Distributed Computing
AgentWeave: Efficient Distributed Agent Serving via Flow Decomposition
Kaibin Guo, Pengtu Li, Zicong Hong et al. (+4 more)
Agent applications are evolving into workflows composed of large
language model (LLM) calls and tool invocations, termed modules. Existing distributed
agent systems typically adopt a disaggregated architecture, deploying different modules on separate devices. While this enables pipeline parallelism, imbalanced module latencies introduce severe pipeline bubbles that limit throughput. A natural alternative is an aggregated architecture that co-deploys all modules across all devices, eliminating bubbles by keeping every device continuously active. However, this introduces request blocking: short requests are forced to progress in lockstep with longer ones in the same batch, and newly arriving requests must wait until the entire ongoing batch completes the full workflow. To address this, we propose AgentWeave, an efficient distributed serving system for
agent applications built on the aggregated architecture. Its core mechanism, flow decomposition, decomposes long
agent requests into fine-grained execution units while preserving
agent semantics, allowing short and newly arriving requests to proceed without being stalled by long-running ones. Complementing this, a flow-adaptive scheduler dynamically balances the throughput gains of decomposition against its scheduling overhead. Experiments across four representative
agent applications demonstrate that AgentWeave improves throughput by 1.81x, reduces average latency by 2.45x, and reduces P90 latency by 2.04x over state-of-the-art baselines.
Read the full abstract ▾
All authors Kaibin Guo, Pengtu Li, Zicong Hong, Zhiyuan Fang, Wuhui Chen, Rachid Guerraoui, Anne-Marie Kermarrec
Combinatorics
Combinatorics of degeneracy in vertical systems
Carles Checa and Nidhi Kaihnsa
We characterize the generic consistency of vertical systems in terms of the iterated
circuits of the column matroid of the coefficient
matrix. More precisely, we prove that such a system is generically consistent if and only if there exists an iterated
circuit such that the dimension of the Minkowski sum of the associated Newton polytopes is equal to the rank of the coefficient
matrix. We also give an equivalent characterization in terms of $G$-spanning flags of the dual matroid. With this, we derive a combinatorial formula for the codimension of the locus of parameter values for which the system has a solution, extending known results on dual defects of toric varieties. When this locus has codimension one, its defining equation yields a generalization of the sparse resultant to vertical systems, which applies also to $\mathcal{A}$-discriminants. Finally, we characterize when the solution set is contained, for generic parameter values, in a finite union of hyperplanes parallel to a coordinate hyperplane.
Read the full abstract ▾
Computer Vision
YOUDrive: Driving Style as a Composable Direction for End-to-End Autonomous Driving
Guantian Zheng, Lidong Yu, Tianyu Gao et al. (+3 more)
End-to-end driving policies can achieve strong task performance while offering limited control over their driving style. Driving style describes how a vehicle accelerates, brakes, and maintains distance. Policies trained without explicit preference conditioning lack a direct mechanism for adjusting these behaviors at inference. We propose YOUDrive, a
vision–
language–action planner for continuous and composable driving-style control. Each teacher's behavior is distilled into a task vector, a learned parameter update that induces a style direction in the planner's representation. These directions can be scaled and combined at inference without retraining the backbone. Conditioned on these directions, a flow-matching decoder models a
multimodal trajectory distribution and generates kinematically feasible motion alternatives. We introduce the Style
Alignment Score (SAS) to describe human-referenced kinematic behavior after safety and task gating. On NAVSIM navtest, a single coefficient monotonically controls assertiveness across six teacher-derived personas, while PDMS remains above 0.90 across the tested individual and blended configurations, reaching 0.954. Further analyses reveal systematic policy-specific kinematics after controlling for scene context, whereas the evaluated discrete style labels provide limited additional explanatory power. Code, checkpoints, and the metric will be released.
Read the full abstract ▾
All authors Guantian Zheng, Lidong Yu, Tianyu Gao, Jiaxing Chen, Jiashu Li, Chen Lv
Instrumentation and Methods for Astrophysics
Mid-IR Water-Vapor Seeing: Implications for ELT/METIS High-Contrast Imaging from VLTI Data
Abhirami S Raghu, Prashant Pathak, Olivier Absil et al. (+7 more)
Direct imaging and characterization of rocky
exoplanets remains one of the central challenges in exoplanetary science. The mid-infrared (mid-IR) offers a favorable high-contrast imaging (HCI) regime, as the
planet-
star contrast is more moderate than at shorter wavelengths. However, mid-IR observations with extremely large ground-based
telescopes (ELTs) are strongly affected by spatio-temporal fluctuations in atmospheric water vapor (WV), which introduce chromatic wavefront errors that are largely invisible to adaptive optics systems operating at shorter wavelengths. This poses a significant challenge for mid-IR HCI instruments such as METIS, which aims to detect thermal emission from rocky
exoplanets in the $N$-band (8–13 $\mu$ m). To quantify WV-induced wavefront errors, we analyze $K$-band fringe-tracker data from the
GRAVITY instrument at the VLTI and derive the optical path difference between group-delay and phase-delay measurements. This difference provides a direct measure of the differential WV column density ($\Sigma$) above the
telescopes. We examine the dependence of the RMS differential WV column density ($\sigma_{\Sigma}$) on precipitable water vapor (PWV), airmass, seeing, and wind speed across multiple VLTI configurations. Among these parameters, PWV shows the strongest correlation with $\sigma_{\Sigma}$. We further explore the influence of the vertical PWV distribution on this relationship. Next, we use power-spectral-density analysis to quantify how increasing PWV modulates $\Sigma$ variability. The wavefront error associated with this phenomenon on ELT scale is approximately 15~nm RMS in the $L$-band and 175~nm RMS in the $N$-band, relative to the $K$-band. These results provide a quantitative framework for assessing and mitigating atmospheric WV effects in next-generation mid-IR HCI instruments.
Read the full abstract ▾
All authors Abhirami S Raghu, Prashant Pathak, Olivier Absil, Benjamin Courtney-Barrer, Gilles Orban de Xivry, Julien Woillez, Alexis Matter, Philippe Bério, Roy van Boekel, Bernhard Brandl
Multiagent Systems
EAIR: Interaction Risks in Enterprise LLM Multi-Agent Systems
Yuchen Ma, Swadheen Shukla, Paulo Salem et al. (+5 more)
LLM
agents increasingly divide enterprise work across roles with different goals, authority, access, and information. Because each
agent often sees only a local snapshot of the enterprise workflow, locally reasonable actions can still violate organization-wide constraints (e.g., several service
agents can each issue credit that is individually permitted while collectively exceeding a shared compensation cap). We introduce EAIR, the first framework for analyzing enterprise multi-
agent interaction risks. We make four contributions. (1) We formalize enterprise interaction risk in terms of the organizational properties that workflow operations must preserve, and operationalize this formulation as a taxonomy of testable risks with explicit failure criteria and evidence requirements. (2) We build EAIR-Worlds, a suite of executable micro-worlds that instantiates selected risks as controlled, observable workflows with explicit roles, policies, permissions, source facts, and execution traces. (3) We develop a novel diagnosis-and-repair pipeline that uses the resulting execution traces to locate contributing interaction boundaries, guide targeted workflow repairs, and retest the repaired system. (4) Large-scale experiments show recurring interaction failures across the organizational properties studied. Together, EAIR is the first operational framework for diagnosing and mitigating enterprise multi-
agent interaction failures.
Read the full abstract ▾
All authors Yuchen Ma, Swadheen Shukla, Paulo Salem, Yue Huang, Robert Sim, Stefan Feuerriegel, Michel Galley, Jianfeng Gao
Artificial Intelligence
CESNET-TimeSeries26: A Multi-Year Network Traffic Time Series Dataset for Annual Seasonality, Concept Drift, and Long-Term Anomaly Detection
Josef Koumar, Karel Hynek, Kamil Jerabek et al. (+1 more)
Forecast-based models are fundamental to network traffic management and anomaly detection. However, it is essential to evaluate these
forecasting models over extended periods of time in order to assess their adaptability to unexpected phenomena and shifts in underlying patterns. This kind of long-term evaluation requires multi-year data to account for yearly seasonality and behavioral drift. Existing public
datasets are too short to support such long-horizon
forecasting tasks. We introduce CESNET-TimeSeries26, comprising 995 days of continuous traffic time series from the CESNET3 network, together with a single subnet spanning approximately 4.5 years. The
dataset captures 278,466 IP addresses, 285 institutions, and 566 institutional subnets at three aggregation intervals. The
dataset provides a definitive longitudinal
benchmark for evaluating
forecasting models in both operational management and security monitoring. We demonstrate that annual seasonality is the dominant periodic component. Moreover, we show that a
forecasting model can exploit the annual component only when its training window covers at least one complete annual cycle, which no sub-year
dataset provides.
Read the full abstract ▾
All authors Josef Koumar, Karel Hynek, Kamil Jerabek, Pavel Siska
Distributed Computing
Accelerating MoE Training through Expert Placement and Replica Coordination
Zhiqiang Tan, Maoxin Wang, Wenxiang Lin et al. (+4 more)
Distributed training of sparsely activated mixture-of-experts (MoE) models is often bottlenecked by workload imbalance and all-to-all (A2A) communication induced by dynamic
token routing. Existing
token deduplication, expert placement, and replication techniques alleviate these bottlenecks but do not jointly capture their distinct load semantics. Under hierarchical
token deduplication, communication counts each
token once per destination group at each hierarchy level, whereas expert computation processes every
token-expert assignment. With replication, both costs depend jointly on physical layout and
token-to-copy mapping, making separate optimization suboptimal. We present PlaceMoE to address this coupling. First, we formulate the joint optimization of expert layout and source-aware
token-to-copy mapping for E experts distributed across P devices, accounting for both hierarchical communication traffic and expert-compute load. Second, we develop a solver that combines co-selection-aware expert grouping, bounded dynamic programming for replica allocation, and alternating refinement of layout and mapping. Third, PlaceMoE overlaps replica-gradient synchronization with
attention backward. We evaluate PlaceMoE on a 32-GPU cluster and a 64-NPU cluster using Qwen3-VL and DeepSeek-V3. Experimental results show that PlaceMoE achieves up to a 6.94x speedup in A2A communication and 1.74-2.33x speedups in end-to-end training over the state-of-the-art training framework VeOmni.
Read the full abstract ▾
All authors Zhiqiang Tan, Maoxin Wang, Wenxiang Lin, Sijie Wang, Qiang Wang, Shaohuai Shi, Bo Li
Analysis of PDEs
The CR Nirenberg Problem, Part I: Blow-up Analysis and Compactness
Jiechen Qiang, Zhongwei Tang and Heming Wang
We establish a priori estimates and compactness results for positive solutions of the CR Nirenberg problem under natural weighted flatness conditions. The proofs combine refined blow-up analysis, Pohozaev identities, and the complete classification of positive entire solutions to the critical CR Yamabe equation recently obtained by Li–Liu–Ma \cite{LLM2026} and Li \cite{L2026}. Existence and degree results are established in our subsequent paper \cite{QTWPartII}.
Read the full abstract ▾
Atomic Physics
Inverse identification of thermal-radiative parameters for reduced effective-temperature uncertainty in an ion optical clock
Jie Zhao, Yani Zuo, Shaoyang Dai et al. (+8 more)
The blackbody-radiation (BBR) shift remains a leading systematic uncertainty of room-temperature single-ion optical clocks. Accurate evaluation of this shift requires the effective temperature at the ion position, which is difficult because RF heating, low and spatially varying surface emissivity and inaccessible trap surfaces make this quantity difficult to measure directly. Here, we developed a hybrid inversion framework that combines radiance-domain infrared thermography, black-painted high-emissivity reference regions, a finite-element model of a 171Yb+ end-cap ion trap and a reduced-order surrogate for repeated forward calculations and parameter identification. The surrogate reproduced the FE test fields with a full-field mean absolute error (MAE) of 0.02220 mK. The calibrated model predicted the ion-position effective temperature with a combined standard uncertainty of 155.0 mK, corresponding to a temperature-related BBR frequency-shift uncertainty of 1.647 x 10^-19 for the 2S1/2-2F7/2 (467 nm) E3 transition. This method provides a reproducible route for reducing and tracing the BBR uncertainty of room-temperature trapped-ion optical clocks.
Read the full abstract ▾
All authors Jie Zhao, Yani Zuo, Shaoyang Dai, Chaowei Wang, Ang Huang, Jian Song, Guangyao Wang, Fangfei Wu, Haotian Shi, Zhengyong Li, Fang Fang
Cryptography and Security
Protecting Shared Agent Workflows from Invalidation Attacks with Signed Trajectories and Runtime Verification
Ruixiao Li, Fahao Chen and Peng Li
Large
language model (LLM)-based
agents increasingly rely on shared workflow repositories to transfer experience across tasks and
agents. By abstracting past executions into reusable workflow graphs, such systems significantly improve planning efficiency and generalization. However, this emerging paradigm introduces a new and underexplored security risk: if the shared repository is compromised, tampered workflows can systematically mislead downstream
agents, inducing persistent execution failures or unsafe behaviors. In this work, we present a principled protection framework for shared
agent workflows based on signed trajectories and runtime verification. We require all cloud-collected trajectories used for workflow generation to be cryptographically signed, with each trajectory representing a concrete execution instance. Prior to repository inclusion, trajectories undergo a unified screening process to ensure correctness and validity. When an
agent retrieves a workflow from the shared repository, it also obtains a small set of signed supporting trajectories and verifies both their cryptographic integrity and their semantic consistency with the workflow. To make this practical for branching workflows, we further propose a runtime, path-aware
retrieval strategy that dynamically fetches only the signed trajectories relevant to the execution path actually taken by the
agent, substantially reducing communication and verification overhead. Extensive experiments across multiple
reasoning benchmarks show that workflow invalidation attacks achieve up to 90.12% success rates, degrading
agent inference performance by up to 73.16 percentage points in accuracy without effective workflow protection. In addition, the proposed protection achieves an average F1 score of 85.26% and a detection rate of 79.97%, outperforming existing baselines by about 2.02x.
Read the full abstract ▾
Dynamical Systems
On the Julia sets of the non-autonomous iteration of exponential maps
Jiaxing Huang and Xiao Yao
This paper studies non-autonomous iterations of $\lambda e^z$, where at each step the parameter $\lambda$ is chosen from a bounded interval. Urbański and Zdunik introduced and developed a thermodynamic formalism for random dynamics of exponential maps, extending the autonomous result established independently by Rees and Lyubich to typical parameter sequences in $[A, B]^{\mathbb{N}}$ with $1/e<A<B<\infty$. In this paper, we show that the same conclusion indeed holds for all parameters in $[A, B]^{\mathbb{N}}$, Our proof is based on carefully crafted uniform versions of pullback arguments, inspired by the original methods of Rees and Lyubich. In a related direction, Urbański and Zdunik established a non-autonomous analogue of Misiurewicz's
theorem for the exponential map, later refined by Lech. We further extend their work by providing a refined condition for the Julia set to be the entire complex plane.
Read the full abstract ▾
Networking
Energy-Aware Agentic Orchestration for Multi-Site Service Deployment
Rodrigo Moreira, Alisson Chaves, Larissa F. Rodrigues Moreira et al. (+6 more)
Energy-aware orchestration across geographically distributed infrastructures is increasingly important as data centers absorb growing Artificial Intelligence (AI) workloads. However, existing approaches often exclude the energy cost of automated decision-making from system-level sustainability assessments. This paper presents an energy-aware agentic orchestration framework that treats Large
Language Model (LLM)
reasoning as a measurable component of the control loop. We compare single-shot tool calling, ReAct, and plan-and-execute strategies on a multi-site Kubernetes testbed spanning two European data centers and instrumented at layers ranging from power distribution units to per-pod energy attribution. Using ten natural-
language orchestration intents with varying decision complexity, we show that
reasoning strategy and intent complexity strongly influence graphics processing unit (GPU) energy consumption during orchestration decisions. ReAct and plan-and-execute generated executable tool calls across all evaluated intents but required multiple LLM inferences, whereas the single-shot configuration was less costly but non-actionable. Prefix caching reduced mean ReAct GPU energy consumption by 29.4%. For the evaluated lightweight workload, the infrastructure-energy changes observed within the one-second measurement window were insufficient to offset the measured GPU decision energy. These findings demonstrate the importance of accounting for
reasoning overhead when assessing the sustainability of AI-native orchestration and motivate energy-budgeted selection of
agent strategies.
Read the full abstract ▾
All authors Rodrigo Moreira, Alisson Chaves, Larissa F. Rodrigues Moreira, Tereza Carvalho, Tiago Barros, Daniel Corujo, Rui Aguiar, David Gomez-Barquero, Flávio de Oliveira Silva
Artificial Intelligence
GREAT-ESDD: Graph Edge Attention Network for Effectively Steered Delivery Decisions
Attila Lischka, Ze Zhou, Xuegang Ban et al. (+1 more)
Online, same-day delivery requires operators to constantly update their routes, and the routing cost for deliveries increases with the number of customers. Out-of-home (OOH) delivery options can help companies handle increasing customer numbers while keeping delivery costs low. We develop a dynamic pricing-and-routing framework for this online ordering scenario with home and OOH delivery options. First, we develop a neural solver for the capacitated vehicle routing problem (CVRP) based on the Graph-Edge-
Attention-Network (GREAT). We train it to solve CVRP with varying instance sizes and starting conditions, allowing it to reoptimize partially executed routes under changing conditions. It is fast while achieving solution quality competitive with traditional baselines. Our solver processes several CVRP instances in parallel in batches and we use it to solve the candidate routing problems that arise from an incoming customer's choice of home or OOH delivery. This provides an estimate of the delivery cost increase per option, which we use to set dynamic prices. Further, it directly provides the updated route once the customer decides. We trade off delivery fees against routing costs and steer customers toward OOH delivery when home delivery would be costly. However, effective steering requires us to know the customer preferences. To learn those online, we use Thompson sampling with a Laplace-approximated posterior. We call the resulting framework GREAT-ESDD (GREAT for Effectively Steered Delivery Decisions). Compared to a traditional routing solver, it increases profit by around 11% under a similar pricing decision time budget. Moreover, we show that we can estimate customer preferences successfully, outperforming a static pricing strategy after only 5 days of simulated operations.
Read the full abstract ▾
All authors Attila Lischka, Ze Zhou, Xuegang Ban, Balázs Kulcsár
Earth and Planetary Astrophysics
Milani and Juventas Optical Navigation Images in Support of the Hera Radio Science Experiment
Edoardo Gramigna, Ozgur Karatekin, Birgit Ritter et al. (+7 more)
The European Space Agency's Hera mission aims to advance planetary defense by investigating the Didymos binary asteroid system following NASA's DART impact on Dimorphos. Incorporating simulated CubeSat images into Hera's orbit determination framework significantly improves positional accuracy, reducing uncertainties by up to three orders of magnitude for Milani and one for Juventas. This data enable more precise estimation of key parameters such as asteroid masses, rotational states, and moments of inertia, with up to 90% uncertainty reduction in Dimorphos' dynamical state under chaotic scenarios. Sensitivity analysis reveals the influence of pointing accuracy, image noise, and imaging frequency. Additionally, dedicated scientific investigations using Dimorphos-focused CubeSat images further advance understanding of the system. These findings underscore the critical role of CubeSat images in refining
navigation and advancing science, offering valuable insights for future small-body missions.
Read the full abstract ▾
All authors Edoardo Gramigna, Ozgur Karatekin, Birgit Ritter, Iosto Fodde, Fabio Ferrari, Felice Piccolo, Paolo Tortora, Riccardo Lasagni Manghi, Marco Zannoni, Ryan S. Park
Artificial Intelligence
Counterfactual Regret for Reflective Agents with Lean-Verified Guarantees
Zhongnian Li, Shi Xu, Mengting Xu et al. (+2 more)
Reflective
language agents reuse feedback and
memory to adapt across tasks without updating model weights. However, existing guarantees for policy selection and action generation do not by themselves establish when local reflection yields long-term performance guarantees for the complete
agent under resource constraints. In this paper, we address this gap by deriving sufficient conditions for sublinear counterfactual regret from local generation and deployment guarantees, while retaining the native controller. Specifically, we establish a resource-sensitive transfer that bounds full-state deployment discrepancies along actual execution paths and preserves the own-history laws of both learner and comparator. Besides, local generation reliability yields a cumulative generation opportunity bound, linking policy acquisition to long-term performance through an explicit quality, coverage, and hard-budget tradeoff. To verify these guarantees, we formalize the conditional proof chain for all thirteen numbered theoretical results in Lean, including a native instance with repeated policy growth and imperfect validation.
Read the full abstract ▾
All authors Zhongnian Li, Shi Xu, Mengting Xu, Peng Ying, Xinzheng Xu
Computation and Language
Training-Free BioNEL-RAG: Advancing Biomedical Entity Linking through Retrieval-Augmented Generation and Large Language Model
Ziyi Chen, Hui Liu, Ziyang Ling et al. (+2 more)
Biomedical Named Entity Linking (BioNEL) is the task of mapping mentions extracted from biomedical texts to canonical entities or concepts in the standardized knowledge base. Although existing methods have achieved remarkable performance in mapping accuracy, most rely on training with domain-specific
datasets, which fundamentally limits their cross-domain adaptability. To address this limitation, we propose BioNEL-RAG, a novel training-free framework that combines
retrieval-augmented generation (RAG) with large
language models (LLMs). Within this framework, the RAG module focuses on retrieving candidate entities, while the large
language model module plays the role of selecting the most appropriate candidate for entity normalization, which is based on the analysis of the mention's context, with the aid of its remarkable text comprehension capabilities. Experimental results demonstrate that BioNEL-RAG achieves performance comparable to existing methods without requiring training on subdomain
datasets, whereas these methods necessitate fine-tuning on specific subdomains such as genes, diseases, or chemical compounds.
Read the full abstract ▾
All authors Ziyi Chen, Hui Liu, Ziyang Ling, Ronald X.Xu, Mingzhai Sun
Quantum Physics
Entanglement generation from degeneracy breaking of superradiant Liouvillian steady states by local operations
Adam Burgess and Erik M. Gauger
Superradiance is a collective optical effect where emitters cooperatively radiate with a maximal intensity scaling proportional to the square of the number of emitters. Despite the superradiant cascade transitioning through highly entangled states, it has been shown that, for dynamics along the Dicke ladder, superradiance need not generate
entanglement between the emitters. By contrast, we show how introducing incoherent local processes – normally considered detrimental to both
entanglement and superradiance – can generate high levels of inter-emitter
entanglement. We develop a general framework, using degenerate non-Hermitian perturbation theory, for the study of superradiant Liouvillian steady-state degeneracy breaking by these local operations from a few emitters up into the thermodynamic limit. From this, we are able to show that
entanglement becomes ubiquitous in this regime and demonstrate its robustness against other forms of symmetric environmental noise. In the regime where pumping dominates, we find that population accumulates towards the maximally entangled minimal pseudo-
spin states, corresponding to multi-emitter singlet states. This provides a deeper insight into many-body
entanglement and the curious connection between local operations and
entanglement generation.
Read the full abstract ▾
Machine Learning
Formalizing Forgetting: Theory of Forgetting for Online Continual Learning
Renjie Cao, Yifei Cheng and Li Shen
Online continual learning (OCL) requires a learner to acquire a sequence of tasks in a single pass over the data, and catastrophic forgetting of earlier tasks remains its central challenge. Despite the many methods proposed to mitigate forgetting, the theoretical understanding of forgetting in OCL is still limited: existing analyses mostly focus on linear, overparameterized, or neural tangent kernel regimes, and it remains unclear which factors govern forgetting and how they affect it. To fill this gap, we analyze single-pass Shuffle SGD under strongly convex and general convex objectives and derive explicit upper bounds on the expected final forgetting. Further analysis of these bounds shows how the per-task data amount, task similarity, and task ordering enter them, and that their effects can differ across objective classes. Moreover, experiments with neural networks on Split-CIFAR-100 and PMNIST show that the trends suggested by our theory also appear in practice, indicating that these insights extend beyond the convex setting.
Read the full abstract ▾
Analysis of PDEs
Global Well-Posedness and Uniform Long-Time Control for a Class of Periodic NLS-Dysthe Equations
Andrea Belloni and Filippo Giuliani
We study a class of periodic NLS-Dysthe (NLS-Dy) equations, also referred to as higher-order nonlinear Schrödinger equations (HNLS), of the form $i\partial_t u+a\partial_x^2u+b|u|^2u+i\varepsilon(c\partial_x^3u+d|u|^2\partial_xu+eu^2\partial_x\overline{u})=0$ with $ x\in\mathbb T$, $a,b,c,d,e\in\mathbb R$, $abc\neq0$, and $\varepsilon>0$. For $0<\varepsilon\ll1$, equations of this type arise as higher-order modulation models for some nonlinear dispersive PDEs, including the Klein-Gordon equation and the Serre-Green-Naghdi equation modeling the
evolution of nonlinear shallow water waves. We prove local well-posedness in $H^s(\mathbb T)$ for $s>3/2$ for arbitrary admissible coefficients, and establish an $H^1$ continuation criterion for $s\ge2$. In the parameter regime $e\neq0$, a lower-order conservation law provides uniform $H^1$ control and, together with the continuation criterion, yields global well-posedness for $s\ge2$. This result does not require the non-resonance condition appearing in Takaoka's periodic theory, while the condition $e\neq0$ naturally arises in the modulation setting considered here. Finally, motivated by the long-time regime relevant for modulation, we establish uniform long-time control without imposing a smallness assumption on the initial datum. Using the conservation laws of cubic NLS and their invariance under the commuting complex mKdV flow, we obtain uniform $H^s$ bounds on time intervals of order $\varepsilon^{-1}$ for integer $s\ge2$. For non-integer real $s\ge2$, the maximal $H^s$ solution exists on the same time scale, while uniform control is obtained at a lower regularity, with a loss of less than one derivative.
Read the full abstract ▾
Quantum Physics
Classical Communication Helps Single-Shot Quantum Compression
Omri Farber and Gilad Gour
Free classical communication can yield an unbounded single-shot saving in
quantum compression, even though it leaves the asymptotic rate of an i.i.d. source unchanged. At every fixed error in $(0,1)$, we exhibit sources whose unassisted compression cost is arbitrarily large, yet a single
qubit suffices with classical communication and no preshared
entanglement. We establish this separation by determining the exact optimal source fidelity for compression through an $m$-dimensional noiseless
quantum channel, with and without free classical communication, while preserving correlations with an inaccessible reference. Without classical assistance, the optimal fidelity is exactly the Ky Fan norm $\|\rho\|_{(m)}$, the sum of the $m$ largest
eigenvalues of the source state. With classical assistance, we obtain an explicit spectral threshold formula and a protocol attaining it. Classical communication strictly improves fidelity whenever $m$ is smaller than the source rank. The corresponding compression costs are two smoothed versions of the log-rank
entropy, also called max-
entropy: trace-distance smoothing under local operations and quadratic smoothing under LOCC. For i.i.d. copies of a fixed source with positive
entropy variance, however, the two costs have identical first-, second-, and third-order asymptotic terms. For nonlattice spectra, we determine the constant-order advantage: classical communication saves $\log_2(e)-1$ bits per block asymptotically, and the ratio of assisted to unassisted minimum communication dimensions converges to $2/e$.
Read the full abstract ▾
Earth and Planetary Astrophysics
Hera Orbit Determination Covariance Analysis of Didymos Binary Asteroid with the Full Two-Body Problem
Edoardo Gramigna, Riccardo Lasagni Manghi, Marco Zannoni et al. (+7 more)
Standard planetary
radio science analyses may be inadequate for accurately interpreting data from the Hera mission at the Didymos system due to the anticipated challenge posed by the coupling between the translational and rotational motions of the asteroids. A more precise dynamical model, the full two-body problem model, which includes a mutual gravitational potential, may be required to fit the Hera data properly. This work presents a covariance analysis of the Hera
radio science experiment using this advanced model. The tested dynamical scenarios are based on recent lightcurve observations indicating that Dimorphos is in an excited non-principal axis
spin state that is, on average, tidally locked. Three non-principal axis scenarios are analyzed using a multi-arc local estimation approach instead of the global one adopted in previous investigations involving non-chaotic states. The results show that tumbling or chaotic cases significantly improve the recovery of the bodies' moments of inertia due to the enhanced observability of the system's dynamics from the excited
spin state. However, the local estimation approach slightly degrades the accuracy of the mass estimation by a factor of 1.3-3 and Dimorphos' relative orbit by about one order of magnitude. Developing a dedicated constrained multi-arc filter is thus required to improve the estimation accuracy of the Didymos system's dynamical state further. Nevertheless, these results still meet the Hera
radio science requirements.
Read the full abstract ▾
All authors Edoardo Gramigna, Riccardo Lasagni Manghi, Marco Zannoni, Paolo Tortora, Michalis Gaitanas, Ioannis Gkolias, Kleomenis Tsiganis, Ryan S. Park, Alex Meyer, Petr Pravec
General Topology
Graphs on Topological Spaces
Peter J. Cameron and Vinod Kumar P. B
Graphs defined on groups, such as power graphs, enhanced power graphs and commuting graphs, form a hierarchy under inclusion of edge sets. We develop an analogous theory for topological spaces. To a space X we attach five graphs on the vertex set X, all invariant under homeomorphisms: the specialization, inseparability, closure-meeting, Urysohn and functional graphs. They form a diamond-shaped hierarchy whose members are null exactly when X is T_1, Hausdorff, Urysohn or functionally Hausdorff. The Urysohn graph always contains the square of the inseparability graph, with equality for Alexandrov spaces, and the hierarchy is compatible with products via the strong product of graphs. Every countable graph, together with isolated vertices, is the inseparability graph of a second countable T_1 space. For finite spaces we characterize the graphs that occur as inseparability graphs (every edge lies in the closed neighbourhood of a simplicial vertex), determine all coincidences between members of the hierarchy, and show that the specialization graph determines the weak homotopy type but not the space itself. Using Dowker duality we prove that the complexes of inseparable and of closure-meeting families are always homotopy equivalent and, under an acyclicity hypothesis, recover the weak homotopy type. We show that if the closure-meeting graph is a cograph then it is chordal, an analogue of a recent result for enhanced power graphs; the same holds for the inseparability graph of an Alexandrov space but not in general. Finally, following Cameron's work on
topologies preserved by permutation groups, we show that for a
topology invariant under a primitive group each graph is null or connected, and that the neighbourhood filters of the five graphs reverse the hierarchy, giving spaces whose specialization graph contains the Rado graph as a spanning subgraph.
Read the full abstract ▾
Systems and Control
Decision-Focused Learning for Optimal Battery Scheduling Under Load and Photovoltaic Power Uncertainty in Microgrids
Marios Saleptsis, Antoine Lesage-Landry, Marco Mussetta et al. (+1 more)
The increasing penetration of renewable energy sources in microgrids demands accurate day-ahead
forecasting to support optimal battery energy storage system (BESS) scheduling. Yet the conventional predict-then-optimize paradigm trains forecasters on statistical accuracy metrics that are agnostic to the asymmetric and time-varying cost impacts of forecast errors. This work proposes a decision-focused learning (DFL) framework that trains gated recurrent unit forecasters for both load demand and photovoltaic generation directly on the downstream operational cost of a two-stage energy management system, where day-ahead BESS commitments are fixed prior to real-time balancing. Gradients are backpropagated from the real-time operational cost through both optimization stages to the forecaster weights via implicit differentiation, enabling the forecasters to internalize the penalty structure of the scheduling problem. The framework is validated on the MG2Lab experimental microgrid testbed at Politecnico di Milano across two operational scenarios of increasing grid constraint and eight
forecasting configurations comprising single-source and combined-forecaster settings. Results illustrate that the DFL-trained pipeline reduces operational cost by 3.78% under normal grid conditions and by 30.60% under grid-constrained conditions relative to the accuracy- optimized baseline, while nearly eliminating load shedding events. Notably, the DFL forecasters exhibit higher statistical errors than their accuracy-trained counterparts yet achieve consistently lower operational costs, confirming that forecast accuracy and decision quality do not share a monotonic relationship in constrained microgrid scheduling.
Read the full abstract ▾
All authors Marios Saleptsis, Antoine Lesage-Landry, Marco Mussetta, Sonia Leva
Cosmology and Nongalactic Astrophysics
TDCOSMO 2026: A 3% measurement of the Hubble constant from strong lensing and stellar kinematics
TDCOSMO Collaboration, Simon Birrer, Michele Cappellari et al. (+26 more)
We present cosmological constraints from 11 strongly lensed
quasars with measured time delays (hereafter, the TDCOSMO-2026 sample). In addition to expanding the sample, we improve many aspects of our blind analysis with respect to our 2025 analysis. Among other improvements, spatially resolved kinematics are available for nine systems, including six with JWST-NIRSpec, compared to one in 2025; for three lens models, JWST-NIRCam images now replace HST; axisymmetric dynamical modeling is applied directly to the spatially resolved data, instead of an axisymmetric correction factor applied to spherical models; and the line-of-sight convergence is now based on the Euclid Flagship simulations instead of the Millennium Simulation. Our baseline and primary result is based on the TDCOSMO-2026 sample, in combination with a prior on $\Omega_{\rm m}$ from the DES-Dovekie
supernova sample. In the flat $\Lambda$ cold dark matter ($\Lambda$CDM) cosmology, we find $H_0=71.1^{+2.4}_{-2.3}$ km.s$^{-1}$.Mpc$^{-1}$, a 3.3% uncertainty that accounts for the mass-sheet degeneracy. The SLACS and SL2S samples are in excellent agreement with the TDCOSMO-2026 sample, improving the precision on $H_0$ in flat $\Lambda$CDM to 3.0%. Using the DESI-DR2
baryon acoustic oscillations (BAO) likelihood instead of DES-Dovekie yields fully consistent results. We also present constraints in the open $\Lambda$CDM, $w$CDM, $w_0w_a$CDM, and $w_{\phi}$CDM cosmologies; the TDCOSMO $H_0$ inference is robust and consistent across all of them. The combination of the TDCOSMO+SLACS+SL2S likelihood with those from DESI-DR2 and DES-Dovekie disfavors $w=-1$ at the $2.3$-$2.5\sigma$ level in all three dark energy extensions that we tested. We find, e.g., $w_0=-0.88^{+0.04}_{-0.05}$ in $w_{\phi}$CDM and $w_0 = -0.85^{+0.07}_{-0.06}$ in $w_0w_a$CDM, while $w_a = -0.48^{+0.46}_{-0.44}$ remains consistent with zero.
Read the full abstract ▾
All authors TDCOSMO Collaboration, Simon Birrer, Michele Cappellari, Frédéric Courbin, Christopher D. Fassnacht, Pierre Fleury, Joshua A. Frieman, Aymeric Galan, Xiang-Yu Huang, Daniel P. Johnson, Sanjana Khanolkar, Shawn Knabel, Hrvoje Krizic, Martin Millon, Takahiro Morishita, Veronica Motta, Pritom Mozumdar, Eric Paic, Hadrien Paugnat, Allan G. Schweinfurth, Anowar J. Shajib, William Sheu, Dominique Sluse, Massimo Stiavelli, Sherry H. Suyu, Tommaso Treu, Felipe Urcelay, Devon M. Williams, Kenneth C. Wong
Logic in Computer Science
StructHCP: Structure-Aware SAT–CEGAR for the Hamiltonian Cycle Problem
Quan Van Vi, Tuyen Van Kieu and Khanh Van To
Many exact
Hamiltonian-cycle solvers use the same edge-level representation for every input, even when graph structure admits a smaller set of global decisions. We present StructHCP, a SAT–CEGAR pipeline that selects a representation through inexpensive structural recognizers. Specialized searches handle high-degree clusters and paired chains, while complete directed and undirected encodings handle cubic and general graphs. Every returned cycle is reconstructed when necessary and verified in the input graph. On the 1,001-instance FHCP Challenge Set, StructHCP solves 826 instances, compared with 741 when the same general SAT–CEGAR engine is applied directly to every unreduced input graph. Relative to this ablation, StructHCP reduces PAR-2 by 32.9 percent. Formulation selection solves 82 more instances than the contracted general encoding; high-degree-cluster and cubic graphs account for 81 of this difference. StructHCP also solves 34 of the 51 instances solved by neither HCup nor Cut-set CEGAR, raising the combined coverage of the three solvers from 950 to 984. Twelve of its solutions lie outside the union of published per-instance SAT results analyzed here, and six also lie outside the published exact result sets in our comparison. These results show that selecting the representation can complement connectivity refinement on structured HCP instances.
Read the full abstract ▾
Functional Analysis
Decomposition of functions of bounded variation in Ahlfors-regular PI spaces
Paolo Bonicatto, Panu Lahti and Enrico Pasqualetto
We prove that, on a non-compact Ahlfors-regular PI space, every function of bounded variation $f$ can be decomposed as a countable sum $\sum_{n=1}^\infty f_n$ of monotone functions of bounded variation $f_n$, in such a way that the total variation of $f$ is the sum of the total variations of the functions $f_n$. Our result extends a Euclidean result of Bianchini and Tonon to the metric setting.
Read the full abstract ▾
Quantum Physics
Avoiding Defects within Heavy-Hexagon Architecture
Hansol Kim and Younghun Kwon
When implementing scalable fault-tolerant
quantum computing using superconducting
quantum processors, one of the critical challenges to overcome is the presence of defective
qubits arising from fabrication and calibration imperfections. A defect on a specific
qubit severely restricts the connectivity between adjacent
qubits, disrupts stabilizer measurements, induces frequent detection events, and consequently inflates the logical error rate. To mitigate this issue, various strategies have been proposed, such as syndrome extraction
circuits based on stabilizer folding and the construction of super-stabilizers via gauge operators. However, these methods have primarily been developed for architectures with relatively high connectivity of 3 to 4. Furthermore, experimental demonstrations showing an actual reduction in logical error rates in the presence of physical hardware defects have been lacking. In this work, we propose a defect-avoidance scheme tailored for the heavy-hexagon architecture, which features a lower connectivity of 2 to 3. While previous approaches required four measurement cycles to measure all gauge operators comprising a super-stabilizer, our proposed scheme completes all gauge measurements within only two cycles. We experimentally validate this approach on IBM
quantum hardware by identifying physical defects and implementing the defect-avoidance
circuits. As a result, avoiding the defect yields an approximately 10% reduction in the logical error rate per cycle for both logical $X$ and $Z$ errors. These findings demonstrate that defects in superconducting hardware can be successfully handled even under the constrained connectivity of heavy-hexagon
lattices, and that defect avoidance substantially reduces logical errors on real
quantum hardware.
Read the full abstract ▾
Artificial Intelligence
LYNX: Analytical Program Synthesis via Grounding and Closed-Loop Refinement
Ruichen Xia, Xiao Han, Dingyu Yang et al. (+5 more)
LLM-powered data analytics aims to translate natural
language instructions into executable analytical programs over underlying data. However, existing approaches, whether treating LLMs as intent-to-code translators or as standalone analytical engines, often fail to reliably align domain semantics with complex schemas and heterogeneous metadata, resulting in brittle execution and semantically inconsistent analytical outputs in real-world workloads. We present LYNX, a multi-
agent framework that formulates LLM-driven analytics as a schema- and knowledge-constrained analysis problem rather than a single-pass generation task. LYNX organizes analytical synthesis into a closed-loop workflow of planning, generation, verification, and refinement through three role-specialized
agents that collaboratively enforce schema grounding, semantic consistency, and execution correctness. To address domain ambiguity and schema misalignment, LYNX integrates Semantic-Structural
Distillation and Stage-Aware Allocation (SSD-SAA). A delta-guided iterative revision continuously diagnoses and repairs semantic and execution-level failures, substantially improving synthesis robustness. We evaluate LYNX on two NL-to-SQL
benchmarks, BIRD-SQL Mini-Dev and the more challenging Spider 2.0-Lite. Experimental results demonstrate strong performance and efficiency, achieving 10.66% to 59.54% execution accuracy across lightweight and commercial models, while reducing
token consumption by 70.53% compared with the prior most efficient method. Our framework displays high generalization over different backbones, from open-source, lightweight models to the latest commercial models. Additional case studies on quantitative finance analytics and complex multi-relational SQL workloads further demonstrate the framework's robustness, scalability, and practical applicability in real-world analytical scenarios.
Read the full abstract ▾
All authors Ruichen Xia, Xiao Han, Dingyu Yang, Luwei Yang, Sai Wu, Dongxiang Zhang, Binhui Yao, Gang Chen
Sound
Training a Compact, Strong, and Realistic Text-to-Speech Model
Runyuan Cai, Yu Lin, Yiming Wang et al. (+1 more)
Text-to-
speech (TTS) systems are conventionally optimized for intelligibility and speaker similarity. However, neither objective captures the prosodic detail that distinguishes human performances from synthetic
speech, leaving a realism gap that larger models and higher-quality data can narrow but not close. In the absence of an objective that directly targets this gap, realism remains a persistent challenge in TTS. We present Audio8-TTS, a compact 0.6B-parameter model for zero-shot voice cloning that achieves state-of-the-art performance in intelligibility and speaker similarity, with particularly strong results in perceptual naturalness. To align the training signal with human perception rather than signal-level fidelity, we construct a quality ladder from widely used open-source TTS systems and train a realism critic to assess generated
speech against matched human recordings of the same sentences. An identity-erasure objective discourages the critic from relying on generator fingerprints. We train the model in three stages:
multilingual supervised fine-tuning; group relative policy optimization (GRPO) with word error rate (WER) and speaker similarity (SIM) rewards; and GRPO guided by the realism critic. The final stage improves perceived prosodic naturalness and speaker similarity while preserving WER performance.
Read the full abstract ▾
All authors Runyuan Cai, Yu Lin, Yiming Wang, Xiaodong Zeng
Superconductivity
Driven Vortex Motion in Periodic Channel Potential Embedded in Superconducting Films
A. Al Luhaibi, A. Glatz, Michael Vogl et al. (+2 more)
In this paper, we investigate
vortex dynamics in periodically width-modulated superconducting channels by numerically solving the time-dependent Ginzburg-Landau (TDGL) equations. The channels are designed with sinusoidally varying widths characterized by their average width and modulation amplitude. The motion of driven
vortices generates an oscillatory electric field with a fundamental frequency, $f_0$, set by the channel periodicity, together with higher-order harmonics. In contrast to the situation of
vortex lattice dynamics in a periodic pinning landscape of circular holes considered by Al Luhaibi et al. [1], we show that the harmonic content of the generated electric field can be effectively controlled through the channel geometry. In particular, appropriately engineered channel profiles strongly suppress higher-order harmonics, yielding an electric-field response that closely approaches a pure sinusoidal waveform. These results demonstrate that
geometric modulation offers a viable route to engineering low-harmonicity microwave signals in superconducting devices and may enable cryogenic on-chip waveform generators for controlling superconducting
qubits. Such devices could provide a useful building block for scalable transmon-based
quantum computing architectures.
Read the full abstract ▾
All authors A. Al Luhaibi, A. Glatz, Michael Vogl, Hocine Bahlouli, J. B. Ketterson