Despite recent progress, medical foundation models still struggle to unify visual understanding and generation, as these tasks have inherently conflicting goals: semantic abstraction versus pixel-level reconstruction. Existing approaches, typically based on parameter-shared autoregressive architectures, frequently lead to compromised performance in one or both tasks. To address this, we present UniX, a next-generation unified medical foundation model for chest X-ray understanding and generation. UniX decouples the two tasks into an autoregressive branch for understanding and a diffusion branch for high-fidelity generation. Crucially, a cross-modal self-attention mechanism is introduced to dynamically guide the generation process with understanding features. Coupled with a rigorous data cleaning pipeline and a multi-stage training strategy, this architecture enables synergistic collaboration between tasks while leveraging the strengths of diffusion models for superior generation. On two representative benchmarks, UniX achieves a 46.1% improvement in understanding performance (Micro-F1) and a 24.2% gain in generation quality (FD-RadDino), using only a quarter of the parameters of LLM-CXR. By achieving performance on par with task-specific models, our work establishes a scalable paradigm for synergistic medical image understanding and generation. Codes and models are available at https://github.com/ZrH42/UniX.

Frontier LLMs are increasingly utilised across academia, society and industry. A commonly used unit for comparing models, their inputs and outputs, and estimating inference pricing is the token. In general, tokens are used as a stable currency, assumed to be broadly consistent across tokenizers and contexts, enabling direct comparisons. However, tokenization varies significantly across models and domains of text, making naive interpretation of token counts problematic. We quantify this variation by providing a comprehensive empirical analysis of tokenization, exploring the compression of sequences to tokens across different distributions of textual data. Our analysis challenges commonly held heuristics about token lengths, finding them to be overly simplistic. We hope the insights of our study add clarity and intuition toward tokenization in contemporary LLMs.

Large reasoning models (LRMs) produce a textual chain of thought (CoT) in the process of solving a problem, which serves as a potentially powerful tool to understand the problem by surfacing a human-readable, natural-language explanation. However, it is unclear whether these explanations generalize, i.e. whether they capture general patterns about the underlying problem rather than patterns which are esoteric to the LRM. This is a crucial question in understanding or discovering new concepts, e.g. in AI for science. We study this generalization question by evaluating a specific notion of generalizability: whether explanations produced by one LRM induce the same behavior when given to other LRMs. We find that CoT explanations often exhibit this form of generalization (i.e. they increase consistency between LRMs) and that this increased generalization is correlated with human preference rankings and post-training with reinforcement learning. We further analyze the conditions under which explanations yield consistent answers and propose a straightforward, sentence-level ensembling strategy that improves consistency. Taken together, these results prescribe caution when using LRM explanations to yield new insights and outline a framework for characterizing LRM explanation generalization.

Recent advances in 3D shape generation have achieved impressive results, but most existing methods rely on clean, unoccluded, and well-segmented inputs. Such conditions are rarely met in real-world scenarios. We present ShapeR, a novel approach for conditional 3D object shape generation from casually captured sequences. Given an image sequence, we leverage off-the-shelf visual-inertial SLAM, 3D detection algorithms, and vision-language models to extract, for each object, a set of sparse SLAM points, posed multi-view images, and machine-generated captions. A rectified flow transformer trained to effectively condition on these modalities then generates high-fidelity metric 3D shapes. To ensure robustness to the challenges of casually captured data, we employ a range of techniques including on-the-fly compositional augmentations, a curriculum training scheme spanning object- and scene-level datasets, and strategies to handle background clutter. Additionally, we introduce a new evaluation benchmark comprising 178 in-the-wild objects across 7 real-world scenes with geometry annotations. Experiments show that ShapeR significantly outperforms existing approaches in this challenging setting, achieving an improvement of 2.7x in Chamfer distance compared to state of the art.

Machine learning models are often used to make predictions about admissions process outcomes, such as for colleges or jobs. However, such decision processes differ substantially from the conventional machine learning paradigm. Because admissions decisions are capacity-constrained, whether a student is admitted depends on the other applicants who apply. We show how this dependence affects predictive performance even in otherwise ideal settings. Theoretically, we introduce two concepts that characterize the relationship between admission function properties, machine learning representation, and generalization to applicant pool distribution shifts: instability, which measures how many existing decisions can change when a single new applicant is introduced; and variability, which measures the number of unique students whose decisions can change. Empirically, we illustrate our theory on individual-level admissions data from the New York City high school matching system, showing that machine learning performance degrades as the applicant pool increasingly differs from the training data. Furthermore, there are larger performance drops for schools using decision rules that are more unstable and variable. Our work raises questions about the reliability of predicting individual admissions probabilities.

While using formal methods offers advantages over unit testing, their steep learning curve can be daunting to developers and can be a major impediment to widespread adoption. To support integration into an industrial software engineering workflow, a tool must provide useful information and must be usable with relatively minimal user effort. In this paper, we discuss our experiences associated with identifying and applying formal methods tools on an electronic warfare (EW) system with stringent safety requirements and present perspectives on formal methods tools from EW software engineers who are proficient in development yet lack formal methods training. In addition to a difference in mindset between formal methods and unit testing approaches, some formal methods tools use terminology or annotations that differ from their target programming language, creating another barrier to adoption. Input/output contracts, objects in memory affected by a function, and loop invariants can be difficult to grasp and use. In addition to usability, our findings include a comparison of vulnerabilities detected by different tools. Finally, we present suggestions for improving formal methods usability including better documentation of capabilities, decreased manual effort, and improved handling of library code.

To what extent is social media research independent from industry influence? Leveraging openly available data, we show that half of the research published in top journals has disclosable ties to industry in the form of prior funding, collaboration, or employment. However, the majority of these ties go undisclosed in the published research. These trends do not arise from broad scientific engagement with industry, but rather from a select group of scientists who maintain long-lasting relationships with industry. Undisclosed ties to industry are common not just among authors, but among reviewers and academic editors during manuscript evaluation. Further, industry-tied research garners more attention within the academy, among policymakers, on social media, and in the news. Finally, we find evidence that industry ties are associated with a topical focus away from impacts of platform-scale features. Together, these findings suggest industry influence in social media research is extensive, impactful, and often opaque. Going forward there is a need to strengthen disclosure norms and implement policies to ensure the visibility of independent research, and the integrity of industry supported research.

To make DNA a suitable medium for archival data storage, it is essential to consider the decay process of the strands observed in DNA storage systems. This paper studies the decay process as a probabilistic noisy torn paper channel (TPC), which first corrupts the bits of the transmitted sequence in a probabilistic manner by substitutions, then breaks the sequence into a set of noisy unordered substrings. The present work devises coding schemes for the noisy TPC by embedding markers in the transmitted sequence. We investigate the use of static markers and markers connected to the data in the form of hash functions. These two tools have also been recently exploited to tackle the noiseless TPC. Simulations show that static markers excel at higher substitution probabilities, while data-dependent markers are superior at lower noise levels. Both approaches achieve reconstruction rates exceeding $99\%$ with no false decodings observed, primarily limited by computational resources.

Smart grid infrastructures have revolutionized energy distribution, but their day-to-day operations require robust anomaly detection methods to counter risks associated with cyber-physical threats and system faults potentially caused by natural disasters, equipment malfunctions, and cyber attacks. Conventional machine learning (ML) models are effective in several domains, yet they struggle to represent the complexities observed in smart grid systems. Furthermore, traditional ML models are highly susceptible to adversarial manipulations, making them increasingly unreliable for real-world deployment. Quantum ML (QML) provides a unique advantage, utilizing quantum-enhanced feature representations to model the intricacies of the high-dimensional nature of smart grid systems while demonstrating greater resilience to adversarial manipulation. In this work, we propose QUPID, a partitioned quantum neural network (PQNN) that outperforms traditional state-of-the-art ML models in anomaly detection. We extend our model to R-QUPID that even maintains its performance when including differential privacy (DP) for enhanced robustness. Moreover, our partitioning framework addresses a significant scalability problem in QML by efficiently distributing computational workloads, making quantum-enhanced anomaly detection practical in large-scale smart grid environments. Our experimental results across various scenarios exemplifies the efficacy of QUPID and R-QUPID to significantly improve anomaly detection capabilities and robustness compared to traditional ML approaches.

We study first-hitting times in Differential Evolution (DE) through a conditional hazard frame work. Instead of analyzing convergence via Markov-chain transition kernels or drift arguments, we ex press the survival probability of a measurable target set $A$ as a product of conditional first-hit probabilities (hazards) $p_t=\Prob(E_t\mid\mathcal F_{t-1})$. This yields distribution-free identities for survival and explicit tail bounds whenever deterministic lower bounds on the hazard hold on the survival event. For the L-SHADE algorithm with current-to-$p$best/1 mutation, we construct a checkable algorithmic witness event $\mathcal L_t$ under which the conditional hazard admits an explicit lower bound depending only on sampling rules, population size, and crossover statistics. This separates theoretical constants from empirical event frequencies and explains why worst-case constant-hazard bounds are typically conservative. We complement the theory with a Kaplan--Meier survival analysis on the CEC2017 benchmark suite . Across functions and budgets, we identify three distinct empirical regimes: (i) strongly clustered success, where hitting times concentrate in short bursts; (ii) approximately geometric tails, where a constant-hazard model is accurate; and (iii) intractable cases with no observed hits within the evaluation horizon. The results show that while constant-hazard bounds provide valid tail envelopes, the practical behavior of L-SHADE is governed by burst-like transitions rather than homogeneous per-generati on success probabilities.

An important part of the information theory folklore had been about the output statistics of codes that achieve the capacity and how the empirical distributions compare to the output distributions induced by the optimal input in the channel capacity problem. Results for a variety of such empirical output distributions of good codes have been known in the literature, such as the comparison of the output distribution of the code to the optimal output distribution in vanishing and non-vanishing error probability cases. Motivated by these, we aim to achieve similar results for the quantum codes that are used for classical communication, that is the setting in which the classical messages are communicated through quantum codewords that pass through a noisy quantum channel. We first show the uniqueness of the optimal output distribution, to be able to talk more concretely about the optimal output distribution. Then, we extend the vanishing error probability results to the quantum case, by using techniques that are close in spirit to the classical case. We also extend non-vanishing error probability results to the quantum case on block codes, by using the second-order converses for such codes based on hypercontractivity results for the quantum generalized depolarizing semi-groups.

Randomized algorithms in numerical linear algebra have proven to be effective in ameliorating issues of scalability when working with large matrices, efficiently producing accurate low-rank approximations. A key remaining challenge, however, is to efficiently assess the approximation accuracy of randomized methods without additional expensive matrix accesses. Recent work has addressed this issue by deriving fast leave-one-out error estimators for the randomized SVD and Nyström decomposition, enabling accurate error estimation with no additional matrix accesses. In this work, we extend the leave-one-out framework to the generalized Nyström decomposition, an approach that can be applied to general rectangular matrices. We do this by deriving three new leave-one-out error estimators and validating their effectiveness through numerical experiments.

In the generative AI era, where even critical medical tasks are increasingly automated, radiology report generation (RRG) continues to rely on suboptimal metrics for quality assessment. Developing domain-specific metrics has therefore been an active area of research, yet it remains challenging due to the lack of a unified, well-defined framework to assess their robustness and applicability in clinical contexts. To address this, we present CTest-Metric, a first unified metric assessment framework with three modules determining the clinical feasibility of metrics for CT RRG. The modules test: (i) Writing Style Generalizability (WSG) via LLM-based rephrasing; (ii) Synthetic Error Injection (SEI) at graded severities; and (iii) Metrics-vs-Expert correlation (MvE) using clinician ratings on 175 "disagreement" cases. Eight widely used metrics (BLEU, ROUGE, METEOR, BERTScore-F1, F1-RadGraph, RaTEScore, GREEN Score, CRG) are studied across seven LLMs built on a CT-CLIP encoder. Using our novel framework, we found that lexical NLG metrics are highly sensitive to stylistic variations; GREEN Score aligns best with expert judgments (Spearman~0.70), while CRG shows negative correlation; and BERTScore-F1 is least sensitive to factual error injection. We will release the framework, code, and allowable portion of the anonymized evaluation data (rephrased/error-injected CT reports), to facilitate reproducible benchmarking and future metric development.

Message delivery respecting causal ordering (causal delivery) is one of the most classic and widely useful abstraction for inter-process communication in a distributed system. Most approaches tag messages with causality information and buffer them at the receiver until they can be safely delivered. Except for specific approaches that exploit communication topology, therefore not generally applicable, they incur a metadata overhead which is prohibitive for a large number of processes. Much less used are the approaches that enforce causal order by buffering messages at the sender, until it is safe to release them to the network, as the classic algorithm has too many drawbacks. In this paper, first we discuss the limitations of sender-only buffering approaches and introduce the Sender Permission to Send (SPS) enforcement strategy, showing that SPS + FIFO implies Causal. We analyze a recent sender-buffering algorithm, Cykas, which follows SPS + FIFO, albeit very conservatively, pointing out throughput scalability and liveness issues. Then, we introduce a novel SPS + FIFO based algorithm, which adopts a new hybrid approach: enforcing causality by combining sender-buffering to enforce SPS and receiver-buffering to enforce FIFO. The algorithm overcomes limitations of sender-only buffering, and achieves effectively constant metadata size per message. By a careful choice of data-structures, the algorithm is also computationally-optimal, with amortized effectively constant processing overhead. As far as we know, there is no other topology-agnostic causal delivery algorithm with these properties.

This article is concerned with tensor field tomography in a fairly general setting, that takes refraction, attenuation and time-dependence of tensor fields into account. The mathematical model is given by attenuated ray transforms of the fields along geodesic curves corresponding to a Riemannian metric that is defined by the index of refraction. The data are given at the boundary tangent bundle of the domain and it is well-known that they can be characterized as boundary data of a transport equation turning tensor field tomography into an inverse source problem. This way the adjoint of the forward mapping can be computed using the integral representation or, equivalently, associated to a dual transport equation. The article offers and proves two different representations for the adjoint mappings both in the dynamic and static case. The numerical implementation is demonstrated and evaluated for static fields using the damped Landweber method with Nesterov acceleration applied to both, the integral and PDE-based formulations. The transport equations are solved using a viscosity approximation. The error analysis reveals that the integral representation significantly outperforms PDE-based methods in terms of computational efficiency while achieving comparable reconstruction accuracy. The impact of noise and deviations from straight-line trajectories are investigated confirming improved accuracy if refraction is taken into account. We conclude that the inclusion of refraction to the forward model pays in spite of increased numerical cost.

Vision-Language-Action (VLA) models are emerging as highly effective planning models for end-to-end autonomous driving systems. However, current works mostly rely on imitation learning from sparse trajectory annotations and under-utilize their potential as generative models. We propose Generative Scenario Rollouts (GeRo), a plug-and-play framework for VLA models that jointly performs planning and generation of language-grounded future traffic scenes through an autoregressive rollout strategy. First, a VLA model is trained to encode ego vehicle and agent dynamics into latent tokens under supervision from planning, motion, and language tasks, facilitating text-aligned generation. Next, GeRo performs language-conditioned autoregressive generation. Given multi-view images, a scenario description, and ego-action questions, it generates future latent tokens and textual responses to guide long-horizon rollouts. A rollout-consistency loss stabilizes predictions using ground truth or pseudo-labels, mitigating drift and preserving text-action alignment. This design enables GeRo to perform temporally consistent, language-grounded rollouts that support long-horizon reasoning and multi-agent planning. On Bench2Drive, GeRo improves driving score and success rate by +15.7 and +26.2, respectively. By integrating reinforcement learning with generative rollouts, GeRo achieves state-of-the-art closed-loop and open-loop performance, demonstrating strong zero-shot robustness. These results highlight the promise of generative, language-conditioned reasoning as a foundation for safer and more interpretable end-to-end autonomous driving.

We present a novel probabilistic approach for optimal path experimental design. In this approach a discrete path optimization problem is defined on a static navigation mesh, and trajectories are modeled as random variables governed by a parametric Markov policy. The discrete path optimization problem is then replaced with an equivalent stochastic optimization problem over the policy parameters, resulting in an optimal probability model that samples estimates of the optimal discrete path. This approach enables exploration of the utility function's distribution tail and treats the utility function of the design as a black box, making it applicable to linear and nonlinear inverse problems and beyond experimental design. Numerical verification and analysis are carried out by using a parameter identification problem widely used in model-based optimal experimental design.

Predictive Process Monitoring is a branch of process mining that aims to predict the outcome of an ongoing process. Recently, it leveraged machine-and-deep learning architectures. In this paper, we extend our prior LLM-based Predictive Process Monitoring framework, which was initially focused on total time prediction via prompting. The extension consists of comprehensively evaluating its generality, semantic leverage, and reasoning mechanisms, also across multiple Key Performance Indicators. Empirical evaluations conducted on three distinct event logs and across the Key Performance Indicators of Total Time and Activity Occurrence prediction indicate that, in data-scarce settings with only 100 traces, the LLM surpasses the benchmark methods. Furthermore, the experiments also show that the LLM exploits both its embodied prior knowledge and the internal correlations among training traces. Finally, we examine the reasoning strategies employed by the model, demonstrating that the LLM does not merely replicate existing predictive methods but performs higher-order reasoning to generate the predictions.

As vision-language models (VLMs) tackle increasingly complex and multimodal tasks, the rapid growth of Key-Value (KV) cache imposes significant memory and computational bottlenecks during inference. While Multi-Head Latent Attention (MLA) offers an effective means to compress the KV cache and accelerate inference, adapting existing VLMs to the MLA architecture without costly pretraining remains largely unexplored. In this work, we present MHA2MLA-VLM, a parameter-efficient and multimodal-aware framework for converting off-the-shelf VLMs to MLA. Our approach features two core techniques: (1) a modality-adaptive partial-RoPE strategy that supports both traditional and multimodal settings by selectively masking nonessential dimensions, and (2) a modality-decoupled low-rank approximation method that independently compresses the visual and textual KV spaces. Furthermore, we introduce parameter-efficient fine-tuning to minimize adaptation cost and demonstrate that minimizing output activation error, rather than parameter distance, substantially reduces performance loss. Extensive experiments on three representative VLMs show that MHA2MLA-VLM restores original model performance with minimal supervised data, significantly reduces KV cache footprint, and integrates seamlessly with KV quantization.

We introduce a smooth variant of the SCAD thresholding rule for wavelet denoising by replacing its piecewise linear transition with a raised cosine. The resulting shrinkage function is odd, continuous on R, and continuously differentiable away from the main threshold, yet retains the hallmark SCAD properties of sparsity for small coefficients and near unbiasedness for large ones. This smoothness places the rule within the continuous thresholding class for which Stein's unbiased risk estimate is valid. As a result, unbiased risk computation, stable data-driven threshold selection, and the asymptotic theory of Kudryavtsev and Shestakov apply. A corresponding nonconvex prior is obtained whose posterior mode coincides with the estimator, yielding a transparent Bayesian interpretation. We give an explicit SURE risk expression, discuss the oracle scale of the optimal threshold, and describe both global and level-dependent adaptive versions. The smooth SCAD rule therefore offers a tractable refinement of SCAD, combining low bias, exact sparsity, and analytical convenience in a single wavelet shrinkage procedure.

Information overload and misinformation create significant challenges in extracting meaningful narratives from large news collections. This paper defines the nascent field of Interactive Narrative Analytics (INA), which combines computational narrative extraction with interactive visual analytics to support sensemaking. INA approaches enable the interactive exploration of narrative structures through computational methods and visual interfaces that facilitate human interpretation. The field faces challenges in scalability, interactivity, knowledge integration, and evaluation standardization, yet offers promising opportunities across news analysis, intelligence, scientific literature exploration, and social media analysis. Through the combination of computational and human insight, INA addresses complex challenges in narrative sensemaking.

Indoor high-capacity connectivity is frequently constrained by significant building penetration loss and the inherent uplink power limitations of a typical outdoor macro-cell deployment. While Mobile Network Operators (MNOs) must optimize spectrum across low-band (<1 GHz) and mid-band (1-7 GHz) frequencies, uplink performance remains disproportionately degraded due to link budget asymmetry. Neutral-host (NH) networking provides a scalable alternative by transparently offloading MNO subscribers via spectrum sharing and shared infrastructure. We present a multi-site measurement study comparing Citizens Broadband Radio Service (CBRS)-enabled NH networks against public MNO 4G/5G macro deployments and Wi-Fi. Our results show: (i) significant building penetration loss with up to 15.5 dB in low-bands and 17.9 dB in mid-bands, resulting in a ~10 dB RSRP deficit for MNO mid-bands compared to low-bands; (ii) NH networks provide a 30 dB higher median indoor RSRP with indoor NH normalized downlink throughput matches MNO outdoor performance, while its uplink performance exceeds MNO levels in both indoor and outdoor settings; (iii) NH proximity enables superior uplink efficiency, utilizing 64-QAM for 56% of transmissions (versus <6% for MNOs) and reducing median UE transmit power by 5 dB; (iv) MNOs rely on low-band spectrum for indoor uplink transmissions, while the NH deployment maintains high-performance mid-band connectivity; and (v) NH outperforms MNOs in end-to-end throughput but trails Wi-Fi in uplink throughput and latency due to packet routing overhead to the MNO core.

In the modern Systems-on-Chip (SoC), the Advanced eXtensible Interface (AXI) protocol exhibits security vulnerabilities, enabling partial or complete denial-of-service (DoS) through protocol-violation attacks. The recent countermeasures lack a dedicated real-time protocol semantic analysis and evade protocol compliance checks. This paper tackles this AXI vulnerability issue and presents an intelligent hardware monitoring system (IMS) for real-time detection of AXI protocol violations. IMS is a hardware module leveraging neural networks to achieve high detection accuracy. For model training, we perform DoS attacks through header-field manipulation and systematic malicious operations, while recording AXI transactions to build a training dataset. We then deploy a quantization-optimized neural network, achieving 98.7% detection accuracy with <=3% latency overhead, and throughput of >2.5 million inferences/s. We subsequently integrate this IMS into a RISC-V SoC as a memory-mapped IP core to monitor its AXI bus. For demonstration and initial assessment for later ASIC integration, we implemented this IMS on an AMD Zynq UltraScale+ MPSoC ZCU104 board, showing an overall small hardware footprint (9.04% look-up-tables (LUTs), 0.23% DSP slices, and 0.70% flip-flops) and negligible impact on the overall design's achievable frequency. This demonstrates the feasibility of lightweight, security monitoring for resource-constrained edge environments.

Retrieval-Augmented Generation (RAG) has emerged as a powerful approach for enhancing large language models' question-answering capabilities through the integration of external knowledge. However, when adapting RAG systems to specialized domains, challenges arise from distribution shifts, resulting in suboptimal generalization performance. In this work, we propose TTARAG, a test-time adaptation method that dynamically updates the language model's parameters during inference to improve RAG system performance in specialized domains. Our method introduces a simple yet effective approach where the model learns to predict retrieved content, enabling automatic parameter adjustment to the target domain. Through extensive experiments across six specialized domains, we demonstrate that TTARAG achieves substantial performance improvements over baseline RAG systems. Code available at https://github.com/sunxin000/TTARAG.

We propose Map2Thought, a framework that enables explicit and interpretable spatial reasoning for 3D VLMs. The framework is grounded in two key components: Metric Cognitive Map (Metric-CogMap) and Cognitive Chain-of-Thought (Cog-CoT). Metric-CogMap provides a unified spatial representation by integrating a discrete grid for relational reasoning with a continuous, metric-scale representation for precise geometric understanding. Building upon the Metric-CogMap, Cog-CoT performs explicit geometric reasoning through deterministic operations, including vector operations, bounding-box distances, and occlusion-aware appearance order cues, producing interpretable inference traces grounded in 3D structure. Experimental results show that Map2Thought enables explainable 3D understanding, achieving 59.9% accuracy using only half the supervision, closely matching the 60.9% baseline trained with the full dataset. It consistently outperforms state-of-the-art methods by 5.3%, 4.8%, and 4.0% under 10%, 25%, and 50% training subsets, respectively, on the VSI-Bench.

Large language models (LLMs) exhibit exceptional performance across various domains, yet they face critical safety concerns. Model editing has emerged as an effective approach to mitigate these issues. Existing model editing methods often focus on optimizing an information matrix that blends new and old knowledge. While effective, these approaches can be computationally expensive and may cause conflicts. In contrast, we shift our attention to Hierarchical Orthogonal Residual SprEad of the information matrix, which reduces noisy gradients and enables more stable edits from a different perspective. We demonstrate the effectiveness of our method HORSE through a clear theoretical comparison with several popular methods and extensive experiments conducted on two datasets across multiple LLMs. The results show that HORSE maintains precise massive editing across diverse scenarios. The code is available at https://github.com/XiaojieGu/HORSE

We address discrete-time consensus on the Euclidean unit sphere. For this purpose we consider a distributed algorithm comprising the iterative projection of a conical combination of neighboring states. Neighborhoods are represented by a strongly connected directed graph, and the conical combinations are represented by a (non-negative) weight matrix with a zero structure corresponding to the graph. A first result mirrors earlier results for gradient flows. Under the assumptions that each diagonal element of the weight matrix is more than $\sqrt{2}$ larger than the sum of the other elements in the corresponding row, the sphere dimension is greater or equal to 2, and the graph, as well as the weight matrix, is symmetric, we show that the algorithm comprises gradient ascent, stable fixed points are consensus points, and the set of initial points for which the algorithm converges to a non-consensus fixed point has measure zero. The second result is that for the unit circle and a strongly connected graph or for any unit sphere with dimension greater than or equal to $1$ and the complete graph, only for a measure zero set of weight matrices there are fixed points for the algorithm which do not have consensus or antipodal configurations.

Microwave linear analog computers (MiLACs) have recently emerged as a promising solution for future gigantic multiple-input multiple-output (MIMO) systems, enabling beamforming with greatly reduced hardware and computational cost. However, channel estimation for MiLAC-aided systems remains an open problem. Conventional least squares (LS) and minimum mean square error (MMSE) estimation rely on intensive digital computation, which undermines the benefits offered by MiLACs. In this letter, we propose efficient LS and MMSE channel estimation schemes for MiLAC-aided MIMO systems. By designing training precoders and combiners implemented by MiLACs, both LS and MMSE estimation are performed fully in the analog domain, achieving identical performance to their digital counterparts while significantly reducing computational complexity, transmit RF chains, analog-to-digital/digital-to-analog converters (ADCs/DACs) resolution requirements, and peak-to-average power ratio (PAPR). Numerical results verify the effectiveness and advantages of the proposed schemes.

This paper studies decentralized stochastic nonconvex optimization problem over row-stochastic networks. We consider the heavy-tailed gradient noise which is empirically observed in many popular real-world applications. Specifically, we propose a decentralized normalized stochastic gradient descent with Pull-Diag gradient tracking, which achieves approximate stationary points with the optimal sample complexity and the near-optimal communication complexity. We further follow our framework to study the setting of undirected networks, also achieving the nearly tight upper complexity bounds. Moreover, we conduct empirical studies to show the practical superiority of the proposed methods.

Deep Differentiable Logic Gate Networks (LGNs) and Lookup Table Networks (LUTNs) are demonstrated to be suitable for the automatic classification of electrocardiograms (ECGs) using the inter-patient paradigm. The methods are benchmarked using the MIT-BIH arrhythmia data set, achieving up to 94.28% accuracy and a $jκ$ index of 0.683 on a four-class classification problem. Our models use between 2.89k and 6.17k FLOPs, including preprocessing and readout, which is three to six orders of magnitude less compared to SOTA methods. A novel preprocessing method is utilized that attains superior performance compared to existing methods for both the mixed-patient and inter-patient paradigms. In addition, a novel method for training the Lookup Tables (LUTs) in LUTNs is devised that uses the Boolean equation of a multiplexer (MUX). Additionally, rate coding was utilized for the first time in these LGNs and LUTNs, enhancing the performance of LGNs. Furthermore, it is the first time that LGNs and LUTNs have been benchmarked on the MIT-BIH arrhythmia dataset using the inter-patient paradigm. Using an Artix 7 FPGA, between 2000 and 2990 LUTs were needed, and between 5 to 7 mW (i.e. 50 pJ to 70 pJ per inference) was estimated for running these models. The performance in terms of both accuracy and $jκ$-index is significantly higher compared to previous LGN results. These positive results suggest that one can utilize LGNs and LUTNs for the detection of arrhythmias at extremely low power and high speeds in heart implants or wearable devices, even for patients not included in the training set.