arXiv preprints from January 1, 2026 through September 10, 2026 — 19:02:54 EST
The macroscopic ideal-MHD stability of an axisymmetric mirror device with sonic levels of plasma rotation is analyzed by approximating the plasma equilibrium as a rotating theta pinch possessing an artificial gravity. An eigenmode equation is derived that governs the stability of the equilibrium to small perturbations in the case of...
When asked about entities outside their knowledge boundary, LLMs routinely fabricate plausible-sounding details rather than backing off to safer, more general claims. We frame this failure through a Gricean lens: a cooperative speaker who is uncertain about a referent retreats up the specificity hierarchy, trading informativeness for...
Thin-film Mn$_5$Si$_3$ is one of the most studied altermagnetic candidates thanks to its metallicity, demonstrated anomalous transport properties, and assumed $d$-wave exchange splitting pattern enabling spin-polarized transport and various spintronic applications. Its postulated altermagnetic structure has zero propagation vector, in...
As language-model-based AI is increasingly deployed in autonomous settings, aligning its goals and values with those of humans becomes critical. Today, alignment, and the assistant identity itself, are typically introduced only after pretraining, once behavioral priors are already established. This can make values a thin overlay,...
We construct a smooth canonically polarized threefold, not biholomorphic to a product of positive-dimensional varieties, whose normalized Kähler-Einstein metric admits a non-integrable infinitesimal Einstein deformation. The same tangent direction is non-integrable as an infinitesimal complex deformation. In fact, the space of...
Compactly representing a variation graph is a core problem in computational pangenomics that is usually attacked with techniques that have been originated on texts and adapted to graphs. In this paper we propose a new framework that takes a topology-centric perspective instead. A variation graph is modeled as a directed acyclic graph...
We report the formation of an original dome-shaped air-water cavity at the impact site of closely-packed plunging water jets. The injector assembly consists of concentric rings of circular jets, similar to a shower-head configuration. We infer from high-speed imaging, LASER-induced fluorescence, and optical phase detection probes that...
We present our submission to Task 3 of the Gen$μ$ 2.0 Challenge on visual concept unlearning. Building on MapRoute, we introduce task-specific training objectives, richer concept representations, and semantic routing for concept-specific mapper selection. Our approach improves robust concept removal while preserving unrelated and...
Combined with numerical analytic continuation techniques, quantum Monte Carlo (QMC) methods enable the extraction of real-frequency dynamical properties from imaginary-time correlation functions. However, the efficient computation of imaginary-time correlation functions by QMC simulations can (depending on the particular model used)...
We present Multi-Agent Reasoning and Coordination (MARC), an open-source framework that replaces monolithic LLM prompting with deterministic multi-agent orchestration for clinical reasoning. MARC coordinates role-specialized agents for extraction, reasoning, answer generation, and evaluation, with explicit context passing and...
Dyson universality typically manifests itself in local eigenvalue statistics. Here we show that its signature survives a nonlinear inverse-spectral reconstruction and reappears in the matrix geometry of the reconstructed operator. Using a dressing transformation, we map each unfolded spectrum to a deformation $f(x)$ of a fixed...
Vision-language-action models (VLAs) are moving rapidly towards deployment as general-purpose manipulation policies, but we currently lack basic tools for understanding what these models represent internally or for monitoring them at runtime. Leveraging ideas from mechanistic interpretability, we probe the residual stream of $π_{0.5}$...
In the effort to determine which low-mass exoplanets have atmospheres, LHS 1140 b remains one of the most favorable targets. Its large size (5.6 $\rm M_{\oplus}$ and 1.7 $\rm R_{\oplus}$) and relatively long orbital period (24.7 days) imply an atmosphere may be likely, and notably, recent interior models favor either a...
Analog circuit design is a time-consuming, iterative process in a nonlinear and high-dimensional design space that relies heavily on expert intuition. Among recent developments, LLMs have introduced a promising approach by bringing natural language reasoning to circuit design tasks. The majority of conventional LLM-based approaches...
This letter proposes the integration of on-off digital noise (OODN) modulation with fluid antenna systems (FASs). A unified analytical framework is developed to evaluate the performance of FAS- assisted OODN receivers over additive white Gaussian noise and generalized \k{appa}-μ fading channels. The analysis incorporates fluid antenna...
Orbiting in the habitable zone of its host star, the 1.7 $R_\oplus$, 5.6 $M_\oplus$ planet LHS 1140 b is a target of great interest. Recently Cherubim et al. (2026) published a detection of metastable He escaping from the atmosphere of LHS 1140 b, simultaneously providing the first concrete inference of an atmosphere on this planet...
We investigate the thermodynamic properties and critical behavior of rotating strongly interacting matter within the two-flavor Nambu--Jona-Lasinio (NJL) model in the mean-field approximation. The phase structure and the critical endpoint (CEP) are determined in the temperature--angular velocity \((T,ω)\) plane. By analyzing the...
For every $0\leqβ\leq1/2$, we construct a nonzero real-valued continuous function $f_β$ in $L^1(\mathbb R)\cap L^2(\mathbb R)$ such that $\widehat {f}_β=f_β$ and $f_β(\sqrt{n}/[\log(e+n)]^β)=0$ for all $n\geq 0$. The case $β=0$ settles in the negative a question raised by Radchenko and Viazovska regarding their Fourier interpolation...
We study the Moreau--Yosida unadjusted Langevin algorithm (MYULA) for the nonsmooth composite target \[ π(dx)\propto \exp\{-f(x)-g(x)\}\,dx, \qquad x\in\mathbb R^d, \] where \(f\) is \(m\)-strongly convex with \(L_f\)-Lipschitz gradient and \(g\) is convex and \(G\)-Lipschitz. Let \(g_λ\) be the Moreau envelope of \(g\), \(π_λ\) the...
When comparison units may also respond to treatment, panel comparisons reflect both the treatment effect and spillovers. If the interference pattern is unknown, observed outcomes alone do not separate the two. I characterize what can nevertheless be learned from panel outcomes under general restrictions, without requiring an exposure...
Concept drift refers to changes over time in the statistical properties of data, as compared to the data that was used to train a learning model. Machine learning models for malware detection or classification are particularly susceptible to performance degradation caused by concept drift, as attackers constantly modify existing...
The JWST Mid-Infrared Instrument Medium Resolution Spectrometer provides the capability to obtain spectra from 5-28 micron. The JWST data reduction pipeline removes the majority but not all of the instrument artifacts and the signal-to-noise (S/N) of the resulting spectra are limited by fixed pattern noise. Building on previous work,...
Modern image classification models excel when trained on single task-specific datasets but often struggle to generalize across domains and difficulty levels. We propose ARMDIL, an Adaptive Router for Multi-Domain Image classification with LLMs. ARMDIL is an ensemble that uses a multimodal large language model (MLLM) agent to...
We identify a general mechanism for the finite-time thermalization of macroscopic observables, such as coarse-grained charge densities, in terms of elementary forms of the quantum dynamics of initial states: (1) aperiodicity, which provides a computable measure of (2) a dynamical partially ergodic exploration of the Hilbert space....
Geometry-aware video object scaling aims to anisotropically resize the object along object-centric axes while preserving geometric plausibility, temporal coherence, and background consistency. Existing text-guided methods mainly operate in the 2D image plane, while depth-guided approaches provide coarse control and mesh-based methods...