arXiv:2309.11965v4 Announce Type: replace Abstract: In this paper, coordination control of discrete event systems under joint sensor and actuator attacks is investigated. Sensor attacks are described by a set of attack languages using a proposed ALTER model. Several local supervisors are used to control the system. The goal is to design local supervisors to ensure safety of the system even under cyber attacks (CA). The necessary and sufficient conditions for the existence of such supervisors are derived in terms of conditional decomposability, CA-controllability and CA-observability. A method is developed to calculate local state estimates under sensor attacks. Two methods are also developed to design local supervisors, one for discrete event systems satisfying conditional decomposability, CA-controllability and CA-observability, and one for discrete event systems satisfying conditional decomposability only. The approach works for both stealthy and non-stealthy attacks. A practical example is given to illustrate the results.
Science Journals
arXiv:2605.18035v1 Announce Type: new Abstract: Hard-thresholding is an important type of algorithm in machine learning that is used to solve $\ell_0$ constrained optimization problems. However, the true gradient of the objective function can be difficult to access in certain scenarios, which normally can be approximated by zeroth-order (ZO) methods. The SZOHT algorithm is the only algorithm tackling $\ell_0$ sparsity constraints with ZO gradients so far. Unfortunately, SZOHT has a notable limitation on the number of random directions % in ZO gradients due to the inherent conflict between the deviation of ZO gradients and the expansivity of the hard-thresholding operator. This paper approaches this problem by considering the role of variance and provides a new insight into variance reduction: mitigating the unique conflicts between ZO gradients and hard-thresholding. Under this perspective, we propose a generalized variance reduced ZO hard-thresholding algorithm as well as the generalized convergence analysis under standard assumptions. The theoretical results demonstrate the new algorithm eliminates the restrictions on the number of random directions, leading to improved convergence rates and broader applicability compared with SZOHT. Finally, we illustrate the utility of our method on a ridge regression problem as well as black-box adversarial attacks.
arXiv:2605.18067v1 Announce Type: new Abstract: Deploying large language model (LLM) on edge device enables personalized LLM agents for various users. The growing availability of diverse personalized agents presents a unique opportunity for peer-to-peer (P2P) collaboration, wherein each user can delegate tasks beyond the local agent's expertise to remote agents more suited for the specific query. This paper introduces PPAI, the first personalized LLM agent interoperability system, which enables users to collaborate with each other based on agent specialization. However, the ever-changing pool of agents and their interchangeable capacity introduce new challenges when it comes to matching queries to agents and balancing loads, compared with existing P2P systems. Therefore, we propose a scalable query-agent pair scoring mechanism based on prototypes to identify suitable agents within a P2P network with churn. Moreover, we propose a multi-agent interoperability Bayesian game to balance local demand and global efficiency, when changes in remote agent load occur too quickly to be observed. Finally, we implement a prototype of PPAI and demonstrate that it substantially broadens the range of tasks that could be carried out while maintaining load balance. On average, it achieves an accuracy improvement of up to 7.96% across multiple tasks, while reducing latency by 16.34% compared to the baseline.
arXiv:2605.18072v1 Announce Type: new Abstract: Detecting AI-generated music is crucial for preserving artistic authenticity and preventing the misuse of generative music technologies. However, existing discriminative detectors typically rely on generated samples during training and often suffer from severe performance degradation when confronted with music produced by unseen generators, which limits their real-world applicability. To address this issue, we formulate a zero-shot setting for AI-generated music detection, where the detector is trained exclusively on real music without access to any generated samples. Under this setting, we propose MusicDET, a generator-agnostic detection framework based on frequency-guided normalizing flows that probabilistically models the distribution of real music features. By evaluating the likelihood of an input sample under the learned real-music distribution, MusicDET enables effective detection of out-of-distribution music signals. Experiments on the FakeMusicCaps and SONICS datasets show that MusicDET consistently outperforms conventional discriminative detectors, particularly when detecting music generated by previously unseen models.
arXiv:2605.18096v1 Announce Type: new Abstract: Offset curves for planar trajectories are interesting in the generation of tool paths for numerically controlled industrial machines and in trajectory planning methods for autonomous driving systems. Theoretical offset curves may exhibit peculiar singularities, including self-intersections, which limit their use in practical applications. Existing approaches address these issue through geometric filtering techniques to detect and remove undesirable features but the computation of accurate and well-behaved offset curves remains a challenging task. We assume a first stage of functional approximation of trajectories by penalized Hermite spline regression enabling the simultaneous fitting of positions and tangents. The regularization is imposed on the second derivatives, effectively mitigating the jerk effect, which is particularly relevant in motion planning and path smoothing applications. Then, taking into account the geometrical pointwise properties of the resulting curve, we design two offset curves through the simultaneous approximation of function values and derivatives. Then, a mathematical model to obtain the so-called bi-offset as most fitting as with the original generator curve is proposed, also relating the offset range and pointwise curvature values. The adaptive reconstruction of the center line from the external boundaries is a topic of interest and is the main focus of our work. Numerical experiments confirm the reliability of our approach at every stage of the resolution process.
arXiv:2605.18111v1 Announce Type: new Abstract: Recent advancements in Large Language Models (LLMs) and Large Vision Language Models (LVLMs) have enabled general-purpose systems to demonstrate promising capabilities in complex reasoning tasks, including those in the medical domain. Medical Visual Question Answering (MedVQA) has particularly benefited from these developments. However, despite Bangla being one of the most widely spoken languages globally, there exists no established MedVQA benchmark for it. To address this gap, we introduce BanglaMedVQA, a dataset comprising clinically validated image-question-answer pairs, along with a comprehensive evaluation of current foundation models on this resource. Consistent with prior findings that report low performance of current models on English MedVQA benchmarks, our analysis reveals that Bangla performance is substantially lower, reflecting the challenges inherent to low-resource languages. Even top-performing models such as Gemini and GPT-4.1 mini fail to accurately answer specialized diagnostic questions, indicating severe limitations in fine-grained medical reasoning. Although certain open-source models, such as Gemma-3, occasionally outperform these models in general categories, they too struggle with clinically complex questions, underscoring the urgent need for top-notch evaluation method.
arXiv:2605.17359v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems have shown strong potential for complex reasoning by coordinating specialized agents through structured communication. However, existing topology-evolution methods typically construct or optimize a collaboration topology for each query from scratch, leading to substantial online search overhead, high inference-time token consumption, and limited scalability in multi-domain settings. We propose TopoPrior, a framework for learning transferable topology priors for multi-agent LLM collaboration across domains. Rather than repeatedly searching for effective collaboration structures online, TopoPrior learns reusable topology priors from reference collaboration graphs collected offline from multiple domains and uses them to generate query-conditioned initial collaboration graphs for downstream refinement. By shifting part of topology search from per-query online optimization to offline prior learning, TopoPrior amortizes search cost while remaining compatible with existing topology-evolution backbones. Technically, TopoPrior contains two key components. First, a transferable topology prior learning module employs a conditional variational graph framework to capture reusable structural regularities across domains in a latent space. Second, a query-conditioned latent adaptation module introduces adversarial alignment to reduce unnecessary domain discrepancy while preserving query-relevant structural variation. Experiments on multi-domain reasoning benchmarks show that TopoPrior consistently improves several heterogeneous topology-evolution backbones while reducing online inference-time token usage, with only modest additional trainable parameters. These results suggest that transferable topology initialization is an effective and lightweight mechanism for improving the efficiency of multi-agent LLM collaboration across domains.
arXiv:2605.18157v1 Announce Type: new Abstract: We introduce a class of cooperative games induced by weighted directed graphs. Specifically, the coalitional value combines an internal interaction term given by the induced subgraph game with an external component based on minimal incoming edges from outside the coalition. The resulting game has a convenient representation in terms of unanimity games. This representation enables closed-form polynomial-time formulas for the Shapley and Banzhaf values. We further establish that the game has a nonempty core and is totally balanced. The class of such games therefore provides an analytically and computationally tractable example of structured network- induced cooperative games in which stability-based allocations and fairness-based solution concepts do not coincide.
arXiv:2605.18177v1 Announce Type: new Abstract: Query-based Vision Transformer segmentation models typically reconstruct dense spatial feature maps to predict masks, inheriting design patterns from convolutional architectures. We show that this explicit image-space reconstruction is not required. We introduce TokenMask, a token-space mask head that computes mask logits directly from query-token affinities and performs interpolation in logit space rather than feature space. This reformulation preserves the original linear scoring mechanism while simplifying the computational structure. Across diverse ViT backbones, datasets and segmentation tasks, TokenMask consistently improves efficiency over prior approaches by reducing computational and memory requirements while maintaining competitive accuracy, leading to tangible speedups on NVIDIA Jetson AGX Orin using TensorRT FP16 inference. Overall, TokenMask yields a simpler and more deployment-friendly design for embedded vision systems.
arXiv:2304.10726v3 Announce Type: replace Abstract: We introduce the Deep Learning Vulnerability Analyzer (DLVA) for Ethereum smart contracts based on neural networks. We train DLVA to judge bytecode even though the supervising oracle can only judge source. DLVA's training algorithm is general: we extend a source code analysis to bytecode without any manual feature engineering, predefined patterns, or expert rules. DLVA's training algorithm is also robust: it overcame a 1.25% error rate mislabeled contracts, and--the student surpassing the teacher--found vulnerable contracts that Slither mislabeled. DLVA is much faster than other smart contract vulnerability detectors: DLVA checks contracts for 29 vulnerabilities in 0.2 seconds, a 10-1,000x speedup. DLVA has three key components. First, Smart Contract to Vector (SC2V) uses neural networks to map smart contract bytecode to a high-dimensional floating-point vector. We benchmark SC2V against 4 state-of-the-art graph neural networks and show that it improves model differentiation by 2.2%. Second, Sibling Detector (SD) classifies contracts when a target contract's vector is Euclidian-close to a labeled contract's vector in a training set; although only able to judge 55.7% of the contracts in our test set, it has a Slither-predictive accuracy of 97.4% with a false positive rate of only 0.1%. Third, Core Classifier (CC) uses neural networks to infer vulnerable contracts regardless of vector distance. We benchmark DLVA's CC with 10 ML techniques and show that the CC improves accuracy by 11.3%. Overall, DLVA predicts Slither's labels with an overall accuracy of 92.7% and associated false positive rate of 7.2%. Lastly, we benchmark DLVA against nine well-known smart contract analysis tools. Despite using much less analysis time, DLVA completed every query, leading the pack with an average accuracy of 99.7%, pleasingly balancing high true positive rates with low false positive rates.
arXiv:2405.14657v2 Announce Type: replace Abstract: Preferential Bayesian optimization (PBO) learns latent utilities from pairwise comparisons, but most existing methods assume homoscedastic comparison noise. This is inadequate in human-in-the-loop settings, where a user may compare some designs reliably and others only hesitantly. We propose a heteroscedastic noise model for PBO: before optimization, the user provides a small set of reliable examples, called anchors, and a kernel density estimator (KDE) turns these anchors into an input-dependent map of user uncertainty. We incorporate this map into preferential GP surrogates and derive risk-averse acquisition functions that trade off utility and ease of comparison. We further show that a risk-adjusted variant of the popular expected utility of the best option (EUBO) preserves the one-step Bayes-optimality guarantee up to an additive constant, and that under an idealized i.i.d. anchor model the KDE estimator enjoys standard consistency and concentration rates. Experiments on synthetic problems and human-preference datasets show improved risk-adjusted performance and clarify how anchor placement affects the method.
arXiv:2506.15588v2 Announce Type: replace Abstract: Differential privacy (DP) protects sensitive data during neural network training, but standard methods like DP-Adam suffer from high memory overhead due to per-sample gradient clipping, limiting scalability. We introduce DP-GRAPE (Gradient RAndom ProjEction), a DP training method that significantly reduces memory usage while maintaining utility on par with first-order DP approaches. DP-GRAPE is motivated by our finding that privatization flattens the gradient singular value spectrum, making SVD-based projections (as in GaLore (Zhao et al., 2024)) unnecessary. Consequently, DP-GRAPE employs three key components: (1) random Gaussian matrices replace SVD-based subspaces, (2) gradients are privatized after projection, and (3) projection is applied during backpropagation. These contributions eliminate the need for costly SVD computations, enable substantial memory savings, and lead to improved utility. Despite operating in lower-dimensional subspaces, our theoretical analysis shows that DP-GRAPE achieves a privacy-utility tradeoff comparable to DP-SGD. Our extensive empirical experiments show that DP-GRAPE can significantly reduce the memory footprint of DP training without sacrificing accuracy or training time. In particular, DP-GRAPE reduces memory usage by over 63% when pre-training Vision Transformers and over 70% when fine-tuning RoBERTa-Large as compared to DP-Adam, while achieving similar performance. We further demonstrate that DP-GRAPE scales to fine-tuning large models such as OPT with up to 6.7 billion parameters, a scale at which DP-Adam fails due to memory constraints. Our code is available at https://github.com/alexmul1114/DP_GRAPE.
arXiv:2605.18262v1 Announce Type: new Abstract: Accurate pedestrian trajectory prediction is crucial for autonomous systems operating in complex environments, such as modular buses and delivery robots in suburban or semi-structured areas. Social Spatio-Temporal Graph Convolutional Neural Networks (Social-STGCNN) have shown strong performance by modeling social interactions; however, producing diverse and well-calibrated future trajectories remains challenging. In this work, we build on a Social-STGCNN backbone and introduce a Conditional Variational Autoencoder (CVAE)-based probabilistic formulation to explicitly model multimodal future trajectories. We evaluate the method on the ETH and UCY pedestrian trajectory datasets as well as on a real-world pedestrian dataset collected by a mobile robot. Results show moderate gains on public benchmarks, but more consistent endpoint accuracy and improved trajectory diversity across different crowd configurations. Evaluation on robot-collected data further demonstrates the approach's effectiveness beyond curated benchmarks and supports its applicability in practical deployments.
arXiv:2605.18272v1 Announce Type: new Abstract: Non-Hermitian lattices can host the non-Hermitian skin effect, a boundary-induced collapse of all bulk eigenstates into exponentially localized edge modes. This effect underlies anomalous bulk-boundary correspondence and remarkable enhancements in non-Hermitian sensing, yet direct energy-resolved access to the eigenmodes of non-Hermitian lattices has remained limited. Here we report band- and energy-resolved eigenmode spectroscopy of skin modes in a frequency synthetic dimension. By introducing strong frequency-domain boundaries in an electro-optically modulated ring resonator, we realize finite non-Hermitian lattices and use laser detuning as a spectroscopic axis for the eigenenergies of the effective Hamiltonian. Site-resolved heterodyne measurements then reconstruct the spatial profile of each mode, revealing boundary-localized skin states throughout the spectrum and their eigenenergy-dependent displacement from the edge. Beyond 1D, the same frequency-boundary architecture, upon incorporating long-range couplings between finite lattices, produces genuine 2D frequency lattices rather than the hitherto-realized folded 1D systems on twisted tubes. In these lattices we observe tunable directional transport and edge localization in two synthetic dimensions. Our results introduce eigenmode spectroscopy as a direct probe of non-Hermitian physics and establish strongly bounded frequency lattices as a flexible platform for Hamiltonian engineering.
arXiv:2605.18292v1 Announce Type: new Abstract: This paper presents a method that learns a regionally stable recurrent neural network model from a set of input-output data generated by an unknown dynamical system. Relying on generalized sector conditions on the deadzone activation function, we first derive sufficient conditions that guarantee forward invariance on a compact set of the state space for any inputs from a given set. Such regional properties lead to less conservative conditions compared to variants that offer a global form of stability, and are in line with the system data that is only observed regionally. Our learning method derives conditions for regional stability using a barrier function approach, leading to models equipped with a certificate of regional stability in a subset of the state space and for a given input set. We illustrate our theoretical result with a numerical example and compare it to methods that impose a global form of stability, which fail to identify the system, and with a method that imposes no stability constraints at all, which does not guarantee a stable behavior within any state or input set.
arXiv:2605.18295v1 Announce Type: new Abstract: Robust pedestrian safety is crucial to the next-generation of intelligent transportation systems. Such systems rely on active pedestrian localization and predictive collision alerts. Pedestrian localization can be supported by Ultra-Wideband technology and Bluetooth 6.0, which offer high-precision ranging and low-latency communication, making them promising candidates for vehicular collision warning systems. This paper assesses the localization accuracy of these technologies for pedestrian alerting and benchmarks their performance against Global Navigation Satellite Systems. Experimental evaluations performed in this paper focused on key performance metrics, including localization accuracy and robustness to environmental conditions. Preliminary results suggest that Ultra-Wideband and Bluetooth 6.0 can serve as viable alternatives or complements to Global Navigation Satellite Systems in certain scenarios, improving situational awareness and enabling timely pedestrian alerts.
arXiv:2502.05462v2 Announce Type: replace Abstract: We propose a real-time implementable motion planning framework for cooperative object transportation by nonholonomic mobile manipulator robots (MMRs) in dynamic environments. Our global planner finds a path from start to goal through the static, obstacle-free regions in the environment and generates a set of convex, static, obstacle-free regions around the path using a novel, fast, and computationally lightweight ellipse-based technique. We introduce a nonlinear Model Predictive Control (NMPC) based real-time implementable planning technique that jointly plans feasible motion for the mobile base and the manipulator's arm and generates a kinodynamic feasible, collision-free trajectory for cooperative object transportation. Simulation and hardware experiments validate the efficiency of our proposed planning framework.
arXiv:2605.18332v1 Announce Type: new Abstract: Behavioral studies of LLM-based software engineering agents extract operational rules about which trajectory shapes correlate with higher resolution rates: that a test step follows a code modification, that error cascades are short, or that trajectories are compact. Each rule is typically derived from a single framework, and whether it transfers, in sign as well as magnitude, to structurally different agent designs has not been directly tested. We address this at ecosystem scale: 64,380 SWE-bench runs from 126 agent configurations spanning 43 frameworks, where each configuration pairs an LLM with a framework (e.g., SWE-Agent, OpenHands) that supplies its tools and workflow. We separate framework effects from LLM effects by holding each layer fixed in turn, then measure one behavior-outcome effect per configuration and examine how those effects agree or disagree. Swapping the framework while the LLM is held fixed produces large behavioral differences in every action feature. On most signals, configurations disagree not merely in magnitude but in direction. Error rate is the cleanest case: 47 configurations resolve more issues when their error rate is lower, while 48 resolve more when it is higher. Five other continuous features and three of seven binary patterns from prior SE literature show similar directional disagreement. Framework identity accounts for more of this variation than LLM family: for mean turns, framework explains 64% of the between-configuration variance against the LLM's 10%. The implication is that the same observable behavioral signal can carry opposite meaning for different agent configurations. Behavioral findings from any single framework therefore warrant cross-configuration validation before being claimed as general.
arXiv:2605.16888v1 Announce Type: cross Abstract: The finite-temperature spin response of the uniform electron gas (UEG) is a fundamental reference for spin-polarized and magnetized electron liquids, including warm dense matter (WDM), yet it remains far less constrained than charge response. Using variational diagrammatic Monte Carlo, we compute the static spin exchange--correlation (XC) kernel $K_{xc}(q;T)$ of the unpolarized UEG at metallic densities across the quantum-degenerate, warm-dense, and classical regimes. The kernel connects smoothly to zero-temperature spin-response parametrizations at low temperature, while heating suppresses the Fermi-surface-scale spin-correlation structure and weakens the XC-driven Stoner enhancement. Its long-wavelength limit provides a direct response test of the spin stiffness implied by thermal local-spin-density-approximation (LSDA) parametrizations, showing low-temperature consistency while exposing a resolved warm-dense residual in current LSDA parametrizations. In the classical regime, the spin XC kernel becomes nearly local on the Fermi-momentum scale, in sharp contrast to the corresponding charge XC kernel. These results provide a first-principles basis for finite-temperature spin-response theory and magnetized WDM modeling.
arXiv:2605.18378v1 Announce Type: cross Abstract: While video compression algorithms effectively reduce bitrate, aggressive quantization often compromises temporal coherence, introducing artifacts such as flicker, motion inconsistency, and unstable textures. Although spatial quality degradation is well-documented, the relationship between compression intensity and temporal stability remains insufficiently characterized. This paper systematically examines the progression of frame-to-frame coherence errors across different bitrate regimes, utilizing multiple codecs (AV1, HEVC, VP9, H.264) and content types. Our findings reveal that temporal consistency degrades non-linearly with increasing compression. Most critically, we identify a "Predictability anomaly" where sequences with unpredictable or irregular dynamics experience disproportionately higher instability than sequences with higher, but more predictable, motion magnitude. This challenges the conventional assumption that motion volume alone dictates encoding difficulty and highlights the necessity of temporal-aware metrics in compression pipelines.
arXiv:2605.18656v1 Announce Type: cross Abstract: Federated Learning is a leading framework for training ML and AI models collaboratively across numerous user devices or databases. We study the trade-offs among estimation accuracy, privacy constraints, and communication cost for differentially private (DP) federated M estimation. The two standard methods in the literature are FedAvg, which may suffer from high federation bias, and FedSGD, which can incur high communication cost. Aimed at improving accuracy at a reduced communication cost, we propose FedHybrid, which uses FedSGD starting with an improved initialization by the FedAvg estimator. We propose FedNewton, which averages local Newton iterations to reduce bias in FedAvg, achieving an estimation accuracy comparable to FedSGD with much fewer communication rounds when the number of clients grows sufficiently slowly. We establish finite sample upper bounds on the mean-squared error rates of the DP versions of these estimators as functions of the number of clients, local sample sizes, privacy budget, and number of iterations. We further derive a minimax lower bound on the MSE of any iterative private federated procedure that provides a benchmark to assess the optimality gap of these methods. We numerically evaluate our methods for training a logistic regression and a neural network on the computer vision datasets MNIST and CIFAR-10.
arXiv:2309.16682v2 Announce Type: replace Abstract: Many simulation applications require the generation of long sequences of pseudo-random numbers. Linear recurrences modulo 2 are commonly used as the fundamental building block for constructing pseudo-random number generators with extended periods and excellent statistical properties. These generators consist of a lengthy binary state vector that evolves iteratively through linear transformations. One widely accepted pseudo-random generator in this category is the Mersenne twister 19937 (MT19937), proposed by Matsumoto and Nishimura, which has been implemented in numerous software libraries and numerical packages. The MT19937's popularity stems from its favorable distribution properties and the simplicity and speed of its algorithm. The linear transformation responsible for evolving the binary state vector can be expressed as a concise set of elementary bit manipulations. However, this transformation does not fully utilize the potential for parallelization through SIMD instructions available on modern hardware, limiting further speed enhancements. This paper introduces a new SIMD-friendly random number generator, which maintains the same statistical properties and period as the MT19937. It combines the random streams of multiple MT19937 instances with state vectors de-phased via jump-ahead transformations, then polls each instance in a round-robin fashion. By evolving their vector states simultaneously, the new generator achieves perfect vectorization, fully leveraging on SIMD hardware capabilities. Comprehensive test results demonstrate that the throughput of the new generator scales approximately linearly with the width of the SIMD registers used. This provides significant speed improvements, especially on modern CPUs equipped with larger SIMD registers, and allows for efficient generation of random numbers for various simulation applications.
arXiv:2410.13846v3 Announce Type: replace Abstract: Scaling language models to handle longer contexts introduces substantial memory challenges due to the growing cost of key-value (KV) caches. Motivated by the efficiency gains of hybrid models and the broad availability of pretrained large transformer backbones, we explore transitioning transformer models into hybrid architectures for a more efficient generation. In this work, we propose LightTransfer, a lightweight method that transforms models such as LLaMA into hybrid variants. Our approach identifies lazy layers -- those focusing on recent or initial tokens -- and replaces their full attention with streaming attention. This transformation can be performed without any training for long-context understanding tasks or with minimal fine-tuning for o1-like long reasoning generation tasks that require stronger reasoning capabilities. Experiments across diverse benchmarks and models (e.g., LLaMA, Mistral, QwQ-STILL) demonstrate that, even when half of the layers are identified as lazy, LightTransfer achieves up to 2.17$\times$ throughput improvement with minimal performance loss ($<1.5\%$ on LongBench) and achieves 53.3\% on math benchmark AIME24 of advanced o1-like long reasoning model QwQ-STILL.
arXiv:2602.12871v2 Announce Type: replace Abstract: Large language models (LLMs) have attracted growing interest as supportive tools for psychiatric assessment and clinical decision support. However, existing mental health benchmarks largely rely on social media data or supportive dialogue settings, limiting their ability to assess whether models can apply formal diagnostic criteria and differential diagnostic rules. In this paper, we introduce MentalBench, a benchmark for evaluating whether LLMs can make DSM-grounded psychiatric diagnostic decisions under varying levels of clinical ambiguity. At the core of MentalBench is MentalKG, a psychiatrist-built and validated knowledge graph encoding DSM-5 diagnostic criteria and differential diagnostic rules for 23 psychiatric disorders. Using MentalKG as an expert-curated logical backbone, we generate 24,750 synthetic clinical cases that systematically vary in information completeness and diagnostic complexity, enabling DSM-grounded evaluation. Our experiments show that although state-of-the-art LLMs perform well on noise-free queries that probe DSM-5 knowledge, they struggle to calibrate their confidence when distinguishing between disorders with overlapping symptoms. These findings raise concerns about the reliability of LLMs as psychiatric decision-support tools and highlight the need for more evaluation that reflects the diverse challenges in real-world psychiatric diagnosis.
arXiv:2502.10665v2 Announce Type: replace Abstract: In this paper, we propose a novel dual-based Lawson's method, termed {b-d-Lawson}, designed for addressing the rational minimax approximation under specific interpolation conditions. The {b-d-Lawson} approach incorporates two pivotal components that have been recently gained prominence in the realm of the rational approximations: the barycentric representation of the rational function and the dual framework for tackling minimax approximation challenges. The employment of barycentric formulae enables a streamlined parameterization of the rational function, ensuring natural satisfaction of interpolation conditions while mitigating numerical instability typically associated with Vandermonde basis matrices when monomial bases are utilized. This enhances both the accuracy and computational stability of the method. To address the bi-level min-max structure, the dual framework effectively transforms the challenge into a max-min dual problem, thereby facilitating the efficient application of Lawson's iteration. The integration of this dual perspective is crucial for optimizing the approximation process. We will discuss several applications of interpolation-constrained rational minimax approximation and illustrate numerical results to evaluate the performance of the {b-d-Lawson} method.