arXiv:2605.26419v1 Announce Type: new
Abstract: Amortized inference promises fast test-time Bayesian inference, but existing methods are inherently tied to fixed models. Extending amortization to unseen models typically requires retraining or costly test-time finetuning. In this paper, we ask: is it possible to build a single inference network capable of generalizing across varying priors, likelihoods, and dimensionality? We introduce Amortized Factor Inference Networks (AFINs), a family of encode-merge-decode inference networks built on dimension-independent modules that map a model specification and its observations to the parameters of a variational posterior. Experimentally, a single trained AFIN achieves posterior accuracy comparable to NUTS and several variational inference methods, while requiring 2 to 4 orders of magnitude less test-time compute. Code is available at https://github.com/joohwanko/AFINs.
Science Journals
arXiv:2605.26781v1 Announce Type: new
Abstract: Advanced Large Multimodal Models (LMMs) have demonstrated impressive performance in K-12 reasoning tasks, exhibiting great promise as intelligent tutors. Realizing this potential requires models to navigate real-world examinations effectively, yet most existing benchmarks fail to capture the complexity of authentic testing environments. Specifically, most datasets are static, prone to data contamination, and are often confined to restricted modalities, disciplines, and evaluation criteria. To address these issues, we introduce LiveK12Bench, a dynamic, holistic, multi-disciplinary benchmark designed to evaluate the reasoning abilities of LMMs in realistic examination scenarios. LiveK12Bench comprises 2K+ verified questions spanning Mathematics, Physics, Chemistry, and Biology, sourced from the latest real-world exam papers and designed to grow over time. Our framework features several core innovations: 1) featuring an automated pipeline that continuously ingests and parses the latest examination papers to mitigate data leakage; and 2) proposing a novel `Mock Exam' evaluation scheme, which assesses the ability to complete end-to-end exams autonomously with accurate and efficient reasoning paths. Extensive experiments on 12 LMMs reveal that advanced models suffer substantial performance degradation under exam-realistic constraints: GPT-5's score drops from 79 to 53 (out of 100) when process rigor and efficiency are jointly evaluated. Our findings expose critical vulnerabilities, such as sensitivity to complex visual layouts, highlighting the gap between idealized reasoning capabilities and true educational readiness. Both code and dataset are publicly available.
arXiv:2508.16801v2 Announce Type: replace-cross
Abstract: We address the stabilization of linear, time-varying parabolic PDEs using finite-dimensional receding horizon controls (RHCs) derived from reduced-order models (ROMs). We first prove exponential stability and suboptimality of the continuous-time full-order model (FOM) RHC scheme in Hilbert spaces. A Galerkin model reduction is then introduced, along with a rigorous a posteriori error analysis for the associated finite-horizon optimal control problems. This results in a ROM-based RHC algorithm that adaptively constructs reduced-order controls, ensuring exponential stability of the FOM closed-loop state and providing computable performance bounds with respect to the infinite-horizon FOM control problem. Numerical experiments with a non-smooth cost functional involving the squared l1-norm confirm the methods effectiveness, even for exponentially unstable systems.
arXiv:2510.16204v2 Announce Type: replace-cross
Abstract: We study the noisy dynamics of periodically driven, discrete-step quantum walks in a one-dimensional photonic lattice. We find that in the bulk, temporal noise that is constant within a Floquet period leads to decoherence-free momentum subspaces, whereas fully random noise destroys coherence in a few time steps. When considering topological edge states, we observe decoherence no matter the type of temporal noise. To explain these results, we derive a non-perturbative master equation to describe the system's dynamics. We experimentally confirm our findings in a time multiplexed photonic lattice implemented in a double-fiber ring setup subject to laser pulse input states, in which we engineer different types of temporal noise.
arXiv:2602.20475v2 Announce Type: replace-cross
Abstract: The High-Luminosity Large Hadron Collider (HL-LHC) at CERN will produce unprecedented datasets capable of revealing fundamental properties of the universe. However, realizing its discovery potential faces a significant challenge: extracting small signal fractions from overwhelming backgrounds dominated by approximately 200 simultaneous pileup collisions. This extreme noise severely distorts the physical observables required for accurate reconstruction. To address this, we introduce the Physics-Guided Hypergraph Transformer (PhyGHT), a hybrid architecture that combines distance-aware local graph attention with global self-attention to mirror the physical topology of particle showers formed in proton-proton collisions. Crucially, we integrate a Pileup Suppression Gate (PSG), an interpretable, physics-constrained mechanism that explicitly learns to filter soft noise prior to hypergraph aggregation. To validate our approach, we release a novel simulated dataset of top-quark pair production to model extreme pileup conditions. PhyGHT outperforms state-of-the-art baselines from the ATLAS and CMS experiments in predicting the signal's energy and mass correction factors. By accurately reconstructing the top quark's invariant mass, we demonstrate how machine learning innovation and interdisciplinary collaboration can directly advance scientific discovery at the frontiers of experimental physics and enhance the HL-LHC's discovery potential. The dataset and code are available at https://github.com/rAIson-Lab/PhyGHT
arXiv:2605.26341v1 Announce Type: new
Abstract: Physics-informed machine learning (PIML) integrates mechanistic knowledge, typically in the form of partial differential equations (PDE), into data-driven models. Despite strong empirical performance, its statistical generalisation properties remain poorly understood, particularly in the regression setting with unbounded losses. Existing analyses rely on approximation or stability arguments and do not fully capture how physical structure influences generalisation from finite data. In this work, we develop a PAC-Bayesian framework for PIML that provides high-probability generalisation guarantees in the presence of unbounded losses. We adopt a multi-task perspective that jointly treats data fidelity, PDE residuals, initial and boundary conditions, avoiding the looseness induced by standard union-bound approaches. Our analysis leverages the structure of physics-informed objectives to derive novel bounds where the complexity scales with input-gradient norms of the losses, revealing a direct link between physical regularity and generalisation. We instantiate this framework under Sobolev and Poincar\'e-type assumptions, yielding two classes of bounds that trade off statistical complexity and smoothness in different regimes. Building on these results, we propose a self-bounding-aware learning algorithm that directly optimises tractable surrogates of the derived bounds, along with a practical procedure to estimate the associated constants in realistic settings. Empirical evaluations on standard PDE benchmarks demonstrate that our bounds are non-vacuous, significantly tighter than union-bound baselines, and can be effectively minimised during training. Overall, our results provide a principled statistical foundation for the generalisation of physics-informed models.
arXiv:2605.26190v1 Announce Type: new
Abstract: This paper presents the HRVConformer, a novel deep learning architecture for the classification of hypoxic-ischemic encephalopathy (HIE) using the instantaneous heart rate (HR) signal. Unlike conventional approaches that rely on handcrafted features, HRVConformer directly processes raw HR signals in an end-to-end manner, capturing both local and long-range dependencies through a hybrid Convolution-Transformer framework. By integrating convolutional layers for local feature extraction and Transformer-based attention mechanisms for global context modelling, the architecture effectively enhances signal representation and classification performance. The model was trained using supervised learning on a large HR dataset consisting of 1,573 one-hour epochs, including 259 one-hour expert-annotated epochs and a substantial set of weakly labelled data. A 314-hour validation set provided a robust performance estimation, while an independent 215-hour dataset with expert annotations was reserved for final testing. HR signals were extracted from electrocardiogram (ECG) recordings using an improved Pan-Tompkins algorithm, which significantly enhanced both signal quality and data availability. Experimental results demonstrate that the HRVConformer achieves an AUC of 83.23\% and accuracy of 74.56\% on the test set. These results surpass the performance of the Transformer, ResNet50 and fully convolutional networks baselines, highlighting the advantages of integrating convolutional and Transformer-based components for HR-based HIE classification. The proposed method provides a promising step toward a more accurate and automated assessment of HIE using HR signals. The code is available at: https://github.com/syu-kylin/HRVConformer.
arXiv:2605.26446v1 Announce Type: new
Abstract: Graph anomaly detection (GAD) aims to identify nodes or substructures whose behavior or attributes deviate significantly from the overall pattern in graph-structured data, with critical applications in financial risk control, social network analysis, and cybersecurity. However, existing GCN-based methods suffer from the fundamental problem of contamination propagation, where anomalous nodes pollute the representations of their neighbors through message passing, leading to degraded detection performance. In this paper, we propose DDGAD, a novel diffusion-based graph anomaly detection framework that leverages trajectory dynamics to distinguish normal and anomalous nodes. Our key insight is that normal nodes exhibit consistent and stable representation trajectories under the coupled effects of diffusion regularization and reliability-aware neighborhood consensus, while anomalous nodes exhibit unstable and conflicting dynamics due to the directional disagreement between the global manifold prior and locally contaminated message passing. To mitigate contamination propagation, we introduce a distributed reliability-aware consensus refinement mechanism and define three complementary anomaly signals: neighbor inconsistency, reliability weight, and dynamical conflict energy. We further provide a preliminary theoretical analysis on normal node stability under the coupled dynamics. These signals collectively characterize anomalous behaviors from the perspectives of local inconsistency, consensus reliability, and dynamical instability. Extensive experiments on five real-world datasets demonstrate the effectiveness of the proposed framework.
arXiv:2605.26273v1 Announce Type: new
Abstract: Semantic segmentation in complex environments such as urban driving scenes remains challenging under adverse lighting conditions, where RGB images alone provide insufficient information. RGB-Thermal fusion leverages the complementary strengths of visible and infrared imagery to improve scene understanding; however, effectively integrating these heterogeneous modalities at varying levels of feature abstraction remains an open problem. In this paper, we propose a multi-modal fusion architecture built upon dual ConvNeXt V2 backbones that employs stage-wise, modality-adaptive fusion strategies. For early-stage features, we introduce a Frequency-Based Fusion Module that decomposes infrared features into low- and high-frequency components via Gaussian filtering, applies dual-branch spatial attention to selectively emphasize thermal patterns and fine-grained boundaries, and integrates them with RGB features through a confidence-gated residual mechanism. For late-stage features, we design a semantic fusion module with cross-modal attention and multi-scale depthwise convolutions to capture semantic correspondences across modalities. The fused features are decoded via a PANet-style bidirectional decoder with deep supervision. Experiments on MFNet and PST900 demonstrate that our lightest variant achieves 61.73\% and 86.24\% mIoU, respectively, with only 35.43M parameters, outperforming recent methods while using substantially fewer parameters and lower computational cost. Code is available at https://github.com/ismailemrecntz/VISIBLE-INFRARED-SENSOR-FUSION
arXiv:2601.21789v2 Announce Type: replace
Abstract: We introduce ECSEL, an explainable classification method that learns formal expressions in the form of signomial equations, motivated by the observation that many symbolic regression benchmarks admit compact signomial structure. ECSEL directly constructs a structural, closed-form expression that serves as both a classifier and an explanation. On standard symbolic regression benchmarks, our method recovers a larger fraction of target equations than competing state-of-the-art approaches while requiring substantially less computation. Leveraging this efficiency, ECSEL achieves classification accuracy competitive with established machine learning models without sacrificing interpretability. Further, we show that ECSEL satisfies some desirable properties regarding global feature behavior, decision-boundary analysis, and local feature attributions. Experiments on benchmark datasets and two real-world case studies i.e., e-commerce and fraud detection, demonstrate that the learned equations expose dataset biases, support counterfactual reasoning, and yield actionable insights.
APEX-Searcher: Refining Credit Assignment with Subgoaling for Agentic Retrieval-Augmented Generation
arXiv:2603.13853v3 Announce Type: replace
Abstract: Retrieval-augmented generation (RAG) connects large language models (LLMs) to external knowledge, but single-round retrieval is often insufficient for complex multi-hop questions. To enhance search capabilities for complex tasks, most existing works integrate multi-round iterative retrieval with reasoning processes via end-to-end training. While these approaches improve problem-solving performance, they still face challenges in task reasoning and model training, especially ambiguous retrieval execution paths and sparse rewards in end-to-end reinforcement learning (RL), which can lead to inaccurate retrieval results and lower performance. We attribute these failures to hierarchical credit entanglement: a single final reward updates planning and execution together, so the model cannot clearly separate plan errors from retrieval errors. We propose APEX-Searcher, which uses a Refining Credit Assignment paradigm: planning is optimized by RL with a plan-level reward, while execution is learned by SFT. Extensive experiments show consistent gains in both multi-hop RAG and task planning across benchmarks.
arXiv:2605.26618v1 Announce Type: new
Abstract: The Abbe diffraction limit, tied to the fundamental spatial bandwidth constraint imposed by any physical aperture, remains the primary barrier to achieving ultimate far-field optical resolution and precise light-matter interactions. However, current efforts to engineer structured light fields beyond this limit often come at the cost of massive sacrifices in energy efficiency. In this work, we mathematically complete the family of non-zero azimuthal-order Circular Prolate Spheroidal Wave Functions (CPSWFs), introducing them as a complete class of band-limited superoscillatory optical vortices carrying helical phase. Compared with classical Laguerre-Gaussian (LG) beams, we rigorously prove that these eigenmodes achieve the theoretical upper bound for extreme energy concentration under strict band-limited constraints. At the scale of light-matter interactions, this optimal concentration directly amplifies the intensity gradients and angular momentum densities that govern optical forces. This advantage translates directly into a 29.9% reduction in the trapping power threshold and a 2.3-fold increase in the subdiffraction orbital rotation speed of nanoparticles. Looking forward, this fundamental physical framework not only establishes strict mathematical boundaries for structured light fields but also serves as an absolute theoretical benchmark for deep-learning inverse design, and next-generation extreme optical micro-manipulation systems.
arXiv:2605.26433v1 Announce Type: new
Abstract: Large language model (LLM) summarization systems may pass compact vector representations of private inputs to downstream retrieval, monitoring, audit, or analytic workflows. Even when source documents remain access-restricted, derived vectors may be handled under different access controls and still support sensitive-information inference, creating a residual information-disclosure risk. We study this issue in clinical discharge-summary generation as a high-stakes case study, using electronic health record (EHR)-recorded race as a controlled sensitive-label audit. We audit two artifacts that a system might retain or expose to downstream components: the final prompt-token hidden state and the mean-pooled prompt representation. Our results show that reducing recoverability of the case-study sensitive label from one exported artifact does not necessarily reduce recoverability from another. As a mitigation case study, we introduce SurfaceLoRA, an exported-vector-targeted parameter-efficient fine-tuning method that uses a gradient-reversal discriminator attached to a designated exported vector. Under a balanced five-way probing protocol, SurfaceLoRA reduces EHR-recorded race recoverability from the targeted final-token artifact toward chance while preserving summarization utility, yet recoverability remains substantially higher from untargeted pooled artifacts. These findings show that privacy auditing and mitigation should be performed on the exact vector artifact retained or exposed to downstream components.
arXiv:2605.26257v1 Announce Type: cross
Abstract: In recent years, increasing aerospace safety requirements have intensified the demand for reliable structural damage detection. This work presents an Operational Modal Analysis approach for accurate modal parameter estimation, with an application to space structure monitoring. The proposed System Identification (SI) method innovatively combines the Natural Excitation Technique (NExT) with the Fast and Relaxed Vector Fitting (FRVF) algorithm, which uses an iterative least-squares optimisation. A preliminary validation is first carried out on a numerical beam model, comparing results with analytical solutions and the established Natural Excitation Technique with Eigensystem Realisation Algorithm (NExT-ERA) and Stochastic Subspace Identification with Canonical Variate Analysis (SSI) methods. Then, operational validation is performed on real acceleration data from the Space Acceleration Measurement Systems aboard the International Space Station. Identified vibration modes from NExT-FRVF and NExT-ERA show comparable results after signal processing, with mode consistency assessed by repeated occurrence and physical interpretation, while SSI fails to identify most. The output-only algorithm proves to be highly reliable, outperforming benchmark methods under noisy conditions on a numerical system and offering reliable identifications on the experimental data.
arXiv:2605.24071v2 Announce Type: replace
Abstract: Training a reinforcement learning agent on-policy means collecting fresh experience at every update, and that experience comes with a hidden problem. Each state in a rollout is the direct output of the previous one, causally chained together by the agent's own actions. Because of this, consecutive transitions are never truly independent. They carry overlapping information, and the gradient signal the network receives ends up far more repetitive than the batch size suggests. The same directions get reinforced over and over, the value network struggles to keep up as the policy shifts, and training becomes quietly unstable in ways that reward curves alone rarely reveal.
This paper asks whether that redundancy can simply be removed. We show that randomly dropping a fixed fraction of transitions from the rollout, at the right stage so the reward signal stays intact, is enough to break the repetitive gradient structure and stabilize training. The change is minimal: one sampling step, no new components, no modification to the core algorithm, and it works with any PPO implementation. Across five environments of increasing difficulty, CartPole-v1, Acrobot-v1, LunarLander-v2, HalfCheetah-v5, and Hopper-v5, the method matches vanilla PPO on reward while producing more consistent training dynamics across KL divergence, policy entropy, and value estimates. Dropping 25% of transitions turns out to be the sweet spot: enough to disrupt the redundancy, not enough to thin the batch.
arXiv:2605.26619v1 Announce Type: new
Abstract: Reconstructing continuous state trajectories of chaotic dynamical systems from sparse, noisy observations remains a fundamental open problem in nonlinear science. We introduce the Physics-Informed Diffusion Model with Dormand-Prince Integration (PIDM-DP), which embeds a fully differentiable 5th-order Dormand-Prince (DP-RK45) ODE integrator directly into the reverse sampling loop of a Denoising Diffusion Probabilistic Model (DDPM). At each denoising step, physics residuals are back-propagated via automatic differentiation, constraining every generated trajectory to satisfy the system's governing equations to 5th-order accuracy. A linear-scheduled guidance mechanism that ramps the physics weight from zero at high noise levels to its full value near the clean-data limit prevents the gradient explosions that cause naive physics-informed approaches to fail on stiff systems with Jacobian eigenvalues of order $O(10^3)$. Evaluated across five benchmark systems of increasing complexity 3D Lorenz, 3D R\"ossler, 5D Hyperchaotic, 20D Lorenz-96, and the stiff 3D Rabinovich-Fabrikant at 10% observation density with additive Gaussian noise ($\sigma=0.05$), PIDM-DP achieves reconstruction RMSE improvements of up to $15.4\times$ over an unconstrained diffusion baseline and decisively outperforms the Ensemble Kalman Filter on stiff systems where ensemble covariance collapses. On the Rabinovich-Fabrikant out-of-distribution benchmark, PIDM-DP attains RMSE $0.1097 \pm 0.0269$ versus $0.9443 \pm 0.5288$ (unconstrained diffusion, $8.6\times$ worse) and $0.3561 \pm 0.3040$ (EnKF, $3.2\times$ worse), with $p<0.001$ in paired Wilcoxon tests ($N = 30$). Topological validation via the Rosenstein Lyapunov estimator confirms that PIDM-DP preserves the chaotic invariant measure.
arXiv:2603.12592v4 Announce Type: replace
Abstract: Routing algorithms for public transport, particularly the widely used RAPTOR and its variants, often face performance bottlenecks during the transfer relaxation phase, especially on dense transfer graphs, when supporting unlimited transfers. This inefficiency arises from iterating over many potential inter-stop connections (walks, bikes, e-scooters, etc.). To maintain acceptable performance, practitioners often limit transfer distances or exclude certain transfer options, which can reduce path optimality and restrict the multimodal options presented to travellers.
This paper introduces Early Pruning, a low-overhead technique that accelerates routing algorithms without compromising optimality. By pre-sorting transfer connections by duration and applying a pruning rule within the transfer loop, the method discards longer transfers at a stop once they cannot yield an earlier arrival than the current best solution.
Early Pruning can be integrated with minimal changes to existing codebases and requires only a one-time preprocessing step. The technique preserves Pareto-optimality in extended-criteria settings whenever the additional optimization criteria are monotonically non-decreasing in transfer duration. Across multiple state-of-the-art RAPTOR-based solutions, including RAPTOR, ULTRA-RAPTOR, McRAPTOR, BM-RAPTOR, ULTRA-McRAPTOR, and UBM-RAPTOR and tested on the Switzerland and London transit networks, we achieved query time reductions of up to 57\%. This approach provides a generalizable improvement to the efficiency of transit pathfinding algorithms.
arXiv:2605.25002v2 Announce Type: replace
Abstract: Memory-backed agents need provenance that can survive leaked or migrated snapshots, where logs, visible outputs, and trusted metadata may be absent. We propose MemMark, a state-evolution attribution watermark that embeds an owner-controlled signal into latent memory-write decisions. At each internal LLM call, MemMark samples among admissible candidates using keyed, distribution-preserving selection, and records cryptographic commitments with signed session anchors and reveal evidence. This makes attribution depend on reproducible backend behavior rather than mutable provenance fields. Across A-Mem and Graphiti on LoCoMo, with three LLM backbones, MemMark preserves memory utility: Overall F1 retains 99.6% of the unwatermarked baseline, while BLEU-1 changes by +0.2%. It also provides usable carrier capacity, with 1.16, 1.14, and 1.26 bits of mean entropy for update-target, link-target, and semantic-realization decisions. In the snapshot-only R3 setting, MemMark recovers the full 40-bit payload from final snapshots, while wrong-key verification remains near chance. Under nine memory-lifecycle attacks, verification distinguishes tampering, evidence deletion, and partial payload recovery. These results show that robust snapshot-only attribution is feasible for long-term agent memory without surviving traces, trusted metadata, or utility-degrading.
arXiv:2605.26151v1 Announce Type: new
Abstract: Contactless diagnosis of musculoskeletal disorders can potentially improve population health as well as robot behaviours in collaborative settings. However, current diagnosis methods require an in-person physical examination in which a trained physician senses, through contact, the force applied by various muscles. Simulation tools exist, but their use for diagnosis with real data is under-explored. In this paper, we propose an algorithm for identifying which upper-limb muscle group is fatigued. Our algorithm compares the realworld free-space motion of the subject with that of a simulated musculoskeletal model, and is therefore contactless: preventing the need for invasive sensing or in-person assessment. Our algorithm simulates various fatigue conditions using a physics-based musculoskeletal model and extracts diagnostic motion features from both real and simulated data, which are compared for diagnosis. Experimental results on real data demonstrate that the proposed method can reliably distinguish between multiple muscle-groups of fatigue. Additionally, through comprehensive performance comparisons, we show how recent advanced musculoskeletal simulators can be properly configured to address the sim-to-real gap in the context of the fatigue diagnosis task. Our approach can potentially spur further research in remote and automated diagnosis, significantly lowering the barrier to large-scale and early detection.
arXiv:2605.26741v1 Announce Type: cross
Abstract: Inverse design of materials has significantly advanced target-driven formulation optimization, yet existing materials machine learning benchmarks remain limited to forward property prediction, failing to systematically evaluate inverse optimization and generation algorithms, a critical gap that hinders the progress of target-driven materials design. To address this limitation, we propose MatFormBench, a novel benchmarking ecosystem tailored to evaluate and guide generative strategies for target-driven formulation. MatFormBench integrates a physics-driven formulation generation scheme to generate synthetic samples that faithfully emulate realistic materials structure-property response relationships, complemented by five escalating difficulty levels to quantify the complexity of these relationships. To rigorously assess algorithm performance, we further propose MatFormScore, a multi-dimensional metric that comprehensively quantifies performance across five critical axes: target success, search efficiency, exploratory capacity, robustness, and stability. We validate MatFormBench by evaluating 39 diverse inverse design algorithms, covering classical surrogate-assisted black-box search, state-of-the-art deep generative models, and increasingly popular Large Language Model (LLM)-based recommendation strategies. Across 1170 standardized algorithm-task evaluations, diffusion-based models demonstrate the strongest overall performance, while Variational Autoencoder (VAE)-based and Genetic Algorithm (GA)-based methods exhibit distinct advantages in specific scenarios. By establishing a unified evaluation standard for target-driven materials formulation, MatFormBench enables reproducible benchmarking, principled algorithm comparison, and diagnostic analysis of inverse design strategies, providing a foundational tool for advancing materials inverse design.
arXiv:2605.27288v1 Announce Type: new
Abstract: Large language models (LLMs) are known to abandon their initial stance to conform to user pushback. While prior research largely attributes this behavior to sycophancy learned during reinforcement learning from human feedback, we hypothesize that conformity is also driven by a model's epistemic uncertainty at inference time. In this paper, we introduce MUSE, a two-stage evaluation framework to disentangle the mechanisms driving LLM conformity. Specifically, MUSE maps a model's epistemic uncertainty in responding to a query against its likelihood to yield to user pushback in a subsequent turn. We demonstrate that the mechanisms driving conformity extend beyond sycophancy alone. Specifically, we characterize two distinct factors that jointly drive conformity: sycophantic conformity, where a model aligns with user pushback even with absolute certainty in its initial response, and uncertainty-driven conformity, where a model's likelihood for conformity increases alongside its uncertainty. Furthermore, we conduct ablation studies to demonstrate that both sycophantic conformity and uncertainty-driven conformity grow with 1) the LLM's perceived expertise of the user and 2) the plausibility of the user's suggestions. More broadly, MUSE informs more targeted intervention strategies by distinguishing alignment-induced sycophancy and training-corpora-driven uncertainty.
arXiv:2605.19267v2 Announce Type: replace
Abstract: Automated Market Makers face a geometric dilemma: expanding liquidity depth to reduce execution slippage increases Liquidity Providers' exposure to toxic arbitrage, quantified as Loss-Versus-Rebalancing (LVR). We study the Hybrid Liquidity-Collateral Pool (HLCP), a stylized architecture that aims to partially decouple execution quality from active risk exposure through an N-scaled virtual invariant and a collateral buffer. The analysis first characterizes the geometric divergence between execution slippage and marginal-price deviation, then uses this divergence to motivate a trigger-based collateral injection rule. In a stylized duopoly model, under hyper-saturated background liquidity and non-zero volatility or collateral yield, adopting the HLCP is a Nash equilibrium and Pareto-improving relative to a standard AMM benchmark. Empirically, we examine two settings. Under a stochastic-volatility-with-jumps stress scenario, the trigger policy avoids one-shot total buffer depletion under the imposed control law and simulated shock path. Using 2025 Uniswap V2 data with zero collateral yield, the HLCP exhibits lower realized LVR and higher net LP return than the standard CPMM benchmark in the sample considered.
arXiv:2605.26438v1 Announce Type: new
Abstract: Large language models can recognize when they are being evaluated (evaluation awareness) and behave differently because of that, which undermines the validity of safety and alignment benchmarks. We propose LURE (Live-Usage Replay Evaluations), a method for constructing deployment-like evaluations by replaying realistic agentic interaction trajectories and appending evaluation prompt at the end. We also introduce an automated pipeline for measuring evaluation realism, combining detection of verbalized evaluation awareness and judge-model estimates of the probability of logs being an evaluation, and validate it on a large dataset of deployment and evaluation transcripts. We find that LURE-based evaluations are substantially less distinguishable from deployment than widely used benchmarks and synthetic evaluation generators, and can approach the realism of real conversations with users. We instantiate LURE in scheming, AI safety sabotage, and sycophancy settings. Our results suggest that evaluation realism is a crucial property of alignment benchmarks and should be reported alongside benchmark results, especially when such results are used in safety cases.
arXiv:2605.27010v1 Announce Type: new
Abstract: We perform point-particle direct numerical simulations (PP-DNS) of particle-laden flow through a linear compressor cascade subjected to synthetic freestream turbulence. Monodisperse particles are advanced in a one-way coupled Eulerian-Lagrangian framework with drag-only dynamics. We quantify blade-particle collisions and resulting blade erosion based on high-fidelity data, and the erosion hotspots are predicted near the leading edge and over the pressure side. On the pressure side, for intermediate Stokes numbers, the onset of collisions correlates with elevated boundary-layer intermittency associated with bypass transition, whereas for larger particles impacts occur farther upstream with a higher probability of multiple rebounds. On the suction side, sparse collisions appear only for the smallest particles and are phase-modulated by separation-induced vortex shedding. Joint distributions of impact velocity and angle show that leading-edge impacts are faster and span wider angles than pressure-side impacts, explaining their greater erosive severity. The present results highlight the role of unsteady boundary-layer dynamics in affecting erosion in compressor cascades.
arXiv:2605.27011v1 Announce Type: new
Abstract: A key challenge in material theory is the formulation of models that satisfy all common mechanical constitutive conditions while retaining sufficient flexibility. In this context, several important modeling aspects remain unresolved for polyconvex anisotropic hyperelasticity. We address some of these challenges and apply our results for physics-augmented neural network (PANN) constitutive modeling. The main contributions of this paper are as follows: (1) We propose a new polyconvex PANN constitutive model for anisotropic hyperelasticity based on triclinic invariants and group symmetrization. For finite symmetry groups, this model fulfills all common mechanical constitutive conditions a priori. (2) We propose a group symmetrization-based method for the construction of polyconvex invariants for finite symmetry groups. Based on this, we derive a new integrity basis for a tetragonal symmetry group and a new functional basis for a cubic symmetry group. To the best of our knowledge, these are the first polyconvex integrity or functional bases for symmetry groups characterized by structural tensors of order higher than two. (3) We provide an extensive introduction to the construction of polyconvex integrity and functional bases, which form the basis of polyconvex invariant-based constitutive models. We discuss polyconvex bases for triclinic, isotropic, transversely isotropic, monoclinic, rhombic, tetragonal, and cubic symmetry groups. (4) We benchmark the polyconvex PANN constitutive models with highly nonlinear homogenization data of cubic metamaterials.