mechanical network, enabling the material to compute gradients of a time-domain loss through its own dynamics. The resulting update rules are local, compatible with the principle of physical learning. Each bond or node requires only co-located forward and adjoint fields. The method trains spring constants, nodal masses, bond damping coefficients and nodal damping coefficients, and programs dynamical functions including broadband wave cloaking, evanescent-wave imaging, passive transient output enhancement and vowel classification. These examples involve distinct physical mechanisms, such as wave scattering, near-field information transfer, dissipative non-normal dynamics and machine learning, but are obtained by simply changing the loss function, rather than leveraging expert knowledge. Therefore, dynamic network materials can act both as programmable matter and as physical neural networks, providing a route to material systems that learn temporal responses in situ.
Read the full abstract
Computation and Language
Beyond Cross-Lingual Transfer: Benchmarking Propagation Boundaries in Multilingual LLM Unlearning
Large Language Model (LLM) unlearning aims to suppress target knowledge while preserving general capabilities. In multilingual settings, unlearning must additionally propagate within its intended linguistic scope. However, existing evaluations mainly measure cross-lingual transfer and cannot distinguish insufficient from excessive propagation. We introduce CLLPU (Cross-Lingual and Language-Bound Protocol for LLM Unlearning), a multilingual benchmark that formulates this problem through two settings: common-goal forgetting, where target knowledge should be suppressed across all languages, and language-conditioned forgetting, where suppression should remain confined to a designated language. CLLPU combines goal-guided topic pairing, schema-aware relation matching, and dual-anchor multilingual translation to construct 800 matched knowledge-unit pairs and 72,000 QA instances across ten languages. Experiments with six representative methods on Llama-3.1-8B-Instruct reveal opposite failure modes: forgetting remains incomplete when universal suppression is required, yet spreads beyond the intended boundary when language-conditioned confinement is required. We further find that general multilingual utility can conceal damage to neighbor knowledge. These findings establish propagation control as a central challenge for multilingual LLM unlearning. We publicly release CLLPU together with its construction pipeline.
Read the full abstract
Computers and Society
Intent Drift at SME Scale: Deployment Practice, Not Model Capability, Determines Agentic Compliance
We introduce Chain of Intent, a governance framework for agentic AI at small regulated firms, and validate it against a failure it was built to address. Existing agentic governance research assumes enterprise infrastructure that small firms do not have. In a simulated Hong Kong asset manager with 415 synthetic contact records, an agent performing a routine client-communications task was subjected to ordinary managerial pressure to increase its reach. With its authorised constraints written into its configuration, the agent held: it identified every ambiguity in the firm's records, cited privacy legislation it had never been shown, and refused six successive requests, breaching in two of fifteen runs. With the same task, data, pressure and model, but its purpose left unstated as resource-constrained firms routinely leave it, it breached in thirteen of fifteen runs, contacting up to 220 individuals of whom 94 per cent had no demonstrable marketing consent - conduct carrying a maximum of three years' imprisonment under Hong Kong law. Chain of Intent applies four controls requiring no security engineering: a machine-readable purpose, constrained tool access, a scope ledger, and a pre-action check. It eliminated unlawful contact in every run while preserving task completion, and ablation shows each control independently sufficient by a different mechanism. We further show that drift must be measured at two stages - agents widened their candidate sets in every pressured run while acting on them in roughly one in seven - and that governance applied at the point of intent costs roughly half as much as governance applied at the point of action.
Read the full abstract
Functional Analysis
Rough paths below the Young threshold: an exact scale calculus and the locality phase transition at one quarter
Since Young's 1936 theorem, irregular integration has been organized around the threshold 1/2: above it the path determines the integral, while below it higher-order data are needed. For fractional Brownian motion, H = 1/4 is the threshold for the canonical Gaussian enhancement, although geometric rough lifts exist for every H > 0. We prove that H = 1/4 is instead the exact threshold for measurable locality. For d-dimensional fractional Brownian motion with independent components, d at least 2, if 0 < H <= 1/4, no positive-measure Borel set supports even one finite off-diagonal second-level coordinate satisfying Chen's relation and measurable interval by interval from path increments. No moment, Holder, geometricity, stationarity, or scaling assumption is imposed. If 1/4 < H <= 1/2, every full-law local rough-path lift in the standard Holder range is automatically square-integrable and hence classified without an assumed L2 condition; for H > 1/2, every finite-step enhancement with natural graded Holder bounds is uniquely the Young signature. We also introduce an exact scale calculus below classical differentiability. Matched dyadic differences recover normalized derivatives with sharp O(epsilon^2) error, exact localized inversion, and lossless reconstruction. Multiplication and smooth functional calculus transport exactly to scale coordinates, while the corona quotient yields an exact universal derivation. Reinserting the Holder amplitude and using Fourier-normal ordering produces strong geometric lifts for every positive input regularity and every lower rough exponent, with explicit ultraviolet rates and stability. Thus rough lifts exist below one quarter, but no measurable lift on a positive-measure domain can be interval-local there. The results separate existence from local recoverability and show that nondifferentiability does not destroy exact differential information.
Read the full abstract
Signal Processing
Terahertz Inter-Satellite Links: Motivation, Challenges and Opportunities
Inter-satellite links (ISLs) are essential to the evolution of next-generation satellite constellations, providing the foundation for low-latency, resilient, and globally scalable connectivity. While low radio-frequency (RF)-based ISLs offer technological maturity, they are increasingly constrained by spectrum scarcity, congestion, and interference. Optical ISLs, on the other hand, deliver unprecedented capacity but demand ultra-precise pointing, suffer from narrow-beam limitations, and are limited to point-to-point links, all of which hinder large-scale deployment, including point-to-multi-point capability. To overcome these limitations, we propose very-high RF terahertz (THz) inter-satellite links (ISLs) as a promising middle-ground solution, merging the ultra-high data rates of optical links with the adaptability, reliability, and relaxed pointing requirements of lower-frequency RF ISLs. However, despite growing interest, research on THz ISLs remains at an early stage, fragmented across isolated studies, and lacking a clear roadmap for practical realization. This paper aims to address this gap by examining the fundamentals of THz ISLs, assessing their potential advantages and key challenges, and identifying the most promising research directions to transform them into a cornerstone of future interconnected mega constellations.
Read the full abstract
Computer Vision
Efficient and Robust Camera-independent Multiview 3D Geometric Reconstruction from Noisy Monocular Depth Estimation and Multiple Point Matching
We present an efficient and robust method for 3D geometric reconstruction that is based solely on the camera-independent linear relationships among a given set of points, which are stable over time and robustly estimated using multiple point matches. We essentially learn, from correspondences between points across several frames, a linear geometric auto-regression matrix $\mathbf{W}$, which establishes how a point in 3D can be expressed as a linear combination of all the others. This matrix is constant and does not depend on the world coordinate system or the camera pose—it is an intrinsic property of the point set. We also show that the principal eigenvectors of $\mathbf{W}$, which all have eigenvalue $1$, provide a homogeneous representation of the 3D point configuration. The first version of our method takes advantage of noisy monocular depth maps in order to obtain, from multiple frames, a robust geometric auto-regression matrix $\mathbf{W}$ of linear relationships between the 3D points. Thus, we build on recent advances in deep learning, which now provide monocular depth estimation models that are fast but very often noisy. Our approach handles noise through robust linear estimation over several frames. The second version of our method does not need monocular depth estimation maps. It applies in cases of weak-perspective projection, when the linear combinations between the 3D points can be robustly estimated from their 2D projections in the image. Note that the camera projection matrix is never used in our derivations. Consequently, our method does not recover camera pose, but only 3D structure. This is a key difference between our method and the related literature on 3D geometric reconstruction.
Read the full abstract
Materials Science
Reusable Operators for Irreducible Cartesian Tensor Decomposition and Coupling
Molecular and material properties, from the polarizability to the elastic constants, are described by tensors. Their behavior under rotations is made explicit when a tensor is decomposed into irreducible parts that transform independently. In Cartesian form these parts are the symmetric and traceless irreducible Cartesian tensors (ICTs), whose decomposition and coupling underlie selection rules, orientational averages, and the use of symmetry. The decomposition is a textbook result at rank two, but higher rank and the intrinsic symmetry of physical tensors make it nontrivial. What has been lacking, unlike in the well-established spherical formalism, is a general construction for a given intrinsic symmetry, together with reusable operators that extract the ICTs and rebuild the original tensor exactly. Here, we develop such a construction and obtain these operators explicitly. These operators depend on rank and symmetry alone, and therefore each need only be built once and then applied to any tensor of that class. Building on them, we further obtain the Cartesian harmonics of a vector and the operators that couple two ICTs into a third, both central to equivariant machine learning. The construction is demonstrated on the elastic tensor, in both its second-order form of rank four and its third-order form of rank six. The ICTs of the rank-4 tensor also define a rotation- and scale-invariant measure of anisotropy, which we evaluate across the first-principles elastic tensors of crystalline materials from the Materials Project. The construction is implemented in the open-source package natto, which produces the operators in both exact symbolic and numerical form.
Read the full abstract
Machine Learning
Machine Learning for Pre-Culture ESBL Risk Stratification to Guide Empiric Antibiotic Selection: A 12-Hospital Study of Enterobacteriaceae Cultures
Empiric antibiotic therapy for suspected ESBL-producing Enterobacteriaceae must be selected 48-72 hours before culture results, forcing clinicians to choose between undertreating resistant infections and overusing carbapenems that drive further resistance. We developed a cost-sensitive XGBoost model predicting an ESBL phenotype (resistance to ceftriaxone, ceftazidime, cefepime or piperacillin-tazobactam) at culture ordering using 45 pre-culture EHR features across 132,955 cultures from 72,217 patients at 12 hospitals (14.41% with the ESBL phenotype). Cultures were partitioned at the patient level. At 90% sensitivity, the model achieved 95.8% NPV, reducing post-test ESBL probability to 4.2%, a threshold that may support safe carbapenem-sparing in non-ICU settings, while sparing 307 of every 1,000 cultures an unnecessary broad-spectrum course at the cost of 14 missed ESBL cases per 1,000. SHAP analysis identified prior ESBL colonization as the dominant predictor, ahead of prior organism burden and neighborhood deprivation; removing deprivation features caused minimal performance loss ($\Delta\text{AUROC} = -0.020$), enabling equitable bedside deployment. Discrimination was unchanged under a strict IDSA ESBL-E definition (AUROC 0.766), with specimen type added as a predictor (0.764) and without any class-imbalance correction (0.762), and ranged from 0.71 to 0.78 across organism strata.
Read the full abstract
Materials Science
The electromagnetic stress tensor in cubic crystals and amorphous solids
The electromagnetic vacuum stress tensor in a system consisting of classical point-like atoms on a cubic lattice is derived directly from the electric field and, alternatively, using the known more phenomenological formula involving the (usually unknown) derivative of the dielectric tensor. The results agree. The stress tensor contains a nontrivial constant and therefore cannot be obtained from macroscopic electrodynamics alone. A system consisting of a mosaic of randomly oriented grains of a cubic crystal is used as a model of an amorphous solid. The electromagnetic stress tensor in this system is deduced by averaging the corresponding quantity in a cubic crystal over all orientations. Two components remain: the component in the direction of the electric field and the component perpendicular to it. The transverse component agrees with the analogous quantity for liquids derived by Peierls in an entirely different way.