arXiv preprints from January 1, 2026 through July 21, 2026 — 15:27:05 EST
Large Language Model (LLM)-based agents are widely used in real-world applications such as customer service, web navigation, and software engineering. As these systems become more autonomous and are deployed at scale, understanding why an agent takes a particular action becomes increasingly important for accountability and governance....
Differential fuzzers detect bugs by executing identical inputs across distinct implementations of the same specification, such as JavaScript interpreters. Validating the outputs requires an oracle and for differential testing of JavaScript, these are constructed manually, making them expensive, time-consuming, and prone to false...
Proton minibeam (pMB) radiotherapy, delivers highly heterogeneous dose distributions alternating high-dose peaks and low-dose valleys. This aims to widen the therapeutic window by improving normal tissue sparing while maintaining the same or even better tumour control. The performance of pMB strongly depends on the collimator design...
Machine-learned interatomic potentials enable large systems to be simulated for long time scales at near ab-initio accuracy. This accuracy is achieved by fitting extremely flexible model architectures to high quality reference data. In practice, this flexibility can cause unwanted behavior such as jagged predicted potential energy...
Decoding visual experiences from human brain activity remains a central challenge at the intersection of neuroscience, neuroimaging, and artificial intelligence. A critical obstacle is the inherent variability of cortical responses: neural activity elicited by the same visual stimulus differs across individuals and trials due to...
The performance of Schwarz Waveform Relaxation is critically dependent on the choice of transmission conditions. While classical absorbing conditions work well for wave propagation, they prove insufficient for damped wave equations, particularly in viscoelastic damping regimes where convergence becomes prohibitively slow. This paper...
With increasing levels of robot autonomy, robots are increasingly being supervised by users with varying levels of robotics expertise. As the diversity of the user population increases, it is important to understand how users with different expertise levels approach the supervision task and how this impacts performance of the...
The economic warehouse lot scheduling problem is a foundational inventory-theory model, capturing computational challenges in dynamically coordinating replenishment decisions for multiple commodities subject to a shared capacity constraint. Even though this model has generated a vast body of literature over the last six decades, our...
In this paper, we propose a novel cross-domain channel estimation (CDCE) algorithm for orthogonal frequency division multiplexing (OFDM) systems, leveraging the unique characteristics of the delay-Doppler (DD) domain channel. Specifically, the proposed algorithm transforms the time-frequency (TF) domain pilot sequence of OFDM into the...
We investigate a dark sector coupled to the Standard Model (SM) through a kinetically mixed dark photon $U$ associated with a new $U(1)'$ gauge symmetry. Kinetic mixing $\varepsilon$ induces an effective coupling to the electromagnetic current, while $U$ interacts with stable dark matter (DM) $χ$ via a dark gauge coupling $g_χ$. Our...
CLIP-based foreground-background (FG-BG) decomposition methods have demonstrated remarkable effectiveness in improving few-shot out-of-distribution (OOD) detection performance. However, existing approaches still suffer from several limitations. For background regions obtained from decomposition, existing methods adopt a uniform...
AI has revolutionised decision-making across various fields. Yet human judgement remains paramount for high-stakes decision-making. This has fueled explorations of collaborative decision-making between humans and AI systems, aiming to leverage the strengths of both. To explore this dynamic, researchers conduct empirical studies,...
A central open question in chiral-induced spin selectivity (CISS) is whether weak electron-phonon coupling in a helical molecular junction can generate a sizable spin polarization in two-terminal transport without invoking additional strong symmetry-breaking ingredients. We address this question by implementing a self-consistent...
Scientific fields are often mapped using citations and metadata, despite knowledge being transmitted primarily through content. We introduce an 'inside-out' approach that reconstructs field structure directly from text by representing each paper as a small set of interpretable knowledge components. Using a large language model to...
Traditional data masking techniques such as anonymization cannot achieve the expected privacy protection while ensuring data utility for privacy-preserving machine learning. Synthetic data plays an increasingly important role as it generates a large number of training samples and prevents information leakage in real data. The existing...
We propose a new formulation of the Korteweg-de Vries equation (KdV) on the real line, via a gauge transform. While KdV and the gauged equation are equivalent for smooth solutions, the latter is better behaved at low regularity in Fourier-Lebesgue spaces. In particular, the admissible regularities go beyond the $H^{-1}$-scale, which...
End-to-end automatic speech recognition (ASR) systems frequently misrecognize domain-specific phrases like named entities, which can cause catastrophic failures in downstream tasks. A new family of named entity correction methods based on large language models (LLMs) has recently emerged. However, these approaches have yet to fully...
Modern CI/CD pipelines integrating agent-generated code exhibit a structural failure in responsibility attribution. Decisions are executed through formally correct approval processes, yet no entity possesses both the authority to approve those decisions and the epistemic capacity to meaningfully understand their basis. We define...
The goal of this paper is to establish a local rigidity result for the integrability of standard-like maps. The main focus of the paper is the remarkable integrable potential discovered by Suris in the 80's. We show that locally, the integrability of this potential is rigid. The proof relies on a similar strategy that was used for...
Surface chemistry on interstellar dust grains is recognized as a central component in astrochemical models, representing a plausible formation route for many of the observed complex molecular species. However, key parameters governing interstellar surface chemistry, such as diffusion energy barriers, remain poorly constrained. In...
Falls are the leading cause of injury related hospitalization and mortality among older adults. Consequently, mitigating age-related declines in gait stability and reducing fall risk during walking is a critical goal for assistive devices. Lower-limb exoskeletons have the potential to support users in maintaining stability during...
Traditional defenses against Deep Leakage (DL) attacks in Federated Learning (FL) primarily focus on obfuscation, introducing noise, transformations or encryption to degrade an attacker's ability to reconstruct private data. While effective to some extent, these methods often still leak high-level information such as class...
The identical charged-kaon correlations induced by quantum-statistics effects and final-state interactions are measured in Pb$-$Pb collisions at $\sqrt{s_{\rm NN}} = 5.02$ TeV. The results of one- (1D) and three-dimensional (3D) analyses show that the obtained system-size parameters (radii) are smaller for more peripheral collisions...
In this work, three InAs/InAs$_{0.65}$Sb$_{0.35}$ superlattices with different periods were investigated using photoluminescence and photoreflectance measurements and their band structure was simulated using a 14 bulk-band kp model. The structures were studied by analyzing the evolution of the spectral features in temperature and...
It is well-known that Leonard pairs have a close connection with bispectral orthogonal polynomials of the Askey scheme. In this paper, we introduce the notion of a Leonard trio $(V,\oV,Z)$, an algebraic structure extending Leonard pairs, for which the overlap coefficients of eigenfunctions of $V$ and $\oV$ are biorthogonal rational...