Papers

The classic multi-armed bandit (MAB) problem tackles the challenge of accruing maximum reward while making decisions under uncertainty. However, in applications, often the goal is to minimize cost subject to a constraint on the minimum permissible reward, an objective captured by multi-armed bandits with cost-subsidy (MAB-CS). Of interest to this paper is the setting where the quality (reward) constraint is specified relative to the unknown best reward and the cost of each arm is known. We characterize the expected sub-optimal samples required by any policy by proving instance-dependent lower bounds that offer new insight into the problem and are a strict generalization of prior bounds. Then, we propose an algorithm called Cost-Ordered Feasibility (COF) that leverages our insight and intelligently combine samples from all arms to gauge the feasibility of a cheap arm. Thereafter, we analyze COF to establish instance-dependent upper bounds on its expected cumulative cost and quality regret, i.e., relative to the cheapest feasible arm. Finally, we empirically validate the merits of COF, comparing it to baselines from the literature through extensive simulation experiments on the MovieLens and Goodreads datasets as well as representative synthetic instances. Not only does our paper develop qualitatively better theoretical regret upper bounds, but COF also convincingly demonstrates improved empirical performance.

Channel knowledge map (CKM) has emerged as a promising technique to leverage prior propagation knowledge in low-altitude wireless networks (LAWNs), yet state-of-the-art grid-based CKM construction methods struggle to support efficient LAWN deployment due to their lack of differentiability with respect to continuous locations of unmanned aerial vehicles (UAVs). To overcome this limitation, we propose a differentiable CKM-triggered trajectory optimization framework for LAWNs. Firstly, we propose a location-oriented CKM construction method that directly maps continuous spatial coordinates to channel gain. In particular, a shared convolutional neural network (CNN) is employed to encode high-level environmental features from conditional inputs. These features are then sampled based on location information to form a fused regressor-conditional multilayer perceptron (c-MLP) or conditional Kolmogorov-Arnold network (cKAN)-for channel gain prediction. We further propose a joint power, bandwidth, and trajectory optimization (JPBTO) method for multi-UAV systems, with the constructed differentiable CKM employed to evaluate the communication performance. The formulated non-convex problem is solved via alternating optimization and successive convex approximation. Numerical results show that the proposed framework enables location-aware differentiability of the CKM, while achieving higher accuracy than the methods without environmental features. Furthermore, the proposed CKM-JPBTO achieves a significantly higher minimum throughput than the conventional statistical channel model-based JPBTO.

The $\mathcal{H}_2$ model reduction problem for high-dimensional linear quantum systems is studied under the constraint of physical realizability (PR). This constraint requires preservation of the canonical commutation relations and the quantum input-output structure, and therefore prevents the direct use of standard projection methods. A symplectic Petrov-Galerkin framework is presented, in which reduced-order models automatically satisfy the PR identities by construction. Within this framework, a symplectic variant of the iterative rational Krylov algorithm is developed and referred to as Quantum IRKA (Q-IRKA). At each iteration, an enriched tangential rational Krylov pool is generated from shifted linear solves. A symplectic basis is then extracted by a Gram-Schmidt-type procedure, paired with symplectic conjugates, and normalized so that the reduced trial space satisfies the canonical symplectic constraint. The interpolation points are updated from selected mirror images of the poles of the current reduced-order model, while the reduced-order matrices are obtained exclusively by structure-preserving projection. Numerical experiments on low-channel oscillator-chain systems and on a bosonic Kitaev-chain-inspired benchmark show that Q-IRKA is effective for large-scale linear quantum systems. Symplecticity and PR are preserved to machine precision, and accurate reduced-order models are obtained with moderate computational cost. The results also show that reduction quality depends substantially on dissipation geometry, channel placement, heterogeneity, and reduced order. These findings indicate that scalable $\mathcal{H}_2$ model reduction of linear quantum systems can be achieved while strictly preserving the underlying physical structure.

We generalize Jan Willems' behavioral approach to a class of discrete-time nonlinear systems in a vector-valued reproducing kernel Hilbert space (RKHS). Apart from linear time-invariant systems, this class covers nonlinear systems modeled by Volterra series and their autoregressive variants, as well as systems admitting Hammerstein-type state-space realizations. We apply the proposed framework to the problem of data-driven modeling of such systems, i.e., when simulation or control objectives for an unknown system are carried out without an explicit system identification step. To that end, we link the behavioral approach to two data-driven modeling methods in a vector-valued RKHS: (1) minimum-norm interpolation and (2) subspace identification.

Data-driven equation discovery is fundamentally an inverse problem that seeks to infer the governing differential equations of a system directly from time-series measurements. A known issue is the ill-conditioned nature of the inverse problem, which frequently produces multiple mathematical models that fit the data similarly well. One path to address this issue is by incorporating known hypotheses and constraints into the training phase beforehand. While this approach effectively reduces the search space, it still results in multiple candidate models, forcing practitioners to rely on post-hoc manual filtering based on their own domain expertise. A recent approach incorporates structural `skeletons' inspired by characteristic curves (CCs), defining a hypothesis-driven methodology. In this methodology, practitioners define a skeleton, which is associated with a family of ordinary differential equations (ODEs), and then add their hypotheses and priors based on their domain knowledge to refine the obtained model iteratively. An important advantage of this approach is that some skeletons have demonstrable structural identifiability properties, which are useful for checking whether the skeleton is correct or should be discarded. Furthermore, this formalism enables the use of multiple equation discovery paradigms due to its modularity (such as neural networks, symbolic regression, and sparse regression). In this work, we present the Python library PyCC, which condenses these efforts into a flexible tool that allows researchers and engineers to seamlessly define their skeletons and hypotheses to discover ODEs from time-dependent data.

Urban flood disaster is one of the most serious natural disasters. Numerous flood simulation models have been proposed and relatively matured. However, two major challenges persist: excessive simplification of the city system and high computational complexity. To break these limitations, this paper develops an Urban Flood Dynamical System Model (UFDSM) based on the concept of the Cellular Automata Urban Flood Model. This model allows flexible customization of cell types and selection of water motion or distribution rules based on actual urban environments to incorporate as much the urban system data as possible. The water motion and distribution rules can be simple, which could reduce the computational complexity, but not arbitrary. So, a sufficient condition is provided so that solutions of dynamical system align with macroscopic physical conditions governing water movement. Then, to preserve the evolutionary properties of the UFDSM, we propose a first-order conservation nonstandard finite difference algorithm. This numerical method ensures positive solutions and conservation of water while maintaining the same fixed-point characteristics as the dynamical system. And, this numerical method is validated by comparing it with an analytical solution.Furthermore, to verify the applicability of our model, we performed an urban flood simulation experiment and compared it to HEC-RAS. There is approximately a 2mm discrepancy in distance dp' and 0.02mm discrepancy in distance d2' , with the relative distance Rp about 7.5% and the relative distance R2 approximately 0.06%. Additionally, the proposed model is easily coupled with other hydrological processes and facilitates data assimilation, thereby offering promising practical applications.

This work studies decentralized novelty detection with global false discovery rate (FDR) control across heterogeneous composite null distributions, without sharing the raw data due to privacy and bandwidth considerations. We propose a framework based on the exchange of quantized surrogate models, allowing independent agents to share low-precision representations of locally learned non-conformity score functions. We prove that evaluating data against these quantized composite scores preserves conditional exchangeability, providing rigorous finite-sample guarantees for global FDR control. Empirical studies on synthetic datasets confirm our theoretical results, demonstrating that the proposed approach maintains competitive statistical power while drastically reducing the communication cost.

Frozen vision foundation models do not merely extract features; they organize images through a learned coordinate system. We ask whether that coordinate system remains internally coherent within a single input. This leads to Representational Consistency: the study of whether a frozen foundation model represents one sample coherently across its semantic subregions. We introduce Dimensional Coactivation (DCA), a per-dimension instrument for measuring this coherence. DCA compares semantic regions by asking whether the same feature dimensions coactivate across them. Unlike classical similarity measures, it deliberately avoids centering, L2 normalization, and full Gram coupling. These operations are useful when comparing different models or distributions, but they are mismatched to the intra-sample setting, where the coordinate system is fixed and raw magnitude carries signal. Deepfake detection provides a natural validation task. Synthetic faces may reproduce plausible eyes, noses, and mouths while breaking the representational structure that links those regions in real faces. Using frozen DINOv3 features, DCA exposes this break: an eyes-mouth-nose fingerprint achieves 0.9106 AUC on CelebDF-v2 and 0.9289 on DFD under FF++ c23 cross-dataset transfer. The design is also sharply validated by ablation: reintroducing centering collapses CelebDF-v2 AUC to 0.459, L2 normalization reduces it to 0.862, and cross-dimension coupling reduces it to 0.478. Finally, replacing DINOv3 with FaRL collapses CelebDF-v2 AUC to 0.582. DCA therefore depends on a stable per-dimension coordinate system, not on region extraction alone. These results position DCA as an instrument for measuring intra-sample representational coherence in frozen foundation models, with deepfake detection as the first validation task.

The interaction of robots with bendable objects in midair presents significant challenges in control, often resulting in performance degradation and potential crashes, especially for aerial robots due to their limited actuation capabilities and constant need to remain airborne. This paper presents an adaptive controller that enables two aerial vehicles to collaboratively follow a trajectory while transporting a bendable object without relying on explicit elasticity models. Our method allows on-the-fly adaptation to the object's unknown deformable properties, ensuring stability and performance in trajectory-tracking tasks. We use Lyapunov analysis to demonstrate that our adaptive controller is asymptotically stable. Our method is evaluated through hardware experiments in various scenarios, demonstrating the capabilities of using multirotor aerial vehicles to handle bendable objects.

This paper details a novel indirect method for solving constrained optimal control problems (OCPs) directly in continuous-time function space. The KKT conditions are embedded in a non-smooth complementarity function, which enables their reformulation as a rootfinding problem in Banach space. This problem is then solved using a non-smooth Newton method. Finally, the paper shows that the Newton update can be obtained by solving a modified differential Riccati equation, where the cost terms are reweighted at every iteration based on the constraint multipliers. Numerical simulations show the effectiveness of the method, which converges superlinearly up to the tolerance of the ODE solver.

Classical optimization theory largely focuses on fixed objective functions, whereas many modern learning systems operate in dynamic environments where data arrive sequentially and decisions must be updated continuously. In this work, we study optimization with streaming data over a distributed network of agents. We adopt a structured, weight-based formulation that explicitly captures the streaming-data origin of the time-varying objective: at each time step, every agent receives a new sample, and the network seeks to track the minimizer of a temporally weighted objective formed from all samples observed across the network so far. We focus on decentralized gradient descent (DGD) with a limited communication/computation budget, where at each time step, only a limited number of DGD iterations can be performed before the objective changes again. For strongly convex and smooth losses, we analyze the tracking error with respect to the time-varying minimizer through a fixed-point theory lens. Our analysis reveals that the tracking error decomposes into a fixed-point tracking term and a bias term induced by data heterogeneity across agents. We specialize the analysis to two natural weighting strategies: uniform weights, which treat all samples equally, and exponentially discounted weights, which geometrically decay the influence of older data. Under uniform weighting, DGD tracks the fixed-point at a rate $\mathcal{O}(1/t)$, whereas discounted weighting yields a non-vanishing fixed-point tracking floor controlled by the discount factor. In both cases, decentralization induces an additional non-zero bias floor under a constant step size. We validate our theoretical findings through numerical simulations.

We propose a novel receiver architecture that preserves the performance benefits of multiport selection in fluid-antenna systems while requiring only a very small number of radio-frequency (RF) chains. The resulting fluid-antenna hybrid multiport (FAHM) receiver effectively decouples port selection from signal combining by integrating a low-complexity analog combining network similar to those used in conventional hybrid multiantenna designs. We develop an effective-port criterion to determine the selected-port dimension before hybrid combining, thereby limiting the performance loss associated with port selection, and then design the hybrid combiner for a given RF-chain budget. The FAHM architecture is evaluated in a multiuser set-up operating under slow fluid-antenna multiple access (FAMA). In this scenario, a FAHM implementation with only 2 RF chains showcases a performance comparable to a fully-digital conventional multiport scheme with a much larger number of RF chains. Additionally, the proposed Rank-1 implementation attains a reduction in average GEPort execution time by about two orders of magnitude.

Recently, true-time-delay (TTD) arrays, also referred to as joint phase-time arrays (JPTA), have been investigated for low-cost frequency-dependent beamforming capabilities to enable various applications, including beam-squint correction, fast beam training, and serving multiple user equipment (UE)s by frequency band to direction mapping, termed as split beampatterns. Several heuristics and optimization-based solutions have been proposed to determine TTD array parameters settings. However, they have practical limitations due to either computationally demanding optimization procedures, requirements for extremely large memory look-up tables, or degradations due to the beam-squint effect. In this article, we propose a novel split-beampattern generation algorithm based on the observed homomorphism between TTD array configuration matrices and corresponding beampatterns. First, we rigorously analyze the beampattern synthesis process and demonstrate the observed homomorphism and mathematical structure. Then, we propose the Homomorphic Directional Beamforming (HDB) algorithm to approximate the desired split beampatterns by utilizing a generator beampattern dictionary that requires a dictionary size orders of magnitude lower than existing approaches without ignoring the beam squint. With extensive simulations, we show that the proposed algorithm can provide a practical implementation with low memory and low computational cost requirements. In addition, HDB design provides close to uniform beamforming gains among UEs in different subbands, enabling fairness in power allocation.

The rise of Internet of Things (IoT) devices in the physical world necessitates voice-based interfaces capable of handling complex user experiences. While modern Large Language Models (LLMs) already demonstrate strong tool-usage capabilities, modeling real-world IoT devices presents a difficult, understudied challenge which combines modeling spatiotemporal constraints with speech inputs, dynamic state tracking, and mixed-initiative interaction patterns. We introduce MIST (the Multimodal Interactive Speech-based Tool-calling Dataset), a synthetic multi-turn, voice-driven code generation task that operates over IoT devices. We find that there is a significant gap between open- and closed-weight multimodal LLMs on MIST, and that even frontier closed-weight LLMs have substantial headroom. We release MIST and an extensible data generation framework to build related datasets in order to facilitate research on mixed-initiative voice assistants which reason about physical world constraints.

This paper presents CARMEN, a runtime-adaptive, CORDIC-accelerated multi-precision vector engine for resource-efficient deep learning inference. The key insight is that CORDIC iteration depth directly governs computational accuracy, enabling dynamic switching between approximate and accurate execution modes without hardware modification. The architecture integrates a low-resource iterative CORDIC-based MAC unit with a time-multiplexed multi-activation function block, supporting flexible 8/16-bit precision and high hardware utilization. ASIC implementation in 28 nm CMOS achieves up to 33% reduction in computation cycles and 21% power savings per MAC stage; a 256-PE configuration delivers 4.83 TOPS/mm2 compute density and 11.67 TOPS/W energy efficiency. FPGA deployment on PynqZ2 validates 154.6 ms latency at 0.43 W for real-time object detection.

Advanced driver-assistance systems (ADAS) require neural compute engines that deliver low-latency inference under strict power and area constraints. Posit arithmetic is attractive for such accelerators because it provides high numerical fidelity at low precision, but its variable-length regime encoding increases encode/decode cost and exposes the datapath to large regime-field fault effects. This paper presents EULER-ADAS, a SIMD-enabled logarithmic bounded-Posit neural compute engine for energyefficient and reliability-aware ADAS acceleration. The proposed datapath combines bounded-regime Posit representation, stageadaptive logarithmic mantissa multiplication with bit truncation, and a SIMD-shared quire accumulation path supporting Posit- (8,0), Posit-(16,1), and Posit-(32,2) execution. The unified architecture enables 4xPosit-8, 2xPosit-16, or 1xPosit-32 operation without duplicating precision-specific hardware. FPGA implementation shows that the proposed configurations reduce LUT count by up to 41.4%, delay by up to 76.1%, and power by up to 71.9% relative to exact Posit neural compute engines, while achieving up to 10x lower energy-delay product than radix-4 Booth-based Posit multipliers. In 28-nm CMOS, the bounded variants occupy 0.013-0.016 mm2 , consume 19.8-22.1 mW, and operate at up to 1.84 GHz. Application-level evaluation across image-classification, ADAS, and edge-inference workloads shows that the evaluated Posit-16 and Posit-32 configurations remain within about 1.5 percentage points of FP32 accuracy. A TinyYOLOv3 prototype on Pynq-Z2 achieves 78 ms latency at 0.29 W and 22.6 mJ/frame, demonstrating the suitability of EULERADAS for low-power real-time ADAS inference.

Fluid antenna (FA) technology has recently emerged as an effective means of exploiting spatial diversity through position-domain reconfigurability. This paper investigates the integration of FA into over-the-air federated learning (OTA-FL) systems with the aim of improving aggregation reliability and user participation under realistic channel conditions. By dynamically selecting antenna positions, FA-equipped users can exploit additional spatial degrees of freedom to realize more favorable channel conditions, thereby increasing the probability of successful contribution to the OTA aggregation process in each communication round. We consider an uplink OTA-FL framework consisting of a single fixed-antenna access point and multiple FA-enabled users operating over spatially correlated fading channels. Unlike existing studies that primarily rely on optimization-based designs or numerical evaluations, we develop a tractable analytical framework that enables a rigorous performance characterization of FA-assisted OTA-FL. In particular, closed-form expressions are derived for the aggregation error outage probability and the expected number of participating users per round. Spatial channel correlation across FA ports is modeled using a copula-based approach, where the Clayton copula is adopted to capture lower-tail dependence relevant to worst-case fading conditions. Numerical results validate the analytical findings and demonstrate that FA-assisted OTA-FL significantly outperforms conventional fixed-antenna schemes in terms of aggregation reliability and participation efficiency, while providing insights under practical system considerations.

We present the design, fabrication, and measurement of a monolithically integrated optical receiver analog front end, where low power operation is a primary consideration with a goal of supporting 56 Gbaud intensity modulated direct detect transceivers. The need for low-power consumption and low-noise operation motivates a monolithic, layout driven design approach which begins with circuit topology selection and analysis. Various transistor unit cell layout configurations are explored, minimizing parasitics, enabling wide analog bandwidth and reduced input referred noise. The post-layout analog front end achieves a 28.9 GHz bandwidth with a low-frequency gain of 61.7 dB{\Omega}. This circuit was designed within the GlobalFoundries FotonixTM monolithic silicon photonics platform. The fabricated device is characterized by its DC operation, noise characteristics, and time domain behavior. The final design was validated by on-off keyed and PAM-4 electrical eye diagram measurements to 64 GBaud, consuming 9.22 mW of power from a 1.2 V supply with less than 737 nA RMS integrated input referred noise current and 0.08 pJ/bit.

We propose a novel extremum seeking control (ESC) method that operates in a lifted Koopman state space to minimize the filtered RMS energy in the dominant subspace. The lifted representation provides linear embeddings of nonlinear dynamics, enabling more accurate gradient estimation and dampening of state interference for more consistent ESC performance. Applied to a parameterized, forced, and time-varying Van der Pol oscillator, we show that the approach yields faster and more robust performance than operating ESC on the measured states. These advantages position the method for a diverse range of applications including vibration suppression, motion control, and subsynchronous oscillation mitigation in inverter-dominated power systems.

Real-time supervisory control of thermal energy distribution systems requires digital twins that are accurate, interpretable, and uncertainty-aware, yet remain data and computationally efficient. High-fidelity simulations alone are costly, while purely data-driven surrogates often lack robustness. To address these challenges, this work proposes an active learning (AL) framework that couples system-level Modelica simulations with four simpler physics-informed and data-driven surrogate modeling approaches: deterministic Sparse Identification of Nonlinear Dynamics with Control (SINDyC), its probabilistic multivariate-Gaussian extension (MvG-SINDyC), feedforward neural network (FNN), and gated recurrent unit (GRU) network. Tailored to each surrogate, model-specific AL query strategies are employed, including Mahalanobis-distance sampling in coefficient space for MvG-SINDyC and error-based sampling in prediction space for SINDyC, FNN, and GRU, allowing the learning process to prioritize dynamically informative trajectories. The proposed approach is demonstrated on the glycol heat exchanger (GHX) subsystem of the Thermal Energy Distribution System (TEDS) at Idaho National Laboratory. Across key GHX outputs–the bypass mass flow rate $\dot{m}_{\mathrm{GHX}}$ and heat transfer rate $Q_{\mathrm{GHX}}$-the AL framework achieves comparable predictive accuracy using as few as one-fifth of the simulation trajectories required by random sampling. Among the evaluated surrogates, the GRU achieves the highest predictive fidelity, while SINDyC remains the most computationally efficient and interpretable. The probabilistic MvG-SINDyC surrogate further enables uncertainty quantification and exhibits the largest computational gains under AL.

We introduce a hybrid quantum-classical pipeline, based on neutral-atom reservoir computing, for medical image classification, focusing on the binary classification task of polyp detection. To deal effectively with the high dimensionality, we integrate a guided auto-encoder. This pipeline learns compact and discriminative representations of image data that are also well-suited for quantum reservoir computing. A key challenge in such systems is the non-differentiable nature of quantum measurements, which creates a 'gradient barrier' for standard training. We overcome this barrier by incorporating a differentiable surrogate model that emulates the quantum layer, enabling end-to-end backpropagation through the entire system. This guided training process is jointly optimized for classification accuracy and for faithful image recovery from the auto-encoder. The learned latent representations are encoded as pulse detuning parameters within a Rydberg Hamiltonian, and quantum embeddings are subsequently obtained through expectation values. These embeddings are then passed to a linear classifier. Our simulations show that this method outperforms some traditional approaches that use PCA or unguided autoencoders. We also conduct ablation studies to assess the impact of various quantum and training parameters, demonstrating the robustness and flexibility of our proposed pipeline for real-world medical imaging applications, even in the current NISQ era.

Denoising electroencephalogram (EEG) is an inherently challenging task, since neural activity is not only subtle but also inseparable from spectrally overlapping noise artifacts. Today, effective EEG denoising is more important than ever, given the rapid adoption of wearables across various applications. Deep learning methods have shown promising results in decomposition-free denoising that handles the time-varying pervasive EEG artifacts. However, training highly expressive neural networks requires artifact-free EEG, which is inherently unobtainable. To address this, we propose Intelligent Partitioning for Self-supervised Denoising (iPSD). Our method eliminates the need for clean references by learning to partition an input EEG segment into independent noisy realizations with the same underlying signal. This enables self-supervision of deep learning denoisers, even in zero-shot settings where only a single EEG segment to be denoised is available. We validate iPSD through extensive experiments, including validations on wearable EEG from in-ear sensors. The results show that iPSD achieves state-of-the-art performance, most notably under extremely low signal-to-noise ratios (down to -10 dB) and challenging artifacts (e.g., EMG), with spectral fidelity orders of magnitude higher than competitive baselines.

Large audio language models (LALMs) increasingly reason over long audio clips, motivating audio compression to reduce memory use and inference latency. However, audio compression can leave the model's overall answer accuracy acceptable while severely degrading accuracy for particular query families. We introduce a framework for auditing a given audio compression method by measuring the increase in an LALM's answer error relative to uncompressed audio. We formalize an acceptance criterion that limits this increase for every required query family and derive a practical protocol for testing compression settings with statistical confidence. Experiments spanning five multiple-choice audio question-answering benchmarks and four LALMs expose compression-induced damage hidden by overall accuracy.

Safety-critical autonomy in adversarial settings demands more than Lyapunov stability of tracking error signals. An agent executing a goal-directed trajectory is intrinsically legible to a passive observer running online Bayesian inference, because the contractive dynamics of any Lyapunov basin of attraction concentrates posterior belief over the latent intent parameters. We initiates the study of intent privacy over a continuous state space as a joint control problem on the physical state combined with the latent belief state of a putative observer. With the main challenges concentrated around the analysis of the belief-state dynamics, the agent dynamics is assumed to be simple, modeled by the differential inclusion $\dot{x}\in u+\bar{d}\mathbb{B}$. That is, the agent is fully actuated with bounded unknown disturbance to the control input. The observer's intent inference process is modeled as a discrete-time stochastic dynamical system evolving over the belief state space of a Rao Blackwellized particle filter reasoning over large random samples of possible agent goals. The agent's control input is modeled as a piecewise constant signal, with jumps matching the RBPF update times. Building on a prior intent-inference framework and its KL-based information leakage measurement, a privacy constraint is imposed, which amounts to maintaining information leakage above a prescribed threshold with high probability, using probabilistic discrete-time control barrier functions. A key technical contribution is the derivation of separate PCBF results for the Bayesian update step and the resampling step of the RBPF, enabling a PCBF result for the full update as well as integration of the privacy constraint with the agent's task-side tracking requirement. Finally, a joint feasibility analysis is carried out by examining the interplay between the privacy constraint and the tracking envelope.

Modern sensors generate rich, high-fidelity data, yet applications operating on wearable or remote sensing devices remain constrained by bandwidth and power budgets. Standardized codecs such as JPEG and MPEG achieve efficient trade-offs between bitrate and perceptual quality but are designed for human perception, limiting their applicability to machine-perception tasks and non-traditional modalities such as spatial audio arrays, hyperspectral images, and 3D medical images. General-purpose compression schemes based on scalar quantization or resolution reduction are broadly applicable but fail to exploit inherent signal redundancies, resulting in suboptimal rate-distortion performance. Recent generative neural codecs, or tokenizers, model complex signal dependencies but are often over-parameterized, data-hungry, and modality-specific, making them impractical for resource-constrained environments. We introduce a Lightweight, Versatile, and Asymmetric neural codec architecture (LiVeAction), that addresses these limitations through two key ideas. (1) To reduce the complexity of the encoder to meet the resource constraints of the execution environments, we impose an FFT-like structure and reduce the overall size and depth of the neural-network-based analysis transform. (2) To allow arbitrary signal modalities and simplify training, we replace adversarial and perceptual losses with a variance-based rate penalty. Our design produces codecs that deliver superior rate-distortion performance compared to state-of-the-art generative tokenizers, while remaining practical for deployment on low-power sensors. We release our code, experiments, and python library at https://github.com/UT-SysML/liveaction .

Weight decay is widely used as a regularizer in large language models, yet its precise role in shaping Transformer loss landscapes remains theoretically underexplored. This paper provides the first rigorous functional-analytic characterization of the standard Transformer objective–cross-entropy loss with $L^2$ regularization–by proving it satisfies Villani's criteria for coercive energy functions. Specifically, we show that the regularized loss $\mathcal{F}$ is infinitely differentiable, grows at least quadratically, has Gaussian-integrable tails, and satisfies the differential growth condition $-\Delta\mathcal{F} + \tfrac{1}{s}\|\nabla\mathcal{F}\|^{2} \to \infty$ as $\|\theta\| \to \infty$ for all $s>0$. From this structure, we derive explicit log-Sobolev and Poincaré constants $C_{\mathrm{LS}} \leq \lambda^{-1} + d/\lambda^{2}$, linking the regularization strength $\lambda$ and model dimension $d$ to finite-time convergence guarantees for noisy stochastic gradient descent and PAC-Bayesian generalization bounds that tighten with increasing $\lambda$. To validate our theory, we introduce a scalable Villani diagnostic $\Psi_s(\theta) = -\Delta \mathcal{F} + s^{-1}\|\nabla \mathcal{F}\|^2$ and estimate it efficiently using Hutchinson trace probes in models with over 100M parameters. Experiments on GPT-Neo-125M across Penn Treebank and WikiText-103 confirm the predicted quadratic growth of $\Psi_s$, spectral inflation of the Hessian, and exponential convergence behavior consistent with our log-Sobolev analysis. These results demonstrate that weight decay not only improves generalization empirically but also establishes the mathematical conditions required for fast Langevin mixing and theoretically grounded curvature-aware optimization in deep learning.

Modern learning systems represent perceptual signals with continuous vectors, but comparison, retrieval, memory, alignment, and reasoning are often symbolic. In language, tokens provide this interface; for speech and audio, it must be learned. Existing audio tokenizers rely on local quantization, clustering, or reconstruction, leaving sequence consistency, compactness, length, termination, and edit geometry only indirectly controlled. We introduce PairAlign, a framework for compact audio tokenization through autoregressive self-alignment. An encoder maps speech to a continuous condition, and an autoregressive decoder emits tokens from BOS to EOS. Given two content-preserving views, PairAlign derives a canonical anchor target and trains both views to predict it, with unrelated in-batch targets as competing sequences. It first learns an autoregressive bridge from VQ targets and then transitions to EMA-teacher self-alignment with grounding and anti-collapse controls. On 3 s speech, PairAlign produces shorter, non-degenerate sequences while retaining ordered paired-view consistency. On the TIMIT retrieval archive, it operates at 8.28 tokens/s, reducing archive token count by 70.65% and the edit-distance work proxy by 91.87% relative to Stage I. Positive-negative probes show strong separation across phone-disjoint, trigram-disjoint, and temporally rearranged negatives, while rate-controlled post-hoc BPE does not recover the same compactness-consistency operating point. These results expose a rate-granularity trade-off: PairAlign does not uniformly outperform denser tokenizers on every local metric, but provides a lower-rate symbolic interface preserving ordered and relational structure. Conceptually, PairAlign follows JEPA-style predictive learning by predicting an abstract target rather than reconstructing the input; here, the target is a learned variable-length symbolic sequence.

Accurate Channel State Information (CSI) prediction is essential for dynamic multiple-input multiple-output (MIMO) systems but remains computationally demanding. This letter proposes a resource-efficient predictor that combines a gated recurrent unit (GRU) encoder with Luong attention, a bottleneck gated fusion module, and a Dimension-wise Separable Linear Head (DSLH). The gated fusion module integrates local recurrent features with global attention context, while the DSLH reduces the cost of the output mapping. Evaluated on 3GPP TR 38.901-compliant channels, the proposed model achieves an average NMSE of -13.84 dB with 26% fewer parameters and approximately 2.3x higher inference throughput than a dimension-matched LinFormer baseline. The proposed model is best suited to LOS and mixed-condition scenarios, offering a practical accuracy-efficiency trade-off for short-horizon CSI prediction at moderate sequence lengths.

We present a statistics-aware compression strategy that processes photon timestamps directly from time-correlated single-photon counting (TCSPC) modules for time-domain fluorescence lifetime imaging (FLIM). Rather than storing or transmitting the full histogram per pixel, timestamps are projected onto sparse, non-uniform one-dimensional spline sketches, with knot positions optimally allocated based on Fisher information. This knot allocation concentrates sketch channels where the decay signal exhibits the greatest statistical discriminability, rather than using a uniform allocation. The proposed approach is extensively validated on synthetic mono- and bi-exponential decay data and on experimental fluorescent dye data, demonstrating comparable accuracy to full-histogram non-linear least-squares fitting (NLSF) and Poisson maximum-likelihood estimation (MLE) at compression ratios of up to 256x. We further validate the feasibility of integrating the timestamp-to-sketch projection directly into firmware via fixed-point (FXP) lookup-table (LUT) simulation, targeting high-spatial-resolution single-photon avalanche diode (SPAD) arrays subject to significant data-throughput constraints.

In this paper, we describe procedures for computing higher-order time derivatives of the Lie-group Newton-Euler, Articulated-Body Inertia, and hybrid dynamics algorithms for floating-base trees, where the base configuration evolves on SE(3) and the attached mechanism is an open kinematic tree with configuration on the (n1+n2)-dimensional manifold T^{n1} \times R^{n2}, using spatial representation of twists. After presenting the algorithms, we collect the resulting recursions into closed-form equations of motion, identifying an admissible Coriolis matrix satisfying the passivity property, and showing that the articulated inertia tensor remains unchanged across all time derivatives. We then apply the developed methods to a 12-DoF aerial manipulator to derive analytical expressions for its geometric forward and inverse dynamics along with their first time derivatives whereas the numerical simulations successfully evaluate these dynamics up to fifth order. Finally, to demonstrate their practical utility, we benchmark the proposed extensions and show that, in the considered tests, their computational cost scales quadratically with the derivative order, whereas the automatic-differentiation baseline exhibits exponential scaling.