Papers

A complete presentation for the diagonal reduction algebra of $\mathfrak{gl}_n$ was found by Khoroshkin and Ogievetsky. In this paper we present a complete set of ordering relations for a reduction algebra $\mathcal{B}_n$ associated to $\mathfrak{gl}_n$ in a parabolic subalgebra of $\mathfrak{sp}_{2n}$. We also study its connection to the diagonal reduction algebra of $\mathfrak{sp}_{2n}$. We use braid group action by Zhelobenko automorphisms to reduce the computations required. Another important tool we employ is the ABRR equation for the extremal projector. As an application we prove that all homogeneous ordering relations for the diagonal reduction algebra of $\mathfrak{sp}_{2n}$ are obtained from the relations in $\mathcal{B}_n$ by applying type $C_n$ Zhelobenko automorphisms.

A face recognition model links two images of a person recorded on separate occasions when their embedding similarity exceeds an operating threshold. We consider making chosen identities unlinkable across separate occasions while the model remains in service for the rest of the population. Deleting their images and retraining does not achieve this, since the model recognises identities never observed in training. Therefore, the embedding space must be altered against these identities, the process of which we call open-set adversarial forgetting. We propose three loss functions, one that disperses an identity's embeddings from their centroid, and two that map each image onto its own near-orthogonal target, learnt with the classifier head or fixed in advance as an almost-orthonormal frame. Each is fine-tuned alongside the classification objective on a subset of each identity's images. We evaluate them against four methods from prior work in verification and identification, at two forget scales and three backbones. Every loss acting on the embedding geometry makes the forget identities nearly unidentifiable. The orthonormal frame alone achieves strong forgetting, which holds wherever an image of that subset enters the comparison and leaves distinct forget identities unlinkable. It also surpasses a concurrent unsupervised method at a higher retain rate.

Given an even-dimensional smooth manifold and a (co)(closed) bundle-valued $n$-form $a$ that "resembles" a scalar $(n,0)$-form, we may (or may not) recover a "$\mathbb{Z}_2$-(almost) complex structure" under certain assumptions. In the case of a $4$-manifold or a quaternion-Kähler manifold, we can recover a compatible $\mathbb{Z}_2$-complex structure on the whole manifold including the zero locus of $a$. In general, we can recover a $\mathbb{Z}_2$-(almost) complex structure wherever $a$ satisfies a certain nondegeneracy condition, but not necessarily on the whole manifold, with a counterexample identified. The above condition of "resembling" a scalar $(2,0)$-form is satisfied by any $\mathrm{U}(1)$-invariant solution to the Vafa-Witten equation on a compact Kähler surface. This hints at a possibility to generalise the notion of $\mathrm{U}(1)$-invariant solutions to oriented Riemannian $4$-manifolds. However, our results above imply any solution satisfying the condition, and thus seemingly any potential "generalised $\mathrm{U}(1)$-invariant solution", must be trivial if the $4$-manifold admits no compatible $\mathbb{Z}_2$-complex structure.

The equal-weighted portfolio is a passive, rule-based strategy that has historically been difficult to outperform, delivering higher returns than the capitalization-weighted "market" benchmark across many markets and periods. Stochastic portfolio theory (SPT) reveals that this relative performance is regime dependent, with the equal-weighted portfolio underperforming during periods of increasing market concentration and high correlations, particularly market bubbles. These observations have motivated us to formulate and solve a stochastic control problem in which an investor actively allocates between the equal-weighted and market portfolios. The investor bases their allocation decisions on forecasts made under a flexible stochastic diversity–dispersion (SDD) model. Using a quadratic surrogate for implementation frictions, we characterize the optimal allocation through a linear forward–backward SDE and obtain an explicit "aiming in front of a moving target" representation of the optimal trading rate, in the spirit of Gârleanu and Pedersen. The penalty parameters are calibrated in sample to match the cumulative wealth effect of proportional transaction costs, while out-of-sample performance is evaluated with those costs deducted directly from portfolio wealth. Using historical S&P 500 data, we show that a mean-reverting SDD specification reproduces several empirical features of market diversity and dispersion. In out-of-sample backtests from 1995 to 2024, the resulting strategies deliver higher cumulative net returns than both the equal-weighted and market portfolios, and higher information ratios than the equal-weighted portfolio after 15-basis-point proportional transaction costs.

We propose a Bayesian framework for uncertainty quantification from the perspective that the working model is mis-specified in settings of a multilevel data-generating process. We focus on settings in which the mis-specification fails to match the functional form of the mean structure, and discuss Bayesian estimation of target parameters under dependence induced by a mismatch between working and data-generating models. The proposal represents a Bayesian semi-parametric procedure aimed at estimating population-level parameters while accounting for cluster- and unit-level variation in the estimating function. The proposal extends the regular Bayesian bootstrap to account for cluster- and unit-level variation using multilevel weights from an enriched Dirichlet model. Simulation studies indicate that the proposed approach has good frequentist properties when the data-generating process and the proposed model induce a partially exchangeable sequence associated with the unknown quantity of interest. Applications to radon (Gelman and Hill, 2007), Programme for International Student Assessment 2022 (OECD, 2023), and tuberculosis (Nobre et al., 2023) datasets are presented for illustrative purposes. The results demonstrate that the proposed method is competitive with variations of multilevel models, with major differences observed in the range of credible intervals, which are justified by the nonparametric assumptions underlying the proposed method.

The relation between HI and dark matter (DM) halo masses provides key insights into gas accretion and the regulation of galaxy formation. In a previous study based on SPARC and LITTLE THINGS samples, we showed that the ratio between HI mass and DM halo mass remains approximately constant with stellar mass for nearby disc galaxies ($\log(M_{HI}/M_{200})=-1.90$ with a $1\sigma$ scatter of 0.37 dex). In this work, we extend that analysis by incorporating galaxies from the Analysis of the interstellar Medium in Isolated GAlaxies (AMIGA) sample and from the Gassendi HAlpha survey of SPirals (GHASP). The AMIGA sample provides the most rigorously selected sample of isolated galaxies in the local Universe, well-suited for testing whether the relation found in the previous samples holds in interaction-free galaxies. We construct mass models from high-resolution rotation curves and infrared photometry, and derive DM halo parameters. We confirm the proportionality between HI and DM halo masses and find a nearly constant HI-mass-to-halo-mass ratio of $\log(M_{HI}/M_{200})=-1.90$ over nearly five orders of magnitude in stellar mass ($7\leq\log(M_{\star}/M_\odot)\leq 11.5$), with a $1\sigma$ scatter of 0.36 dex. The AMIGA and GHASP samples are statistically consistent with the relation previously found for the SPARC and LITTLE THINGS samples, indicating that it is robust across galaxies spanning a broad range of isolation levels. In contrast, cosmological hydrodynamic simulations such as SIMBA, IllustrisTNG, and NIHAO predict a dependence on stellar mass, with a break at the high stellar mass end for disc galaxies ($M_{\star}>10^{10}M_{\odot}$). Our results thus further demonstrate that the HI-mass-to-halo-mass ratio is remarkably self-similar across rotationally-supported disc galaxies, hinting at mass-independent self-regulation mechanisms that are not yet understood in current theoretical models.

Image quality assessment predicts a mean opinion score (MOS), but for image aesthetics the mean hides the individual, so personalized assessment must start from the few images a new user rates. Fitting each person from scratch fails with ten ratings; networks fine-tuned per person carry over poorly to new kinds of images. We present Undertone-P, which learns from other raters a Gaussian prior over per-person regression weights on read-outs of frozen vision-language model (VLM) judges: the probability-weighted rating and the hidden state at the score token. Since people differ from the crowd in ways other raters reveal, a few ratings need only choose among those ways: a new user costs one closed-form solve, and no network is trained. On Ryu and Yanaka's PARA benchmark, the prior alone reaches a Spearman correlation of .640, above the best published 10- and 100-rating results (.581 and .620); 10 and 100 of a person's ratings raise it to .648 and .676. These ratings add most where tastes depart from the crowd: with 100 ratings, +.067 for the raters who agree least with it and +.206 on LAPIS artworks, against +.037 overall. Without labels, the judges' expected scores predict PARA's MOS at .835 (best published model we ran: .694). PARA-trained networks still lead under PARA's original protocol, likely through a stronger end-to-end MOS model.

Light-matter interaction in time-varying metasurfaces brings about phenomena that transcend the limits of static systems. Temporal modulation enables energy exchange between light and matter, and electromagnetic fields are coupled in both momentum and frequency, with implications across a broad range of photonic applications. Nevertheless, a precise description of such systems necessitates a theory that captures the intertwined effects of spatiotemporal variations while accounting for dispersion in a causal manner, as realistic optical materials exhibit frequency-dependent response and finite temporal memory. Here, a self-contained Floquet-Bloch theory is developed to capture the response of dispersive, time-varying metasurfaces, respecting causality through physically consistent constitutive relations. The theory treats spatial periodicity, nonadiabatic temporal modulation, and material dispersion within a unified formalism, providing access not only to the metasurface's scattering response but also to its inherent modal structure. The formulation is validated against full-wave simulations. As illustrative examples, it is first applied to asymmetric Floquet harmonic generation in an excitonic time-varying metasurface, where excitonic dispersion enables selective harmonic enhancement. It is then used to investigate metasurface-based photonic time crystals, revealing how modal dispersion governs momentum-bandgap formation and dynamics. This work establishes a comprehensive platform for understanding dispersive, time-varying metasurfaces and their underlying physical mechanisms.

For the purpose of modelling ferroelectric switching in wurtzite-structured materials, four cost-effective density-functional theory (DFT) methods (PBE, PBEsol, r2SCAN, and r2SCAN-rVV10) are considered and compared to various ab initio approaches based on the random-phase approximation (RPA), including RPA with singles corrections (RPAR+S), as well as second-order Møller-Plesset perturbation theory (MP2). The purpose is to determine whether an ab initio approach could act as a 'gold standard' for estimating the reliability of DFT, thus determining an optimal DFT method for use in exhaustive tasks such as the training of machine-learning interatomic potentials (MLIP) for large-scale simulations of materials of arbitrary composition and structure such as Al1-xScxN and Zn1-xMgxO, using AlN, Al0.5Sc0.5N, ZnO, and Zn0.5Mg0.5O as model materials. Applications of active learning (AL) are now common, in which results from MLIP simulations are used to enhance the DFT training data set, but future extended active learning (EAL) methods will also need to systematically assess the DFT methodology against a gold standard. Herein, the variability of the ab initio results is found to exceed that required for a robust gold standard, but the DFT and ab initio approaches appear to converge on RPAR+S and r2SCAN-rVV10 as optimal method choices to initiate EAL. The electron correlation energy is found to be dominated by covalent binding effects associated with the high-electron-density anions involved, but the van der Waals dispersion force is seen to be too significant to ignore in ferroelectric modelling.

We study equilibrium and risk evaluation in large sparse economic networks with heterogeneous responses, shocks, and bilateral exposures. Our approach approximates the distribution of equilibrium outcomes through local computations on a limiting rooted network. Under marked local weak convergence, we prove convergence in probability of the empirical equilibrium distribution in two settings: uniformly contractive responses and bounded monotone nonexpansive responses whose lower and upper root iterations coalesce. In the latter setting, the limit is independent of the measurable finite-network equilibrium selection, even when equilibria are not unique. Under bounded domain and continuity assumptions, this convergence extends to law-invariant risk measures, including conditional value-at-risk, with recursion-depth error bounds under additional regularity. Numerical experiments on production networks and payment clearing systems illustrate the accuracy and computational benefits of the local approximation.

Here, magneto-optical magnetometry measurements on Py/NiO bilayers reveal a pronounced magnetic-field-induced wavelength shift, demonstrating strong magneto-optical coupling in this antiferromagnetic system. A systematic and monotonic spectral shift of up to ~ 400 nm is observed as the applied magnetic field increases, saturating at higher fields. Quantitative analysis shows that the associated magneto-optical energy variation is on the order of 107 eV, comparable to the magnon energy scale in NiO. Owing to the large NiO thickness, the observed effect originates from magnons intrinsic to the antiferromagnetic NiO layer rather than from spin currents injected by the Py underlayer. These results provide direct experimental evidence for magnetic-field control of antiferromagnetic magnon energies via magneto-optical interactions, establishing Py/NiO bilayers as a promising platform for optically probing and manipulating antiferromagnetic spin dynamics.

Generalized planning aims to compute a plan that solves all instances of a planning domain. Recent work has used LLMs to automatically generate and debug such generalized plans in the form of Python programs and achieved perfect test data coverage for several domains. However, whether these generalized plans are actually complete, i.e. solve all instances of the domain, could only be determined by manual evaluation. Here, we present an approach for automatically generating generalized plans in Lean together with proofs of their completeness relative to a specification of the domain constraints provided as input. We introduce a semantic-preserving PDDL-to-Lean conversion, and use an LLM to generate both the generalized plan and the formal proof that it solves every instance satisfying the domain constraints. The correctness of the completeness proof is determined by Lean's kernel. We evaluate our approach on 13 commonly used benchmark domains, using GPT-5.6-Sol as the LLM. For 12 of the domains we obtain generalized plans together with valid completeness proofs. This is a major advancement of the state of the art in automatic generalized-plan completeness proofs.

Generative visualization increasingly embeds large language models within direct manipulation and multisensory interaction. Speech, gaze, touch, gesture, sound, and haptics can make probabilistic inference feel like familiar, deterministic tool use. I call this gap a deterministic-affordance mismatch: deterministic interaction cues persist while AI weakens predictability, locality, reversibility, or provenance. Reading the malleable interfaces of Iron Man 2 against systems from Blade Runner 2049, I propose productive friction, including cross-sensory renderings whose detail reflects model uncertainty. Four frictions expose the inference boundary, match action to effect, make stochastic branches tangible, and attribute sensory agreement.

We consider the dynamic Schrödinger bridge problem on a finite state space and exploit its Markov structure to decompose the problem into an endpoint coupling and a collection of conditional Markov bridges. To sample from the conditional bridges, we develop a quantum Gibbs sampler based on quantum walks on the bridge path space. Under uniformly positive transition probabilities, we derive a spectral-gap bound and provide an explicit procedure for preparing the initial state. Compared with the update bounds of the corresponding classical Gibbs sampler, our quantum walk-query complexity improves the dependence on the time horizon $T$ from quadratic to linear. For the endpoint coupling, we adapt a quantum box-constrained Newton method to compute the Schrödinger potentials, improving the matrix-scaling complexity, at fixed accuracy, from quadratic to $N^{3/2}$ in the state-space size $N$. Finally, we show that exponential reweighting extends both the Schrödinger bridge formulation and the Gibbs spectral-gap estimate to models with additive running costs.

In this paper we consider area-preserving twist maps $f:\mathbf{T}^2\to\mathbf{T}^2$ with vertical rotation sets reduced to a single irrational number $\alpha$. Our first result states that under these hypotheses, there exists a unique $f$-invariant foliation $\mathcal{F}_G=\{\operatorname{Graph}(\phi_t)\}_{t\in\mathbf{T}^1}$, where each function $\phi_t:\mathbf{T}^1\to\mathbf{T}^1$ is $K$-Lipschitz for some constant $K=K(f)$. Moreover, we also prove that if $\mathcal{F}_G$ is a Lipschitz foliation, then $f$ is minimal.

We report the first experimental observation of resonantly enhanced propagating phonon transport induced by magnon pumping in a ferromagnetic/piezoelectric bilayer. Surface acoustic waves (SAWs) generated in a 128$^{\circ}$ Y-cut LiNbO$_3$ delay-line device resonantly excite magnetization dynamics in an adjacent Co film through magnetoelastic coupling. The enhancement of the transmitted acoustic signal occurs exclusively when the SAW frequency satisfies the ferromagnetic resonance condition predicted by the Kittel dispersion, providing a direct experimental fingerprint of resonant magnon-phonon coupling. A systematic comparison between the Co/LiNbO$_3$ bilayer and the bare LiNbO$_3$ substrate demonstrates that the observed transmission enhancement originates solely from the dynamic interaction between propagating phonons and coherent magnetization precession. Furthermore, measurements performed at different SAW harmonics reveal that only the harmonic satisfying the FMR condition produces a measurable enhancement, confirming the frequency-selective nature of the phenomenon. These findings establish an efficient mechanism for transferring energy from magnons to propagating phonons and provide a new strategy for actively controlling coherent acoustic transport in hybrid spintronic, straintronic, and quantum phononic platforms.

Dark pools are designed to provide pre-trade privacy, liveness, and post-trade confidentiality - concealing order flow before execution and limiting information leakage after. Decentralized dark pools, such as Renegade, aim to replicate these properties without custodial risk, using secure multi-party computation (MPC) and zero-knowledge proofs for private order matching and verifiable settlement. We show that Renegade's cryptographic guarantees do not deliver these dark pool properties in practice. MPC-with-abort ensures correctness but not fairness: a party may learn the match result and abort without penalty, breaking pre-trade privacy. We demonstrate that the protocol's discovery layer further leaks trading intent before MPC even begins, and that sustained probing via selective abort can probabilistically reconstruct counterparty order history, threatening post-trade confidentiality. We also show that the absence of input-consistency checks prior to MPC execution enables a griefing attack using invalid state commitments requiring no real token holdings that continuously locks honest users' wallets and wastes compute, breaking liveness under sustained conditions. We further analyze over 700,000 Renegade transactions on Base and probe the P2P layer, finding that the network is effectively centralized: 88% of traffic routes through a handful of relayers, with only four nodes sustaining the P2P layer. Since relayers hold their users' wallet state in plaintext, this concentration means the system operates as a centralized orderbook in practice - reproducing off-chain the information asymmetry that dark pools are designed to eliminate. Together, our results show that cryptographic privacy does not imply dark pool security: pre-trade privacy, liveness, and post-trade confidentiality each require additional protocol-level guarantees beyond MPC correctness.

Surgical robots require highly dexterous and compact robotic systems capable of operating effectively within confined anatomical spaces. However, due to limited access provided by a single incision, the miniaturization and maneuverability of these robots still need to be improved. In this paper, we propose the design and modeling of a single-port three-arm robotic tool containing one major cannula (7.14 mm outer diameter (OD)) and three steerable minor cannulas (1.93 mm OD). By integrating the proposed 12 degrees-of-freedom (DoFs) steerable robotic tool with a 7-DoF robotic arm, this robotic system can potentially achieve multi-arm manipulation capability. We present the design of the steerable robotic tool consisting of tendon-driven joints controlled by a compact actuation system, derive the kinematic model, and validate both the static and kinematic models through experiments. The performance is evaluated with the root mean square error (RMSE) and mean absolute error (MAE) computed between the experimental data and the kinematic model.

Several experiments reported measurements of soft photons in hadronic collisions far above the predictions from QED. This work investigates possible production during the hadronisation process. It is shown that the assumption of isotropic emission of photons during the hadronisation process leads to a consistent description of the experimental data.

Due to progressing miniaturization, material fatigue is expected to gain importance on microscopic scales. Nevertheless, most non-destructive testing techniques are optimized for macroscopic scales. Therefore, new approaches applicable to miniaturized samples need to be explored. High spatial resolution imaging of a material's magnetic stray field enables sensitive detection of microstructural changes due to the local interplay of magnetic properties, strain, and defects. Up to now, this approach has rarely been used for materials testing since high-resolution magnetic sensing is traditionally challenging. Novel quantum sensing techniques offer the potential to close this gap. One of the most prominent quantum sensors, the nitrogen vacancy center in diamond, is investigated for non-destructive testing applications within the scope of this work. An experimental system based on a widefield sensing approach was developed to image magnetic stray field distributions within seconds to minutes over areas up to 1 x 1 $mm^2$. Magnetic sensitivities below 10 $\mu$T/$\sqrt{\text{Hz}}$ and a spatial resolution of 1-2 $\mu$m were achieved. The technique offers high mechanical stability, which is crucial for non-destructive testing applications. Comparison with magneto-optical Kerr effect measurements showed that the magnetic field maps are strongly correlated with the material's surface domains while containing additional information from deeper inside the sample. Characteristic changes in the magnetic stray field distribution were detected in an electrical steel sample after cyclic loading. A potential marker for early fatigue damage was deduced from the splitting gradient distribution, while 2D Fourier transform analysis provided additional insight.

Motzkin paths of length $n$ and standard Young tableaux (SYT) with $n$ cells and at most three rows are both counted by the Motzkin numbers, and many bijections between them are known. The Riordan numbers count the subfamilies of Riordan paths (Motzkin paths with no horizontal step on the $x$-axis) and of tableaux whose three row lengths have the same parity, but none of the known bijections restricts to these subfamilies. We introduce the set of $321$-avoiding permutations in which every left-to-right maximum is either a descent or a fixed point. This family is counted by the Motzkin numbers, and its fixed-point-free elements are the “Riordan permutations” of Callan. We give a bijection from Motzkin paths to these permutations under which Riordan paths correspond to Riordan permutations. We then give a bijection from these permutations to SYT of height at most three, obtained from Robinson–Schensted insertion followed by a parity correction, under which Riordan permutations correspond to tableaux with rows of equal parity and the number of left-to-right maxima becomes a simple tableau statistic. Via Dyck paths, we connect these objects to further families counted by the Riordan numbers, including derangements of genus zero and SYT of shape $(k,k,1^{n-2k})$.

Robotic embodiment encompasses the sensing, kinematics, dynamics, geometry, actuation, and control through which an agent physically interacts with the world. These properties vary across robots and change over time. We argue that general embodied intelligence requires learning that accumulates across these differences. Prevailing methods that engineer correspondences to bridge embodiment differences offer immediate practical gains, but their assumptions limit the scope of transfer in the long run. Instead, a more general approach should discover representations that support transfer to a larger range of embodiments as experience grows. We propose embodiment diversity as a promising axis of scaling, and identify broad learned priors as a complementary ingredient. We call for evaluations that better characterize embodiment gaps and transfer performance. More broadly, cross-embodiment learning connects the practical challenge of learning from heterogeneous robot experience with a broader scientific pursuit inspired by nature - physical intelligence that adapts and co-evolves with its embodiments to gain agency over its behavior and physical forms.

Automatic tagging is a core task in Music Information Retrieval (MIR), yet most tagging systems exploit only audio. Live music performance is inherently multimodal, as semantic labels such as instruments, regional styles, and dance forms are encoded simultaneously across acoustic, visual, and embodied performance cues. This is especially true of culturally specific repertoires such as Greek traditional music, which remain underrepresented in MIR benchmarks. In this paper, we investigate whether the use of dancer pose provides complementary information for automatic tagging in Greek traditional music beyond audio. Using the Lyra dataset, we extend prior audio-only work by extracting aligned video features and pose-derived skeleton streams, enabling an experimental setting for multimodal auto-tagging. We further introduce an automated pipeline for extracting primary-dancer skeleton sequences from in-the-wild dance footage, combining dance-scene detection, multi-person tracking, dancer selection, pose estimation, and quality filtering. We compare unimodal, all bimodal combinations, and trimodal systems using multiple fusion strategies. Audio remains the strongest single modality (AST: macro ROC-AUC 0.821), while skeletons, though weak in isolation, enhance performance through multimodal fusion. The best trimodal system improves macro ROC-AUC by about 4 percentage points over the strongest audio baseline.

The 12-h wave in meridional winds in the mesosphere and lower thermosphere (MLT) during the solar minimum 2018-2019 Arctic winter is investigated using sodium lidar observations at Poker Flat Research Range (PFRR), Chatanika, Alaska (64N,147W). Nightly 12-h wave amplitudes increased significantly during December-January, with amplitudes exceeding 130 m/s above 97 km on several days. This was more than double the 12-h wave amplitudes observed outside this time period. Meteor radar winds over Chatanika also showed significant increase in meridional wind 12-h wave amplitudes at altitudes between 82 and 97 km during this time period with lower amplitudes than lidar measurements. The strong variation in 12-h wave amplitudes was not correlated with SME index, though the largest amplitudes coincided with the sudden stratospheric warming (SSW) in early January. Measurements were compared to a seasonal WACCM-X model run and four days of HIAMCM. For the four dates of available HIAMCM data, 12-h wave amplitudes over Chatanika were found to be similar between both HIAMCM and WACCM-X and the lidar below 97 km, with amplitudes measured by lidar exceeding the models at altitudes between 97-105 km. All measurements followed a similar seasonal trend with increasing amplitudes at the end of December/early January. Fits of SW2 from WACCM-X show the SW2 tidal amplitude following similar seasonal trends to 12-h wave measurements. These high-resolution lidar measurements indicate that localized 12-h wave amplitudes are larger than previously reported by studies using meteor radar measurements.

Many node-classification GNNs apply a linear classifier to a nonnegative mixture of local messages. An error can reflect either poor mixture weights or a reachable logit set poorly positioned for the classifier. We separate these causes with an exact-mass linear program and two learned post-hoc repairs. Every reweighted prediction has an equivalent centered logit translation, but only translations in a message-induced displacement set are realizable by reweighting. Across eight datasets, eight GNN backbones, and ten splits, mean accuracy rises from 62.6% for the frozen models to 63.8% with reweighting and 65.3% with set-conditioned translation. A parameter-matched node-only translator reaches 64.6%, showing that translation explains most of the gain while the message set supplies a smaller additional benefit. Although oracle reweighting can correct many errors, label-free reweighting captures little of this potential: local evidence is often present but hard to select, and relaxing the evidence constraint is more effective than learning within it.

The growing adoption of blockchain technologies, particularly the Ethereum platform, has amplified the critical role of smart contracts in decentralized applications. However, the increasing complexity and financial value of these contracts make them prime targets for cyber attacks. In this work, we present a transformer-based approach for the detection of vulnerabilities in smart contract fragments written in Solidity. Leveraging the representational power of pre-trained Large Language Models (LLMs), we construct a robust pipeline that includes the definition of a ground truth dataset, labeling code fragments as vulnerable or safe. We then fine-tune a BERT-based architecture on this dataset, enabling the model to capture the syntactic and semantic patterns specific to Solidity code. Our fine-tuned model demonstrates strong performance, achieving an F1 score of 92%, and highlighting the effectiveness of LLM adaptation in enhancing smart contract security through deep contextual understanding.

Multi-passage corpus poisoning often repeats one target claim across similar documents, creating correlated lexical and semantic patterns that similarity- and conflict-aware defenses can suppress jointly. We introduce DnD (Divide and Doubt), a targeted attack based on two principles: distributing support for the target answer across stylistically diverse passages, and including a passage that casts doubt on evidence for the reference answer. The first disperses poison-passage representations in embedding space, while the second strengthens target adoption when multiple poisoned passages are retrieved. We evaluate DnD on two open-domain QA datasets across three LLMs and nine RAG configurations, under both black-box and white-box access to the retriever. Across these settings, DnD matches or outperforms prior attacks in most configurations, with its largest gains against clustering- and conflict-aware defenses.

We establish a Harder-Narasimhan theory for torsion coherent $F$-gauges over a perfect field $k$ of positive characteristic and completely classify its stable objects. Our construction uses Ekedahl's derived equivalence with coherent complexes over the Raynaud ring, where the theory refines his half-integral type filtration on diagonal dominoes. At each finite rational slope $d/q$ in lowest terms, there is a unique stable diagonal domino up to Breuil-Kisin twist, constructed explicitly from the corresponding Christoffel word. The semistable category of this slope is equivalent to a product of $q$ copies of the zero-slope category. The classification of stable objects corrects Ekedahl's classification of weakly simple diagonal dominoes of type $\frac{1}{2}$. The theory upgrades to a locally finite Bridgeland stability condition on $\mathrm{Perf}_{\mathrm{tor}}(k^{\mathrm{Syn}})$.

Elongate multi-legged robots use coordinated body waves and distributed legs to move through cluttered terrestrial environments. However, as housing actuators for independent leg control can require bulky body segments, their non-streamlined body and limb structure makes it difficult to achieve swimming capability comparable to their terrestrial locomotor performance. At the water surface, we found that the multi-legged robots we tested unexpectedly moved backward: their body waves traveled in the same direction as their displacement, i.e., swimming with a direct wave. We found similar behavior in the centipede Lithobius forficatus, which swims with a direct body wave and periodic leg movement. To study how distributed legs contribute to direct-wave swimming, we analyze animal kinematics and develop a multi-legged robophysical model that allows independent variation of leg morphology and stiffness, body-wave direction, and leg coordination. Robophysical experiments show that direct body waves produce consistent forward motion under the tested conditions and that swimming performance depends on body–leg coordination. Additionally, directionally compliant legs increase displacement from approximately 0.08 to 0.21 body lengths per cycle relative to rigid legs under matched anti-phase actuation. These findings clarify how distributed appendages contribute to surface swimming and establish gait and morphology design principles for extending multi-legged field robots from terrestrial locomotion into aquatic environments.

Organizations increasingly use LLMs for policy, compliance, risk, and operational decision support, requiring evidence validation, review routing, version control, and auditability. We introduce Policy-as-Skill (PaS), a modular runtime that packages these functions as executable, versioned policy capabilities. Thirteen methods are evaluated with a fixed Gemma4 backend on 600 development tasks. PaS+Audit achieves 53.8% exact accuracy, macro-F1 0.346, review F1 0.854, citation precision 1.000, policy-reference recall 0.984, and audit completeness 1.000, outperforming LLM+RAG on most governance and review metrics. Deterministic control raises aggregate accuracy to 61.2% but is strongly task dependent, supporting selective rather than universal rule-based intervention.