Multi-Hop RIS ISAC for Target Positioning: A Tensor Decomposition-based Approach

Reconfigurable intelligent surface (RIS) has demon- strated remarkable potential to enhance the performance of integrated sensing and communication (ISAC), particularly when the line-of-sight (LoS) paths are obstructed. By controlling the reconfigurable elements on the surface, RIS can establish virtual LoS paths and provide considerable passive beamforming gains, thereby significantly improving the received signal quality. In this paper, we design a novel multi-hop RIS ISAC system for target positioning, where multiple RISs are deployed to assist the communication from a transmitter to associated users while simultaneously enhancing receiver sensing performance in target positioning. Specifically, we formulate an optimization problem to minimize the root mean square error (RMSE) of the target detection while guaranteeing the communication requirements of the users. To solve this problem, we first unfold the cascaded sens- ing channel through parallel factor decomposition, and develop a low-rank CANDECOMP/PARAFAC decomposition (CPD)-based scheme to extract the location parameters (i.e., angle of arrival, angle of departure and delay) of the sensing targets. Then, we develop a scheme for jointly selecting the transmit beamforming and RIS phase shift configurations to maximize the sensing energy at the receiver, which in turn leads to improved accuracy in target positioning. We also provide a uniqueness analysis, complexity analysis, and Cramér-Rao lower bound (CRLB) of the parameters estimated by our methodology. Simulation results validate the improvement in target positioning obtained by our design relative to baselines.

Baseband-Free Wideband MU-MIMO OFDM via Stacked Intelligent Metasurfaces

This paper proposes a novel baseband-free wideband MU-MIMO OFDM transmitter architecture enabled by two cascaded stacked intelligent metasurfaces (SIMs). Unlike conventional wireless transmitters, the proposed design shifts symbol loading, MU-MIMO precoding, and OFDM modulation from digital baseband to the wave domain. Specifically, in the first SIM (SIM$_1$), a programmable symbol-loading interface first loads one stacked virtual-subcarrier coefficient vector onto a common monochromatic carrier, while the remaining layers perform channel-adaptive MU-MIMO precoding from the data streams to the port-subcarrier domain. Then, the second SIM (SIM$_2$) realizes offline-configured wave-domain OFDM modulation by implementing the inverse discrete Fourier transform (IDFT) and cyclic-prefix (CP) insertion operator directly in the wave domain. Based on this architecture, we develop a wave-domain system model that explicitly characterizes the virtual-to-physical transition from block-level coefficients to radiated OFDM subcarriers. The designs of SIM$_1$ and SIM$_2$ are formulated as two operator-fitting problems, respectively targeting a block-diagonal wideband precoder and the ideal OFDM modulation operator. Under practical discrete phase constraints, we develop a quantization-aware training-based gradient descent (QAT-GD) framework for both SIM stages. Numerical results verify the effectiveness of the proposed optimization, the correctness of the synthesized wave-domain functionalities, and the strong end-to-end MU-MIMO OFDM performance of the resulting architecture.

CoCurve: Cross-Module Co-Pruning Curvature for Training-Free Structured LLM Pruning

Structured pruning compresses large language models (LLMs) by removing whole computational units, such as attention heads and feed-forward (FFN) channel groups. Most training-free methods, however, rank these units independently, implicitly treating the loss from pruning a set as the sum of its individual losses. This view fails for Transformers, whose sublayers are coupled through a shared residual stream. Two individually weak units can thus be jointly indispensable, yet independent scoring is blind to such dependence and removes them together. We introduce CoCurve (Cross-Module Co-Pruning Curvature), a calibration-only, fine-tuning-free method that prunes attention and FFN units jointly. A second-order Taylor expansion of the token-level KL between the frozen model and its masked copy yields a single Fisher matrix whose diagonal is classical node saliency and whose off-diagonal entries are co-pruning curvature edges: the extra damage of removing two units together. Under a single-ablation additivity approximation this matrix reduces to a Gram product of single-unit ablation features, so the full M x M interaction is recovered from M forward passes, with no pairwise sweeps or gradients. Pruning then reduces to one budgeted quadratic program, solved in a single shot under a shared attention--FFN budget, with no labels, fine-tuning, or recovery.

Task-Oriented Sensing and Covert Transmissions for Collaborative Multi-AUV Systems

In underwater covert cooperative missions, autonomous underwater vehicles (AUVs) often cannot rely on active sonar to continuously obtain complete information, since active sensing and frequent communications increase the risk of exposure. As a result, AUVs primarily rely on passive observation, an approach that yields incomplete local perception and limited task efficiency. Although underwater acoustic communications can mitigate this limitation through information sharing, they are simultaneously constrained by long delays, severe interference, low reliability, and the risk of covert exposure. Existing communications-oriented multi-agent reinforcement learning (MARL) studies often model communication as an ideal information flow, whereas traditional communication optimization primarily focuses on link-level performance. However, both are insufficient to characterize the actual contribution of perceptual information to cooperative tasks under realistic conditions of covert physical communications. This paper proposes a Sensed Information Value Realization Multi-Agent Reinforcement Learning (SVR-MARL) framework that leverages practical information to characterize the utility of information for cooperative tasks and learns distributed cooperative policies under realistic communication and covert constraints. Through a case study of covert multi-AUV cooperative localization and tracking, the potential of the proposed framework to improve collaborative task efficiency while reducing unnecessary communication and exposure risks is demonstrated.

CarbonCLIP: Enhance Carbon Prediction from Satellite Imagery via Integrated Street-View Semantics and Temporal Context Training

Accurately estimating urban carbon emissions is critical for sustainable urban planning, yet many existing approaches remain difficult to apply consistently across cities due to data-source heterogeneity and the lack of fine-grained semantic-temporal context in remote sensing data. We propose CarbonCLIP, a task-oriented multimodal distillation framework that improves satellite-based carbon emission prediction by transferring contextual knowledge into a unified satellite representation through dual-branch contrastive learning. Unlike conventional methods that rely on static visual features, CarbonCLIP explicitly bridges the gap between top-down satellite views and ground-level human activities. Specifically, the spatial branch uses fine-grained textual descriptions automatically generated from street-view images by Large Multimodal Models (LMMs) to provide semantic priors reflecting building functions, infrastructure, and urban activities, while the temporal branch employs a month encoder to encode temporal priors associated with monthly emission variation. CarbonCLIP requires multimodal data only during the pretraining phase; during inference, it relies solely on satellite imagery, thereby supporting scalable deployment when ground-level data are unavailable at inference. Experiments on Beijing and Singapore demonstrate that CarbonCLIP outperforms baselines in both study cities. The results validate that our method effectively transfers multimodal knowledge into satellite representations, offering a robust solution for satellite-based urban carbon modeling.

Conditional Co-Ablation: Recovering Self-Repair Backups in Transformer Circuits

Mechanistic interpretability often relies on component-level interventions to discover how a model produces a behavior. This guides attribution, capability knockout, and model pruning downstream to operate by scoring each unit by the effect of ablation in isolation. Such first-order scoring is natural when component importance is additive, but becomes misleading when a transformer self-repairs: after a primary component is removed, a dormant backup can take over, muting the primary's measured effect while the backup itself appears irrelevant on the intact model. We recast this failure as a recovery task, conditional circuit completion, and introduce Conditional Co-Ablation (CoAx), a label-free, output-grounded score that asks how much each remaining unit's ablation effect grows once a primary set has been removed. This conditional growth exposes the second-order interaction that single-unit scores discard. On the GPT-2-small IOI circuit, CoAx raises backup-head recovery from 0.33 to 0.91 ROC-AUC, outperforming all baselines, including self-repair-aware gradient scores (best 0.82); counterfactual patching verifies that the recovered heads causally carry the repair. The same label-free procedure transfers to induction across eight models. Beyond discovery, the recovered backups correct self-repair-masked attribution, identify the components required for capability knockout, and yield repair-aware structured pruning scaling from 124M to 7B. Component importance is therefore not merely an isolated-unit property: in robust circuits, the components that matter can become visible only under the interventions that make them necessary.

XDomainBench: Diagnosing Reasoning Collapse in High-Dimensional Scientific Knowledge Composition

Large Language Models (LLMs) are increasingly deployed for knowledge synthesis, yet their capacity for compositional generalization in scientific knowledge remains under-characterized. Existing benchmarks primarily focus on single-turn restricted scenarios, failing to capture the capability boundaries exposed by real-world interactive scientific workflows. To address this, we introduce XDomainBench, a diagnostic benchmark for interactive interdisciplinary scientific reasoning. We formalize the composition order and mixture structure to enable systematic stress-testing from single-discipline to inter-disciplinary, comprising 8,598 interactive sessions across 20 domains and 4 task categories, with 8 realistic trajectory patterns covering difficulty and domain-mixture dynamics, simulating real AI4S scenarios. Large-scale evaluation of LLMs reveals a systematic reasoning collapse as composition order increases, stemming from two root causes: (i) direct difficulty increases induced by domain composition, and (ii) indirect interaction-amplified failures where trajectory patterns trigger error accumulation, reasoning breaks, and domain confusion, ultimately leading to session collapse.

Signal-Dependent Shot Noise Modeling of Rydberg Atomic Quantum Receivers: A Design Perspective

In this paper, we develop a communication-oriented complex baseband equivalent model for superheterodyne Rydberg atomic quantum receivers (RAQRs). The model explicitly captures photodetection-induced signal-dependent shot noise and its coupling with the optical operating point. By leveraging an atomic superheterodyne architecture and a strong local oscillator, we construct a complex baseband representation for both the received signal and the signal-dependent shot noise under both direct incoherent optical detection and balanced coherent optical detection. The derived model reveals that the optical operating point jointly determines the normalized effective receive gain and the equivalent noise background, thereby establishing a traceable gain-noise tradeoff governed by system design. More importantly, the proposed model shows that neglecting signal-dependent shot noise may lead to inaccurate operating-point design. Finally, by extending to the multiple-input-multiple-output (MIMO) case, we derive a lower bound on the achievable rate while considering the signal-dependent shot noise. Our analysis \textcolor{black}{reveals} that the non-zero asymptotic rate of RAQ-MIMO and its superiority over conventional RF-MIMO hinge on the normalized noise floor of the RAQ receive chain falling below that of RF MIMO. Simulation results validate our analysis and yield practical, closed-form design guidelines for RAQR front ends, revealing parameter regimes in which RAQ-MIMO outperforms conventional MIMO systems.

Antenna Elements' Trajectory Optimization for Throughput Maximization in Continuous-Trajectory Fluid Antenna-Aided Wireless Communications

Fluid antenna (FA) systems offer novel spatial degrees of freedom (DoFs) with the potential for significant performance gains. Compared to existing works focusing solely on optimizing FA positions at discrete time instants, we introduce the concept of continuous-trajectory fluid antenna (CTFA), which explicitly considers the antenna element's movement trajectory across continuous time intervals and incorporates the inherent kinematic constraints present in practical FA implementations. Accordingly, we formulate the total throughput maximization problem in CTFA-aided wireless communication systems, addressing the joint optimization of continuous antenna trajectories in conjunction with the transmit covariance matrices under kinematic constraints. To effectively solve this non-convex problem with highly coupled optimization variables, we develop an iterative algorithm based on block coordinate descent (BCD) and majorization-minimization (MM) principles with the aid of the weighted minimum mean square error (WMMSE) method. Finally, numerical results are presented to validate the efficacy of the proposed algorithms and to quantify the substantial total throughput advantages afforded by the conceived CTFA-aided system compared to conventional fixed-position antenna (FPA) benchmarks and alternative approaches employing simplified trajectories.

Indirect and Direct Multiuser Hybrid Beamforming for Far-Field and Near-Field Communications: A Deep Learning Approach

Hybrid beamforming for extremely large-scale multiple-input multiple-output (XL-MIMO) systems is challenging in the near field because the channel depends jointly on angle and distance, and the multiuser interference (MUI) is strong. Existing deep learning methods typically follow either a decoupled design that optimizes analog beamforming without explicitly accounting for MUI, or an end-to-end (E2E) joint analog-digital optimization that can be unstable under nonconvex constant-modulus (CM), pronounced analog-digital coupling, and gradient pattern of sum-rate loss. To address both issues, we develop a complex-valued E2E framework based on a variant minimum mean square error (variant-MMSE) criterion, where the digital precoder is eliminated in closed form via Karush-Kuhn-Tucker (KKT) conditions so that analog learning is trained with a stable objective. The network employs a grouped complex-convolution sensing front-end for uplink (UL) measurements, a shared complex multi-layer perceptron (MLP) for per-user feature extraction, and a merged constant-modulus head to output the analog precoder. In the indirect mode, the network designs hybrid beamformers from estimated channel state information (CSI). In the direct mode where explicit CSI is unavailable, the network learns the sensing operator and the analog mapping from short pilots, after which additional pilots estimate the equivalent channel and enable a KKT closed-form digital precoder. Simulations show that the indirect mode approaches the performance of iterative variant-MMSE optimization with a complexity reduction proportional to the antenna number. In the direct mode, the proposed method improves spectral efficiency over sparse-recovery pipelines and recent deep learning baselines under the same pilot budget.