Papers

Automatic speech recognition models suffer from catastrophic forgetting when adapted to new domains, accents, or downstream tasks. This problem becomes increasingly important with the growing use of speech foundation models, where adaptation should be both memory-efficient and safe, preserving the broad capabilities learned during pretraining. Gradient Low-Rank Projection (GaLore) has recently been proposed as a memory-efficient fine-tuning method that keeps model parameters full-rank while reducing the optimizer memory through low-rank gradient projection. However, GaLore does not account for catastrophic forgetting. We propose Curvature-Guided GaLore (CGaLore), which incorporates old-task curvature information when selecting the low-rank projection bases. Specifically, CGaLore filters current-task gradients using Kronecker-factored approximate curvature from previous tasks before computing the gradient subspace. Our experiments show that CGaLore enables effective adaptation while alleviating forgetting, outperforming state-of-the-art continual learning baselines. Extensive ablation studies confirm the practical applicability of CGaLore.

Effective evaluation of zero-day jamming detectors requires robust adversarial models. However, existing attack models often assume prior knowledge of the target receiver, limiting their utility as evaluation benchmarks. On the detection side, existing detectors fail to capture the global temporal-spectral structure of jamming behavior and cannot differentiate zero-day strategies as they emerge. This paper addresses these limitations through a two-pronged framework. First, an online detection framework is introduced that combines a graph attention network (GAT) for temporal-spectral representation learning with Dirichlet process (DP)-means clustering. This framework jointly classifies known and discovers zero-day strategies within a unified learning objective. Second, an inference-driven reinforcement learning (RL) jammer is proposed as an adversarial benchmark. The jammer treats the target receiver as a black-box, infers the detector state via hypothesis testing, and optimizes the trade-off between attack impact and stealth. Simulation results show that the proposed RL jammer outperforms benchmarks, achieving 33% higher attack efficacy and 67% higher stealth. The proposed detection framework against the proposed RL jammer is shown to achieve 20% higher detection accuracy than the benchmarks.

In the early stages of a pandemic, identification of a small number of infected individuals through large-scale screening is critical for pandemic control, yet remains challenging under limited reagents and testing capacity. Existing group testing methods suffer from either high computational complexity or low identification accuracy. Even worse, no available methods provide theoretically rigorous analysis for sparse identification with hard constraints caused by the sample usage constraint and the dilution effect existing ubiquitously in practical applications. In this article, we propose the Logic Screening method (LoSc), an ultra fast, accurate, and theoretically grounded framework for large-scale screening. LoSc introduces a novel decoding algorithm with a very simple selection strategy, achieving identification of all positives with only O(klogn) pooled tests. The decoding relies only on logical operations, enabling direct hardware implementation and yielding ultra fast computational implementation. Moreover, LoSc explicitly incorporates dilution and sample usage constraints into pooling designs, and establishes theoretical guarantees to guide optimal pooling configurations. Extensive simulations confirm the superior effectiveness, efficiency, and scalability. We believe LoSc offers a fast and reliable solution for automatic large-scale screening.

As all generators are distributed in different areas among large scale power systems, the cooperative manipulation of the multi-generator system cannot be done well without consideration of the distance information of the generators, A distributed step-by-step finite-time consensus scheme for the heterogeneous Battery Energy Storage System (BESS) is proposed in this paper, where the coordinated consensus can be come into reality within a limited time, which is appealing for the electrical engineering community. To be concrete, at first, all BESSs are classified into several clusters according to their locations, and in each cluster, there is an active leader in charge of information receiving from outside. Then after, in order to coordinate the multi BESSs, five inputs, which are function oriented, were used to achieve energy level balancing, active/reactive power sharing, and voltage/frequency synchronization of the multi BESSs. To be further, the frequency and voltage restoration to the nominal values of the main grid was made possible by the introduction of a virtual leader, which is actually an external leader. Compared with the centralized methods, this control scheme is entirely distributed, and each BESS only utilizes the information of its own and its neighbors. Besides, this control is robust to the load perturbation and the plug-and-play of the communication topology. Finally, some simulation experiments are executed on the modified IEEE 57-bus system to verify the suggested scheme.

Discrete-time non-symmetric algebraic Riccati equations (DTNAREs) arise in game-theoretic computations of Nash equilibria. Solving such equations at large scale is often computationally prohibitive. This paper develops a numerical approach for large-scale DTNAREs whose solutions are low rank. A low-rank alternating direction implicit (ADI) method is introduced that recursively constructs a low-rank stabilizing solution without explicitly solving any projected DTNARE. Through the pole-placement property of the low-rank ADI iteration, the method ensures that the implicitly solved projected DTNARE always admits a stabilizing solution. An automatic shift-generation strategy is also developed for the ADI iterations. Once an initial shift is provided, the algorithm computes the low-rank solution without further user intervention. Numerical experiments on DTNAREs with dimensions between \(10^6\) and \(10^7\) demonstrate the accuracy and efficiency of the method. The results confirm that the proposed low-rank ADI algorithm is an effective solver for large-scale DTNAREs that would otherwise be computationally prohibitive.

We consider the problem of H-infinity control in the presence of a digital communication channel between the observer and the controller and investigate the fundamental trade-off between the required communication data rate and H-infinity performance. The system follows linear dynamics and has no additive internal noise, so that the uncertainty is confined to the initial state. For the scalar system, using a deterministic time-zero covering argument, we establish a lower bound on the required data rate for a given H-infinity cost and demonstrate why logarithmic quantization (Elia, 2000) is a natural choice. We further develop an achievability scheme and show that, for the scalar system, its rate matches the converse asymptotically in the high-data-rate limit.

The cloud radio access network (C-RAN) mitigates inter-cell interference by jointly processing the observations of distributed remote units (RUs) at a centralized unit (CU), but limited fronthaul capacity forces each RU to compress its received signal. Under transform-compress-forward, an RU transforms its signal and quantizes the resulting coefficients, with bit allocation distributing a finite bit budget across them. Classical reverse waterfilling assumes Gaussian sources, yet practical finite-alphabet symbols carry mutual information that saturates at $\log_2 M$, leaving bit allocation for such inputs unresolved. We address this by formulating bit allocation as maximizing the finite-alphabet generalized mutual information (GMI) achieved after linear MMSE (LMMSE) detection at the CU. Via the I-MMSE relation, this yields a fixed-point update whose converged solution decomposes into a vessel height, a shared water level, and a finite-alphabet mercury level; we term it {reverse mercury/waterfilling} (RMWF). Numerical results show that RMWF sustains end-to-end rate under tight fronthaul budgets and remains robust under antenna scaling, which is increasingly consequential as antenna counts outpace fronthaul capacity in modern C-RAN.

We study online control of a scalar storage system with adversarial nonnegative arrivals, known retention coefficient, and convex costs depending on both state and action. Each action must respect current resource availability and is chosen before the current arrival and cost function are revealed. For the existing simplex disturbance-action policy class, we give an exact reparameterization by cumulative allocation fractions and a decay-weighted projected subgradient update. The resulting regret bound is independent of policy memory length. For fixed retention coefficient and cost constants, the controller achieves $O(\sqrt T)$ regret against the best fixed infinite-memory policy in this class, using $O(\log T)$ memory and arithmetic operations per round and one cost-subgradient query. A storage-specific block construction gives a matching lower bound against every causal feasible controller, including randomized controllers. Writing $\tau=(1-\alpha)^{-1}$, the minimax expected regret is $\Theta(\sqrt T\min\{T,\tau\}^{3/2})$ for every finite-memory simplex policy class and its infinite-memory extension, when $\alpha\in[1/2,1)$, $T\ge4$, and the positive cost constants are fixed. This identifies the joint horizon and retention-time dependence for these policy benchmarks.

Vision-language-action (VLA) models map visual observations and natural-language instructions to robotic actions, but distribution shifts can compromise their reliability. Because these models may still succeed under out-of-distribution (OOD) conditions, detecting OOD inputs alone is insufficient to predict execution failure. In this paper, we introduce VLA-Scope, a two-stage framework that combines input-shift characterization with execution history to predict failure during OOD rollouts. The first stage uses pooled image and language representations to detect OOD inputs and classify their shift categories. For inputs flagged as OOD, the second stage combines the predicted category, action-prefix features, and execution progress features. A logistic regression model shared across shift categories updates failure risk as execution proceeds. We evaluate the framework with OpenVLA on ten LIBERO-Spatial tasks using leave-one-group-out cross-validation. OOD detection achieves a ROC-AUC of 0.9454, and shift classification achieves 91% accuracy. Evaluated independently of the OOD gate on all 1,400 OOD rollouts, the failure predictor achieves a ROC-AUC of 0.8497 after 60 executed actions, compared with 0.7906 without execution progress features. It also achieves a higher ROC-AUC than the evaluated ActProbe and SAFE-MLP baselines. These results suggest that combining action features with temporally aggregated execution step representations improves failure prediction under input shifts.

Discrete audio tokenization has become the critical interface between raw waveforms and autoregressive modeling in recent music generation. As a result, music tokenizers must simultaneously support high-fidelity reconstruction and produce discrete sequences that remain amenable to language modeling. Existing reconstruction-oriented tokenizers often mix musical structure with fine acoustic details, producing high-entropy tokens that are hard to model. In contrast, semantics-guided alternatives are designed for speech and do not fit music well, often hurting reconstruction quality. We address these trade-offs by rethinking music tokenization around a measurable notion of music semantic content grounded in downstream Music Information Retrieval tasks. Guided by this definition, we propose MuSeC, a music semantic codec that factorizes semantic and acoustic content directly from mixed signals without source separation. MuSeC preserves information required for high-fidelity reconstruction while producing more LM-friendly discrete units. Empirically, it improves reconstruction quality and yields more predictable token sequences, providing a practical foundation toward high-fidelity LLM music generation. Demos are available at https://longwaytog0.github.io/MuSeC/.

We study in-market uniform pricing and out-of-market bid cost recovery (BCR) payments in rolling-window dispatch for real-time power system operations with energy storage resources (ESRs). Due to intertemporal state-of-charge (SOC) constraints, ESR operations are temporally coupled, and existing in-market locational marginal pricing (LMP) may fail to compensate ESR's intertemporal opportunity costs, thereby triggering out-of-market BCR payments. We show that positive BCR is unavoidable when dispatched generators or ESRs have supply-side bids higher than the demand-side bid. We further identify an intertemporal coupling indicator associated with binding SOC constraints and demonstrate empirically that positive BCR arises only when this indicator is active, revealing that BCR is fundamentally driven by intertemporal coupling. In the simulation, we compare a BCR-minimizing uniform pricing scheme (UP-BCR) with existing real-time pricing methods under forecast uncertainty and show that UP-BCR substantially reduces BCR and demand payments relative to LMP while maintaining zero merchandising surplus in a copper-plate model.

This paper proposes a multimodal learning framework that integrates camera images, LiDAR point clouds, and radio-frequency (RF) pilots to realize multi-user multiple-input single-output (MU-MISO) beamforming optimization in mixed line-of-sight (LoS) and non-line-of-sight (NLoS) environments. In LoS-dominant environments, beamforming can be designed by utilizing sensing information related to user locations. However, LoS paths cannot be guaranteed to exist for all users in realistic dynamic environments due to blockages. To characterize NLoS propagation, detailed three-dimensional (3D) information of surrounding obstacles is required. This paper integrates camera images and LiDAR points with RF pilots to capture detailed 3D geometric, object-class, and material information of users and surrounding objects. The obtained multimodal information provides useful cues for both direct LoS paths and reflected NLoS propagation, and provides rich channel-related information for beamforming design. Specifically, a bounding-box detection neural network is first used to extract two-dimensional spatial, object-class, and material information from camera images. The image-derived information then guides the selection of LiDAR points to obtain 3D geometric information of users and surrounding obstacles, and assigns object-class and material labels to the selected points. The image-guided LiDAR features, together with RF pilots, are jointly processed by a designed neural architecture to optimize the beamforming vectors. Simulation results show that the proposed method achieves a higher minimum rate than benchmarks.

Large language models (LLMs) are moving onto mobile devices for increasingly diverse workloads over text, images, video, and audio. These applications often require long contexts, making the Key-Value (KV) cache a dominant memory bottleneck because it grows linearly with sequence length and is accessed at every decoding step. Prior work reduces KV-cache footprint through low-rank compression, token eviction, or flash offloading, but the resulting reconstruction overhead, irreversible token loss, or I/O stalls can offset the benefit of saving memory. We present TierKV, a mobile LLM inference framework built on Predictive Multi-Tier Cache Optimization (PMCO). Before decoding starts, PMCO predicts future cache demand from prefill hidden states and jointly assigns tokens to exact, low-rank, and flash-offloaded tiers under the device memory and accuracy budgets. This formulation retains access to the full context, removes the circular dependency of reactive eviction, and admits a closed-form solver that selects tier boundaries and per-layer ranks at runtime. Across eight text, vision, and audio models on three mobile SoCs, TierKV improves prefill throughput by up to 17.6x over existing mobile LLM frameworks, reduces RAM-resident KV cache by 12.5-34%, thereby enabling substantially longer contexts under the same memory budget, while incurring only minor accuracy degradation.

Recent speech enhancement systems combine self-attention and Mamba to capture global interactions and long-range dependencies. Yet these hybrids usually operate as sequence mixers and do not explicitly exploit harmonic periodicity, a strong cue for preserving voiced speech under noise. Perceptual optimization poses another challenge. PESQ is non-differentiable, so many methods train auxiliary metric discriminators that increase complexity and introduce adversarial instability. We propose \ours, a harmonic-aware and discriminator-free speech enhancer built around two components. (i) The Time-Frequency Harmonic-aware Attention-Mamba (TF-HAM) block runs self-attention and bidirectional Mamba in parallel along both spectrogram axes, then applies a speech-adapted autocorrelation feed-forward network to encode local periodic structure. (ii) Metric-explicit perceptual refinement (MEPR) combines differentiable PESQ and log-likelihood-ratio losses to expose perceptual metric structure without a learned surrogate. On VoiceBank+DEMAND, \ours achieves 3.69 PESQ and 4.41 COVL with only 2.39\,M parameters, outperforming or matching discriminator-based systems. Inference-time perceptual contrast stretching further raises PESQ to 3.79 without retraining. The source code will be available at https://github.com/shangfuu/HAMMER.git.

Safety-critical control systems must contend with disturbances whose extreme deviations occur far more frequently than classical light-tailed models predict. Existing stochastic MPC (SMPC) formulations tighten constraints using an assumed distribution, a moment bound, or a finite scenario sample, each of which degrades under an unknown heavy-tailed disturbance. This paper develops an SMPC formulation for linear systems under heavy-tailed disturbances that is only assumed to be regular varying, replacing these approaches with an explicit extreme value theory (EVT) characterization of the tube error tail that is asymptotically exact. We further show that closed-loop dynamics induce temporal clustering of rare excursions across the prediction horizon, and characterize this clustering through a closed-form extremal index estimable from data. The resulting $\theta$-corrected constraint bounds the probability of a rare-event episode over the horizon, rather than only the marginal per-step exceedance probability. Simulation on a nonlinear unicycle navigating past an obstacle under Student-$t$ disturbances validates both approaches and demonstrates reduced frequencies of safety constraint violations.

Independent control of tonescale regions (e.g., shadows, highlights) is essential for painters, photographers and cinematographers to bring 2D images to life. In image manipulation software this is most directly addressed by color grading modules, which use intensity thresholds to segment distinct illumination regions for local manipulation. In this work we develop an open source color grading tool and use it to annotate a large dataset of video frames with tonescale region thresholds. Using these thresholds we conduct modeling experiments with strategies based on both practitioners' conventional wisdom and machine learning. Results show that K-nearest neighbors is an effective prediction strategy, outperforming state-of-the-art end-to-end methods for image enhancement. This outcome demonstrates the benefit of focusing on a compact set of core parameters when modeling creative stylization processes. Our adaptive color grading interface and data are available at https://github.com/SamsungLabs/adaptive-color-grading.

Voxel-based volumetric mapping is fundamental to 3D reconstruction, yet fixed-resolution grids remain inherently inefficient - wasting memory in uniform regions and losing detail in complex ones. Existing adaptive methods, such as MAP-ADAPT, partially address this by varying resolution based on geometry and user-defined semantic class lists, but these heuristics require expert tuning, lack generalization to unseen objects, and provide no explicit mechanism to control memory usage. We propose an adaptive framework that refines voxels based on semantic entropy, which captures label uncertainty, together with geometric curvature and texture richness as scene complexity cues, yielding principled resolution allocation without reliance on semantic taxonomies. To make the accuracy-memory trade-off explicit and user-controlled, we further introduce a reinforcement learning agent that learns voxel subdivision policies under a user-specified target memory budget, replacing hand-tuned thresholds with a single intuitive control parameter. The resulting multi-resolution TSDF achieves higher geometric accuracy, better semantic consistency, and improved memory-accuracy trade-offs compared to MAP-ADAPT and fixed-resolution baselines on both synthetic and real-world datasets. Our code and models are available at https://github.com/alpayozkan/UnRL.

Intrathecal contrast-enhanced MRI shows how a tracer injected into the cerebrospinal fluid spreads through the brain, but requires repeated scans over 24–48\,h. We therefore ask how much of the later tracer distribution can be predicted from the pre-contrast scan and elapsed time alone. Using 104 training patients (878 MRI volumes), we compared a direct image-to-image regression model with a more complex generative model based on conditional flow matching, and tested both on 23 held-out patients and on 51 patients from a separate cohort with idiopathic normal pressure hydrocephalus (INPH). The direct model reduced the error of simply copying the pre-contrast scan by about 60% in held-out patients and 25% in the INPH cohort, and its predicted change maps agreed with the observed tracer enhancement rather than only reproducing anatomy. The generative model was less accurate in both cohorts. A substantial part of the later tracer distribution is therefore predictable from the patient's brain anatomy as seen on pre-contrast MRI, and the added model complexity gave no benefit at this data scale. With prospective validation, such forecasts could be studied for patient stratification and treatment planning in intrathecal therapies.

Large neighborhood search improves an incumbent by releasing selected variables and repairing the resulting subproblem under a time limit. FOVEA casts variable selection as dense semantic segmentation on a fixed $128\times128$ canvas. A small U-Net reads twelve semantic channels and predicts cell scores, which are decoded into a variable set subject to a fixed variable budget. Family-specific layouts connect the search state to the canvas, while one set of network weights serves four problem families. The FP32 network input occupies $786$\,kB across instance sizes. Our analysis bounds constraint coupling for spatially coherent regions, gives a rank certificate for rounds with zero possible improvement, and bounds the coupling advantage attainable on an expander family. On the tested families and hardware, FOVEA trails the strongest graph encoder at $10^{4}$ variables and overtakes it near $1.5\times10^{4}$; the cost model provides an approximate crossover estimate. At larger sizes, where the graph-encoder baselines exceed device memory, FOVEA reduces the primal integral by $12$ to $16\%$ relative to the best runnable baselines. On the expander family, FOVEA performs comparably to random selection, with a coupling advantage close to one.

This letter addresses budgeted stealthy false-data-injection (FDI) scheduling against remote state estimation, where a resource-constrained adversary may corrupt at most a fraction $\bar{\Gamma}$ of transmissions. Existing event-triggered schedulers invert a Gaussian innovation tail to set the firing threshold and certify stealth by covariance matching; both are exact only under Gaussianity, which real cyber-physical residuals routinely violate. We propose a distribution-free scheduler pairing the worst-case FDI action with a split-conformal calibrated trigger, requiring neither the plant matrices nor any distributional model. We establish exact pathwise stealth against every magnitude-measurable detector, for any firing rule and innovation law; a finite-sample distribution-free bound on the mean firing rate, with almost-sure budget attainment under stationarity and ergodicity; and a steady-state degradation identity linear in a single scalar energy capture $\psi$, maximized by the same order statistic that delivers the budget guarantee. A conditional sign-symmetry condition delimits when the certificate extends to sign-sensitive detectors, the residual exposure being governed by a fourth cumulant that vanishes under Gaussian noise. Monte-Carlo studies and a heavy-duty truck CAN record confirm the bounds and quantify what the Gaussian assumption costs outside its regime.

The 30 and 40 GHz instruments of the QUIJOTE radio astronomy experiment use very sensitive receivers which need to be characterized both to ensure their proper functioning and for their calibration. Given the age of the PXI-1031DC device used up until now for this purpose, the design of a new device, the Calibrator TFGI QUIJOTE, has been proposed. This new device has been designed with the idea of being simple and modular, easy to use and upgrade. It is built on a Raspberry Pi 5 system, using an MCC 118 as analog reading module. This system, widely known among both makers and professionals, not only meets the requirements of modularity and ease of use, but also ensures that the device is much more economically competitive than the PXI. A first version of the device has been built and tested, obtaining good results in accuracy and ease of use.

Coaxial rotor drones have generated considerable interest because of energy efficiency and small size, but they are afflicted with inherent underactuation for roll and pitch control, although systems like swashplates have circumvented this limitation at the cost of greater mechanical complexity. This work presents a novel coaxial drone supplemented by a two-degreesof-freedom pendulum mechanism for active thrust vectoring that offers a less mechanically complicated alternative. We develop a comprehensive Lagrangian dynamic model that does not ignore the inertial contributions of all the components, including body, servo arms, and motor assembly. A Linear Quadratic Regulator(LQR) is designed based on the linearized dynamics around the hover equilibrium. High-fidelity simulations taking actuator dynamics and sensor noise into account validate the proposed architecture. An Extended Kalman Filter (EKF) blends GPS, barometer, and IMU estimates with high accuracy for state estimation. The findings verify the potential and reliability of this approach for power-saving, rapid coaxial UAVs.

As Urban Air Mobility (UAM) systems scale toward high-density operations, managing autonomous Unmanned Aerial Vehicle (UAV) traffic requires control frameworks that are both tractable and safety-critical. This paper presents a principled optimal control-theoretic foundation for routing in multi-destination UAM networks subject to vertiport capacity and flow constraints. We first model the network as a destination-conditioned Continuous-Time Markov Chain (CTMC) to capture the stochastic transitions between queueing, service, and flight states. To ensure tractability, we employ a mean-field fluid approximation and derive the underlying system dynamics as a set of coupled ordinary differential equations. A key contribution of this work is the formal proof of the positive invariance of the queue-free state space. We demonstrate that under specific underloaded conditions, a system initialized without queues will remain queue-free indefinitely. This result allows us to transform a complex, infinite-dimensional continuous-time optimal control problem into a tractable, finite-dimensional algebraic optimization. The resulting framework jointly optimizes for travel time and multi-hop efficiency while ensuring network-wide stability. We validate the approach by characterizing the steady-state flow equilibria and providing sufficient conditions for safe, congestion-free operation in large-scale mobility systems.

Modern automatic speech recognition (ASR) systems trained on extremely large datasets can produce transcripts with numbers written in Arabic numerals. This creates a need for fair comparison with models that output verbatim texts and proper processing of reference transcripts. Popular approaches often reduce text normalization to lowercase and remove punctuation, with no additional normalization applied to languages other than English. In this work, we analyze the impact of normalization of numerical expressions in the evaluation of ASR systems in various languages, using Polish as an example of a highly inflective language. We perform experiments on VoxPopuli and The Polish Parliamentary speech datasets and estimate word error rate (WER) differences for different text normalization approaches. We show that the difference due to the lack of number normalization in WER may be substantial - more than 2 percentage points, and often higher than the differences between systems in popular multilingual benchmarks.

Quadcopters offer great utility in many applications, but their nonlinear nature and disturbance sensitivity present great control challenges. Basic PID controllers are generally not sophisticated enough to cope with these complexities. This paper suggests a control system that integrates the Asynchronous Advantage Actor-Critic (A3C) algorithm with a PID controller for quadcopter attitude and trajectory tracking. The A3C controller uses parallel agents to optimize PID parameters dynamically using a neural network. A system identification module for the complementary system makes predictions about system states for optimal control policy. The proposed framework was compared with a standard actor-critic (A2C) model. Simulation results verify that they both track accurately. However, the A3C-based controller converges much more for the loss function, as evidenced by reward figures and loss curves, demonstrating better parameter optimization. This shows that A3C-based approach results in improved performance for the control of quadcopter, effectively integrating reinforcement learning and traditional control to achieve higher adaptability.

Recent work revealed a striking Clever Hans effect in the Pitt dataset, where Alzheimer's detection achieved nearly 100% accuracy using only silent audio segments. This raises serious concerns about hidden confounding factors in speech-based health datasets. We systematically investigate whether similar biases exist across five widely used clinical speech corpora: DAIC-WoZ (depression), TORGO (dysarthria), Neurovoz (PD), MDVR-KCL (PD), and UCLASS (stuttering). For each dataset, we compare classification using the first second of audio, silent segments, and full recordings, and evaluate both raw and denoised signals. Across all datasets, silence-only classification frequently matched or exceeded full-audio performance, suggesting that classification performance may be influenced by dataset-specific confounds in addition to disorder-related speech characteristics. These findings question the reliability and generalisability of speech-based biomarkers and call for stricter methodological reporting, preprocessing transparency, and bias mitigation.

The reliance on scarce and expensive accelerators such as GPUs and TPUs in modern datacenters places unprecedented demands on backend infrastructure. For workloads characterized by heterogeneous service times and complex multi-stage processing, such as Generative AI, conventional load balancing techniques are often inadequate, relying heavily on costly overprovisioning to maintain service level objectives. This paper introduces DLB, the Distributed Load Balancer, a novel system designed to minimize end-to-end user latency for large-scale, heterogeneous workloads. DLB employs a scalable, distributed design with peer-to-peer probing to maintain real-time visibility into server capacity across large-scale, geographically distributed infrastructure. The system continuously learns latency models to estimate the latency impact of routing decisions, allowing it to effectively manage heterogeneous hardware and diverse model architectures. We provide a novel theoretical analysis of our routing algorithms that establishes their stability and global performance guarantees over time. We also evaluate DLB through extensive simulations, which show substantial gains compared to state-of-the-art load balancing algorithms. Finally, following a 22-month deployment of DLB at Google, where it facilitates large-scale Generative AI inference for thousands of different machine learning models and millions of requests per second, we detail the design choices and practical experiences gained from the system in production. Analysis of production migrations demonstrates that DLB yields statistically significant latency reductions compared to the legacy baseline, including a 17% decrease in median latency and a 13% decrease at the p95 tail.

Geofencing refers to controlling the availability of wireless services within a network perimeter. In this paper, we study how RIS can be used to achieve geofencing in outdoor-to-indoor network scenarios. Particularly, we propose GeoRIS, a RIS controller that achieves geofencing by exploiting beam management procedures to control beam alignment in networks with steerable directional transmission links. GeoRIS does not require control over the outdoor base station and can work with or without channel state information. Through simulations based on 3GPP network models, we validate GeoRIS and demonstrate that it can "shield" an indoor network area (for example, by weakening communication links to the point where indoor users cannot meet the minimum requirements of 5G eMBB services). We also highlight an interesting insight: the same RIS can play a dual role, extending or inhibiting outdoor-to-indoor communication. In our simulations, we show that GeoRIS can easily change an indoor space from a "strongly covered area" (e.g., supports eMBB services in approx. 90% of network area) to an "out-of-service area" (e.g., inhibits eMBB services in approx. 90% of network area), or vice-versa, an attractive characteristic in scenarios in which dynamic indoor coverage control is needed.

This paper proposes structured channel-parameter estimators for reconfigurable intelligent surface (RIS)-assisted multiple-input multiple-output (MIMO) systems employing uniform rectangular arrays under line-of-sight propagation. By exploiting the Kronecker and multilinear structures of the cascaded channel, we introduce the Successive Kronecker Factorization (SKF) alongside two Third-Order Tensor Parameter Estimation (TOPE) methods: TOPE-ALS and TOPE-HOSVD. These estimators recover spatial signatures from the pilot-filtered signal matrix, and then decouple the corresponding horizontal and vertical spatial frequencies. This structured formulation achieves benchmark estimation accuracy while reducing computational complexity, offering a efficient solution for RIS-aided systems.

In strategic interactions, deterrence is an outcome where one party incentivizes the other party to not participate. This concept can be applied to Colonel Blotto games between two generals where troops can be deterred from following orders by the actions of the opposing general. In this setting, deterrence depends not only on how generals allocate troops, but also on the information available to troops when they decide whether to follow orders or defect. In this work, we study a two-battlefield Colonel Blotto model of deterrence under multiple information structures, ranging from decisions made with only aggregate information to decisions made after local battlefield conditions are observed. We show how different information structures alter the structure of the induced game, affect the existence of equilibria, and substantially change expected utilities. Our findings also prove that the ratio of troops available to each general combined with a capture probability threshold is enough to inform generals about what information they want their troops to have access to.