Papers

Linearizability is a standard correctness condition for concurrent objects. It takes a correct by construction sequential specification of a concurrent data structure, and produces a composable specification under atomic interleavings of operations. However, some operations like atomic exchangers are not specifiable atomically, making it difficult to reason compositionally about their specifications. Interval linearizability is the state of the art in relaxing linearizability's constraints. Formally it is a powerful framework, but it requires an interval-sequential specification of the target data structure, which we show is challenging to correctly generate or understand, even for simple operations on a single concurrent register. We propose substep linearizability, which we show is equivalent to interval linearizability using X-CCorrectness, a relaxed concurrency framework we created to compare existing linearizability semantics. We show that substep linearizability's sequential specifications are suitable for previously infeasible specification tasks, while being comparable in complexity to linearizability's sequential specifications. For evaluation, we model an atomic exchanger, a popular non-linearizable queue structure, and snapshot isolation between database transactions.

Precise manipulation of qubits (spins) requires strong suppression of noise arising from couplings between qubits and their environment, as well as from driving field fluctuations. We propose a double-pulse sequence and an eighteen-pulse sequence to realize a single spin pi-rotation under off-resonance conditions (instead of conventional on-resonance or near-resonance conditions). It is proved that the latter sequence performs much better than the former in simultaneously suppressing both dephasing and driving field noise. As a proof of principle, we experimentally demonstrate, by implementing both sequences in an NMR spin system with a polarized initial spin state, that the off-resonance pi-rotation fidelity reaches as high as 0.9992 +/- 0.0008 for the double-pulse sequence and 0.9997 +/- 0.0006 for the eighteen-pulse sequence. In addition, the noise-tolerance range of the eighteen-pulse sequence is about 20 percent of the driving field strength, approximately three times wider than that of the double-pulse sequence. Besides its potential application in the NMR community, our proposal provides an alternative practical way to accurately control qubits by suppressing the coexisting transversal and longitudinal noises.

We develop a theoretical model for radiation-regulated spherical accretion with an inner accretion disk. In our model, the outer accretion flow remains quasi-spherical and optically thick, while the luminosity generated by the inner disk is self-consistently coupled to the accretion rate and irradiates the inflowing gas. By solving the global transonic structure with radiative transfer treated in the diffusion approximation, we investigate how radiation regulates the flow structure and the accretion rate. We find that the luminosity is constrained by a generalized Eddington limit, leading to a saturated accretion rate typically lower than the classical Bondi prediction by orders of magnitude. At sufficiently low densities, the enthalpy flux of the spherical inflow contributes substantially to the emergent luminosity, in which case the inner disk luminosity is reduced and the accretion flow enters an unsaturated state. A criterion is derived to estimate whether the accretion flow is saturated. We further demonstrate the applicability of our model by considering fallback accretion in failed supernova systems. The predicted nearly constant emergent luminosity, constrained by the generalized Eddington limit, provides a possible explanation for the observed luminosity plateau in the failed supernova candidate M31-2014-DS1. The presence of an inner disk also provides sufficient radiation to avoid the under-luminous predictions of a purely spherical fallback model extending down to the central object. Our model provides a framework for studying radiation-regulated accretion flows in systems where low but finite angular momentum leads to the formation of an accretion disk within a larger-scale quasi-spherical inflow.

In this work, using nonlinear electrodynamics (NLED) as a source in the field equations of extended scalar-tensor-Gauss-Bonnet (E-STGB) gravity, a novel type of static and spherically symmetric spacetime is found. We show that this novel type of spacetime can be used to explain the rotation curves and gravitational lensing of spiral galaxies. The NLED model used to derive our solution, in the weak and strong field limits, reduces to Maxwell electrodynamics. Furthermore, for a wide range of parameter values where our solution describes the gravitational field of a spiral galaxy, the NLED model satisfies the null and weak energy conditions. Hence, our results suggest that the extended scalar-tensor-Gauss-Bonnet gravity minimally coupled to nonlinear electrodynamics (E-STGB-NLED) provides a viable alternative to explain flat rotation curves of spiral galaxies without invoking dark matter.

We derive the covariant equations of nondissipative chiral hydrodynamics from their equal-time Hamiltonian formulation and identify the Lie–Poisson structure implicit in the chiral brackets. This structure combines fluid motion with local chiral transformations. For a parity-invariant Hamiltonian quadratic in chiral charge densities and pion gradients, without external backgrounds, we derive and solve the constraints imposed by Lorentz boosts. Their coupled nonlinear sector is equivalent to the linear boost transformation laws of a symmetric tensor, which determine the velocity dependence from rest-frame data. For finite, positive rest-frame coefficients, finite ordinary-fluid energy density and thermodynamic derivatives, and a momentum–velocity relation that can be solved in either direction, the chiral Hamiltonian coefficients remain finite at every fluid speed below light speed if and only if the rest-frame pion propagation speed does not exceed the speed of light. This formulation makes symmetry constraints explicit and provides a starting point for future thermodynamically consistent dissipative extensions.

We use centered differences to construct Chernoff families and derive even asymptotic expansions for their iterated products. Richardson extrapolation is then used to cancel polynomial part and thereby raise the convergence order. As illustrations, we apply the construction to uniformly parabolic, Schrödinger-type second-order operators and to a third-order dispersive operator, obtaining, at extrapolation depth $m$, the rates $ O(n^{-2m}) $ on the Sobolev scales $H^{2m+\delta}$, $H^{4m}$, and $H^{5m}$, respectively, with arbitrary $\delta>0$. In the analytic category, quantitative all-order estimates allow the extrapolation depth to grow with the base resolution and yield the super-fast rates $e^{-cn^2}$ for the two second-order models and $e^{-cn} $ for the third-order model. The constructions are explicit in terms of the coefficients of the underlying differential operators and do not use the target semigroup in their definition. This provides a constructive positive answer to the challenge of super-fast Chernoff approximation formulated by Remizov in Problem 30 of arXiv:2410.00416.

Predicting single-cell perturbation responses requires mapping between two unpaired populations of cells measured before and after the perturbation. Optimal transport (OT) provides a principled framework for constructing such mappings by minimizing the transport cost, and neural OT methods (e.g., CellOT) have been employed for this prediction task. However, these methods rely on a fixed transport cost, typically the squared Euclidean distance. This cost may connect cells with the wrong responses. Lineage tracing provides the information needed to learn the cost, as heritable barcodes link the cells of the same clone before and after the perturbation. Here, we propose \CloneOT, a neural inverse OT method that learns both the transport cost and the transport map from partially lineage-traced single-cell datasets. \CloneOT\ represents the cost as a neural network. It learns the cost by minimizing the extra cost that the lineage adds to the optimal transport, so that the cheapest transport keeps the cells of each clone together. Our experiments demonstrate that \CloneOT\ recovers the correct transport on 2D datasets, while the squared Euclidean cost fails. Moreover, \CloneOT\ outperforms five baselines on 14 lineage-tracing perturbation scenarios of cancer cells, for both seen and unseen clones. Furthermore, we show that \CloneOT\ reveals the clone-specific transcriptional programs of drug-tolerant persister cells and forecasts the fate of each clone from its cells before treatment. Our implementation is available at https://github.com/poseidonchan/CloneOT.

We introduce a multiplicity one defect that quantitatively measures the failure of multiplicity one for the Jacobian of the Shimura curve \(X^D(M)\) attached to an indefinite rational quaternion algebra of discriminant \(D\) and level \(M\), localized at a maximal ideal of the Hecke algebra. By constructing a certain lattice inside the local Tate module, we obtain an explicit upper bound for the defect in terms of congruences between eigenforms and the lattice-Tate module discrepancies associated with the individual newforms. We then show that the local ratio between the congruence exponent and the modular exponent of modular abelian varieties attached to newforms is bounded above by the multiplicity one defect. Finally, we examine the relation with the Gorenstein property and give an explicit example at level \(126\) where the local Hecke algebra is Gorenstein even though multiplicity one fails. In this example, we compute the multiplicity one defect to be \(3\), which is also the local congruence–modular exponent ratio.

Layered van der Waals CuInP2S6 crystals exhibit a rich sequence of temperature- and pressure-driven phase transitions, giving rise to multiple structural phases with distinct physical properties. Despite extensive experimental and theoretical efforts, unambiguous identification of these phases remains challenging because of their structural similarities and complex vibrational responses. In this work, we employ first-principles density functional theory calculations to systematically investigate the vibrational properties of the reference paraelectric C2/c and P-31c phases and their corresponding dynamically stable ferroelectric Cc and P31c phases, as well as the antiferroelectric P-1 and P-3 phases. From calculated phonon spectra and symmetry analysis, we determine the Raman- and infrared-active modes of each candidate structure and identify characteristic vibrational fingerprints that enable their phase determination through vibrational spectroscopy. Furthermore, using Porto notation, we establish polarization- and crystal-orientation-dependent Raman selection rules for different backscattering geometries and identify the modes that are expected to emerge or vanish under specific experimental configurations. Thus, our results provide a comprehensive theoretical framework for interpreting Raman and infrared measurements of CuInP2S6 and for reliably identifying its competing structural, ferroelectric, and antiferroelectric phases across temperature- and pressure-driven phase transitions.

Deep neural networks are capable of both learning generalizing solutions on natural tasks and memorizing arbitrary mappings. We argue that understanding the former requires understanding the latter, which is much less studied. In what ways do networks that generalize or memorize differ in their behavior or their training dynamics? Do memorizing networks simply function as a lookup table where input-output mappings are stored independently? We investigate these questions in the context of CNNs trained on image classification tasks with randomized labels or pixels. We first study the behavior of such memorizing networks on unseen images. We find that independently initialized networks trained on the same random labels learn different predictive functions: they make different predictions on test examples, in contrast with generalizing networks. We then examine the effect of dataset size on memorization through the required network capacity and training time to interpolate the training set. Counter-intuitively, we find that the number of presentations of a given example required for complete memorization can decrease with dataset size even when input-output pairs share no common structure. Although memorizing networks do not learn a universal solution, they exhibit interesting behaviors which call for further investigations into the underlying mechanisms.

Efficient spin transport with high spin mobility is of both fundamental and practical significance for spintronics. Spin can be transferred by moving particles. However, particle scattering limits spin transfer and prevents high spin mobilities. In this work, we study the kinetics of spin transport mediated by a quantum system of indirect excitons (IXs) in a van der Waals heterostructure composed of atomically thin transition-metal dichalcogenides (TMDs). This spin transfer mechanism allows for the realization of efficient spin transport characterized by a high spin mobility reaching 10$^5$ cm$^2$/(eV$\cdot$s), which is orders of magnitude higher than that achieved via standard diffusive spin transport in TMD materials. The suppression of IX scattering enables high-mobility spin transport.

We develop a unified multi-scale framework for sparse domination of multilinear operators beyond the multilinear Calderón–Zygmund theory. More precisely, under mild hypotheses on a support condition, weak type and restricted strong type uniform bounds for finite partial sums, and single-scale improving and regularity estimates, we establish $(r_1, \ldots, r_m, s')$-sparse form bound for a family of localized $m$-linear operators uniformly over all finite partial sums. This sparse bound yields weighted norm inequalities for the corresponding multilinear Muckenhoupt classes $A_{\vec{p}, (\vec{r}, s)}$, including the quasi-Banach range. As applications, we determine the optimal open sparse range for multilinear rough singular integrals, and characterize sparse domination for bilinear pseudo-differential operators. Further applications include sparse bounds for bilinear oscillatory Fourier multipliers, bilinear spherical maximal operators, and bilinear Bochner–Riesz means.

We introduce Confidential Error-correcting Codes (CECs), a new cryptographic primitive that enables reliable decoding from perturbed codewords while preserving message confidentiality. We formalize this primitive through three security properties: confidentiality, robustness, and soundness. To realize this goal, we develop a general framework for constructing CECs based on a Randomness-Security Separation (RSS) structure, which first encodes a nonce to introduce randomness and then masks the encoded payload with a pseudorandom function. Based on our RSS framework, we instantiate CECs, called Conc-CEC, using concatenated Reed-Muller and Reed-Solomon codes. The resulting Conc-CEC admits provable guarantees for all three properties. Notably, compared to PRCs (CRYPTO'24), another recent cryptographic primitive with error-correcting capability, CECs achieve significantly stronger concrete security guarantees under practical parameter settings. As a primary application of CECs, we further present CECMark, a watermarking scheme for AI-generated images. Our experiments on real-world generative image models show that CECMark consistently outperforms state-of-the-art watermarking methods, including Gaussian Shading (CVPR'24), PRCMark (ICLR'25), and T2SMark (NeurIPS'25). In particular, CECMark achieves higher robustness across all 56 evaluated perturbations and ensures a false positive rate below $10^{-5}$ without introducing side effects. Taken together, our work lays a novel cryptographic foundation for trustworthy watermarking schemes that, for the first time, unify provable security guarantees with strong practical reliability.

Real-world public health policies are often enacted by local jurisdictions, but their effects can ripple through connected populations. For causal evaluation, this creates a difficult spatiotemporal problem: intermittent policy updates create path-level treatments, and because spillover depends on neighboring treatment paths along with the focal-unit's own history, controlling time-varying confounding requires high-dimensional, multimodal information from both sources. To address this challenge, we develop a neural spatiotemporal spillover framework that extends potential-outcome causal analysis to networked policy-systems with unit-level effects and structured interference. The framework indexes each unit's potential outcomes through own-treatment and neighborhood spillover-exposure paths, yielding direct, spillover, and total effects. We introduce an LSTM-based representation-learning encoder that integrates each unit's own and neighborhood histories into compact sequential summaries for downstream modeling and causal estimation. These summaries feed a self-normalized kernel inverse-probability-weighting estimator for continuous spillover exposure. Beyond theoretical guarantees, simulations show improved performance over outcome-regression baselines; in a practically motivated county-level COVID-19 policy study, the framework finds both business-and-economic restrictions and education-and-childcare policies associated with lower subsequent case burden, with larger estimates for business-and-economic restrictions. Overall, it provides a scalable, policy-interpretable framework for evaluating unit-level effects in high-dimensional, multimodal spatiotemporal systems when interventions spill over across space and time.

Prestack seismic inversion often suffers from inherent ill-posedness and is significantly affected by complex factors such as the wavelets, parameter coupling, heterogeneous media, fault displacements, and stratigraphic variations, making it difficult for the conventional inversion methods to accurately recover subsurface information. To address these challenges, this paper proposes a global inversion method based on multi-dimensional matrix multiplication to optimize the inversion results under structure constraints. This method first represents the inversion model as a four-dimensional tensor representing spatial and parameter dimensions. By performing multi-dimensional matrix multiplication operations in various directions of the model, a multi-dimensional tensor of observed data is effectively constructed. Secondly, multi-dimensional regularization are introduced. By analyzing the impact of different dimensional constraints on the inversion results, targeted parameter selection and adjustment are achieved, significantly enhancing the stability of the inversion process and effectively suppressing noise. Finally, stratigraphic and fault constraints extracted from seismic interpretation are incorporated to improve the geological coherence and reasonableness of the inversion results. Both synthetic and field data experiments demonstrate that the proposed method effectively enhances the robustness of inversion in complex structural areas while maintaining high computational efficiency, validating its effectiveness in practical applications.

Activation steering adds a vector to a language model's internal representations at inference time. It is a lightweight alternative to fine-tuning for behavioral control. The standard construction, contrastive activation addition (CAA), is fragile, as on many inputs it shifts behavior in the wrong direction. To explain this fragility, we first study the geometry of the steering vector. We introduce steering Hessian, the matrix of second order derivative of the model's behavioral loss with respect to the steering vector. It captures the output behavior response sensitivity of the model under small perturbations of the steering vector along each direction. Its top eigenvectors are sharp directions: tiny perturbations along them can cause large, erratic, input-dependent behavioral swings. Bottom eigenvectors are flat directions: behavior is robust along them. Removing the sharp component of a steering vector approximately preserves its direction and magnitude (cosine > 0.974, norm reduction < 3% for typical settings) while reducing the vector's directional curvature by up to 12x (median 4x across 36 conditions). We evaluate on three model families (Qwen2.5-7B, Llama-3.1-8B, Mistral-7B), with four behavioral tasks, and 36 token-level conditions. Steering along sharp directions reverses the intended behavior in every condition tested, with anti-steer rates of 53-100%. Sharp directions account for only 1-5% of the CAA vector's squared norm. However, they drive a disproportionate share of the gap between the steered and unsteered next-token distributions during autoregressive generation. We term this one-line correction as sharp removal; it outperforms standard CAA in open-ended generation across all 25 conditions tested. A jailbreaking experiment further shows that sharp directions act as noise rather than signal: they disrupt coherent behavioral control regardless of steering polarity.

We study nematic liquid crystal phases confined to bounded, two-dimensional (2D) domains, in the reduced Landau-de Gennes (rLdG) theoretical framework. The rLdG framework is equivalent to the celebrated Ginzburg-Landau (GL) theoretical setting, under certain assumptions. In the asymptotic limit of large domains (or low temperatures), we recall the concept of the finite-dimensional renormalized energy in the GL theory [Bethuel et al., GL Vortices, Prog. Nonlinear Differ. Equ. Appl. 13, Birkhäuser, 1994.] and generalize it to polygons with sharp corners. The renormalized energy is fully defined by the triplet, $(\mathbf{a},\mathbf{c}, \mathbf{d}$), of topological degrees of the interior defects ($\mathbf{d}$) and the winding number contribution of polygon vertices ($\mathbf{c}$) and interior defect locations ($\mathbf{a}$), subject to sum rules dictated by the imposed Dirichlet boundary condition. We use the critical points of the renormalized energy to construct initial conditions for the rLdG solver, recover known energy-minimizing and non-energy minimizing saddle points of the rLdG energy on a disc, regular hexagon and rectangle along with the discovery of previously unreported rLdG solution branches. The renormalized energy approach is computationally substantially more efficient than a direct PDE-solver in the rLdG setting but misses solution branches with line defects and finite-size effects.

Heterogeneous medical image data motivate domain-specialized diffusion experts, but expert specialization alone does not determine how their predictions should be combined during sampling. Learning a global router is particularly challenging when the server has limited, potentially unrepresentative images. We study two router-learning strategies for a fixed expert pool. Server-Only learns input- and timestep-dependent expert assignments from server images with component labels. Anchor-Delta (AD) uses the same images and frozen experts to obtain additional evidence from denoising losses. It distills targets that combine this evidence with the current router's assignments, while a proxy gate controls a subsequent refinement step. Component labels describe data origin; expert losses measure empirical prediction performance. We evaluate these distinct signals through routing analyses and paired medical image generation. An AD-trained router improves generation metrics over Server-Only in a heterogeneous four-domain setting, and a separate server-data study shows gains in held-out routing over a corrected-routing comparator. Additional paired evaluations show that the effect of inference-time evidence mixing varies across benchmarks. These findings establish server-data availability and expert performance as distinct considerations in learning routers for multi-domain diffusion models.

We consider the three-dimensional Fröhlich polaron at fixed coupling strength $\alpha>0$ and total momentum $P \in \bbR^3$. We show that the bottom of the spectrum reaches the essential spectrum at a finite value of $|P|$, implying that the fiber Hamiltonian has no ground state at sufficiently large total momentum. This extends the known weak-coupling result to arbitrary coupling. Our approach describes states near the spectral threshold by a fully interacting polaron with small residual momentum and one additional phonon carrying almost all of the total momentum.

We find a new presentation of the orbifold fundamental group of the moduli space of cubic surfaces, based on the Artin group of the Petersen graph. This group has two well-known quotients, which appear in the simplest possible way: killing the squares resp. sixth powers of our generators yields the Coxeter group $W(E_6)$ resp. a certain arithmetic group in $\operatorname{PU}(1,4)$. Along the way we present a certain central extension by $\mathbb{Z}$, of the genus zero mapping class group with any number of punctures. The $6$-puncture case is essential in this paper, and the general case may have independent interest.

In this paper, we present a discrepancy-based approach to cost-preserving rounding under hard constraints. Starting from a feasible fractional solution, our rounding procedure preserves the designated hard constraints, never increases the linear objective, and bounds the error in the remaining constraints by the discrepancy of a suitable column-difference matrix. Our second discrepancy-theoretic result bounds the discrepancy of a matrix in terms of a novel column parameter, called mountain distance. This notion captures ordered structural features of a matrix which are invisible to the classical Beck–Fiala and Komlós column-norm parameters. We show a variety of applications of these two results. In particular, we obtain new results for min-cost unsplittable flows, bounded-turn routing in grids and tori, rainbow prefix hypergraph decomposition, and $2$-multicast scheduling.

Recent advances in self-improving LLM agents have demonstrated the potential of AI systems that drive their own intelligence. However, many domain-specific applications rely on specialized foundation models that possess strong domain expertise but remain dependent on human-designed improvements. Extending recursive self-improvement to these models presents a key challenge: as a specialist improves, the strategies used to advance it must also adapt. To address this challenge, we propose Heterogeneous Agentic CoEvolution (HACE), a framework that jointly evolves specialized foundation models and LLM agents through reciprocal learning. HACE improves the specialist through supervised learning while training the agent with on-policy reinforced self-training, guided by verified performance gains on the current specialist. The agent's exploration, in turn, informs subsequent specialist training, enabling both models to adapt to each other's progress. To sustain this reciprocal learning, HACE further employs evolutionary search to optimize their interaction harness, adapting how the models exchange information and select training experiences as their capabilities evolve. We evaluate HACE across three challenging domains: recommendation systems, protein function prediction in biology, and crystal stability prediction in material discovery. We show that HACE outperforms the evaluated self-improving methods on the primary task metrics across all three domains. These results shed light on heterogeneous coevolution as a universal approach to improving specialized foundation models through reciprocal interaction with a learning agent.

Strong-field physics uses intense optical fields to probe and control electron motion on ultrafast timescales. Recent advances in bright squeezed vacuum have brought macroscopic quantum states of light into the ultrafast strong-field regime, opening prospects for high-harmonic generation and photoionization. Yet linking its phase-space structure to subcycle electron dynamics remains challenging. Here, we demonstrate strong-field tunneling ionization of xenon using a phase-referenced combination of coherent and squeezed fields, with the coherent field defining the ionization clock and quadrature fluctuations of the zero-mean squeezed field providing control. The squeezed field selectively modulates atomic tunneling barriers and electron wave packets, imprinting signatures on photoelectron momentum distributions, number statistics, and energy spectra. Using coincidence measurements of photon number and phase-sensitive photoelectron momentum distributions, we perform quantum state tomography of the driving field by reconstructing its Wigner function. These results establish a bidirectional interface between quantum optics and nonperturbative strong-field electron dynamics, in which the driving quantum field shapes subcycle electron motion while photoionization provides a phase-sensitive probe of the field's quantum state. This work opens new routes toward controlling electron dynamics with quantum light and phase-sensitive quantum state metrology through ultrafast strong-field electron dynamics.

Nonverbal vocalizations (NVVs), such as laughter and sighs, convey meaning alongside words. NVV-aware speech systems must recognize and generate these events while preserving lexical accuracy and perceptual quality. Yet differences in annotation, resources, and evaluation complicate comparisons, and aggregate scores can obscure specific weaknesses. We review ten system descriptions from the NVVSpeech Challenge 2026, covering understanding and generation in Chinese and English across 16 event categories. Our analysis examines supervision construction, task adaptation, and inference-time or preference-based feedback. Results reveal persistent recognition difficulties with breath and gasp and trade-offs between event recognition and lexical accuracy, as well as between generation intelligibility and acoustic quality, with human listening providing complementary evaluation evidence. These findings identify priorities for advancing NVV-aware speech understanding, controllable generation, and comparable evaluation.

The increasing integration of variable renewable generation introduces weather-driven variability and uncertainty into power-system operations. System operators manage the resulting deviations between day-ahead forecasts and real-time net load by scheduling upward and downward flexibility reserves. These requirements have traditionally been set using simple deterministic rules. Newer data-driven methods tailor reserves to forecasts and operating conditions. However, their reliability may deteriorate when conditions depart from the historical data. We address this limitation with a sequential framework that updates day-ahead reserve requirements as operating outcomes become available. Within this framework, Adaptive Conformal Inference targets the frequency of insufficient reserve, whereas Rolling Risk Control targets expected uncovered deviation, the mean residual beyond the scheduled reserve. We prove finite-horizon bounds on realized miscoverage and capped uncovered-deviation loss. We test the framework on real ERCOT data in a chronological deployment simulation that respects operational timing: outcomes from one day first affect the reserve schedule two days later. Training and validation use 2022–2023 data; the uninterrupted test covers 2024–2025. Coverage calibration achieves approximately 90% coverage, with upward and downward miscoverage near 5%. Risk calibration brings expected uncovered deviations close to their budgets while using 5.3% less reserve than the uncalibrated requirements. The framework thus produces adaptive reserve requirements that provably control the selected reliability metrics over finite horizons without assuming stationarity or exchangeability.

Co-circulating influenza, COVID-19, and respiratory syncytial virus (RSV) create a shared hospital-admission burden, because beds, staff, testing capacity, and respiratory support are allocated to total demand rather than to one pathogen at a time. This makes respiratory forecasting a target-construction problem as well as a modeling problem: disease-specific forecasts preserve attribution, whereas hospitals also need a single planning signal for combined burden. We developed a supervised 1- to 4-week-ahead framework for weekly admissions and compared two total-burden strategies: summing independently generated pathogen-specific forecasts and forecasting the combined admission target directly. Evaluation included retrospective held-out testing and prospective operational forecasts issued before the corresponding admission weeks were observed, with performance summarized using MAE and Percent Agreement (PA; the smaller of model-generated and observed counts divided by the larger). In retrospective total-burden evaluation, the combined-burden model had PA values of 0.85, 0.78, 0.74, and 0.76 and MAE values of 80.63, 200.09, 183.97, and 156.13 from week 1 to week 4 ahead, compared with summed-forecast PA values of 0.85, 0.80, 0.69, and 0.69 and MAE values of 100.14, 200.06, 205.65, and 217.57. Prospectively, the combined-burden model had higher PA and lower MAE at all four horizons, with the largest separations at weeks 3 and 4 ahead. These results support layered respiratory forecasting: pathogen-specific forecasts remain necessary for etiologic interpretation, while combined-burden forecasts better align with hospital preparedness and shared-capacity planning.

As active power dispatch becomes increasingly coupled with reactive power support and voltage security in modern power systems, AC security regions (AC-SRs) under exact steady-state AC power flow (AC-PF) constraints provide valuable insights into admissible operating boundaries and security margins for security assessment. However, the nonconvexity of exact AC-PF constraints makes it difficult to achieve both accuracy and computational efficiency in AC-SR characterization. To this end, this paper proposes an AC-aware iterative exclusion refinement (AC-ITER) approach that refines a given convex-relaxed AC-SR toward the exact AC-SR by progressively constructing and removing convexlocal exclusion subregions around numerically identified insecure points. Boundary-point searches and dual-based local exclusion cuts provide a tractable approximation of relaxation-induced insecure portions. A statistical certification method based on uniform sampling is developed to support high-confidence assessment of residual-region security and guide termination, while an AC-PF security similarity-based acceleration (AC-SSA) approach is introduced to further accelerate the sampling-based search for insecure operating points. Numerical studies on academic test systems and real-world systems demonstrate the characterization accuracy of AC-ITER and the computational efficiency gained through AC-SSA.

An exactly solvable lattice with periodic inverse-square singularities exhibits two physically distinct mechanisms of spectral-gap closure. The model arises from the generalized trigonometric Pöschl-Teller potential in the coupling range where both local solutions at each singularity are square-integrable. A transmitting, current-preserving self-adjoint matching condition connects neighboring cells and generates a nontrivial Bloch spectrum. For a generic lattice with two inequivalent singular sublattices, the spectrum contains two alternating families of gaps, and the band edges reproduce the scale-invariant spectra of the associated finite singular system. One gap family closes when the two sublattices become equivalent. In this symmetric limit, however, the primitive period halves, so the apparent band crossing results from Brillouin-zone folding. A different family closes at a critical but genuinely asymmetric configuration. Because the primitive period is preserved, adjacent bands then meet at the center of the original Brillouin zone in a genuine massless Dirac-like cone. Near either critical limit, a small finite gap acts as an effective Dirac mass: the dispersion remains quadratic near a band edge but crosses over to a massive Dirac-like form beyond that narrow region. The band curvatures define effective electron and hole masses. Exact Bloch eigenfunctions further reveal that isospectral lattices can display different single-particle probability densities. This model therefore provides an analytically controlled one-dimensional setting in which conventional quadratic carriers, massive Dirac-like band dispersions, and a tunable massless contact coexist within the same singular lattice.

Reliable deployment of LLM-based multi-agent systems calls for runtime auditors that predict terminal outcomes from observed execution prefixes. These auditors typically learn from a limited corpus of trajectories with verified outcomes. Expanding this corpus requires costly execution and verification, yet broader failure coverage alone provides no explicit criterion for identifying prefixes that the current auditor finds difficult to classify. To this end, we propose Dream2Discern, a runtime auditor trained through implicit augmentation guided by its own prediction loss. During Dream, we select perturbations of the auditor's representation operators to increase prediction loss on recorded prefixes, yielding hard latent variants. During Discern, the auditor learns to predict execution outcomes from these variants using the original labels. The variants are regenerated as the auditor learns, turning a limited corpus into a renewable source of hard examples and requiring no additional trajectory collection or outcome verification. Extensive evaluations on AFTraj-2K and StepShield show that Dream2Discern outperforms competitive failure auditors and achieves over 90% trajectory-level weighted accuracy on both benchmarks while balancing prediction accuracy and timeliness. Project page: https://zbox1005.github.io/dream2discern-project/

Frontier models are now trained for complex multi-agent tasks, from running a company to engaging in negotiations. These tasks necessitate training models to be persuasive. But what does this optimization pressure mean for model alignment? In this work, we study how reinforcement learning (RL) training for persuasion affects models' use of misaligned strategies such as evidence fabrication and belief coercion, and what techniques can reduce such misalignment. We find that optimizing sender models solely for receiver belief shifts causes the average number of distinct misaligned strategies used per game to increase by more than 25%, and fabrications per game increase to more than 4 times relative to pre-RL models. We compare various mitigation approaches, including stronger strategy guidance, clean demonstrations, penalizing misaligned strategy usage, and improving the model's ability to detect fabrications. We find that navigating the persuasiveness-misalignment trade-off in this setting is challenging, with the best trade-offs coming from combining guidance on aligned persuasion strategies with reward penalties on misalignment. Overall, our work provides a testbed for assessing misalignment stemming from RL-based persuasion training and shows concrete paths for reducing misalignment. Importantly, it highlights that without direct intervention, persuasion-involved RL environments are likely to naturally lead to large increases in the use of misaligned strategies.