arXiv preprints from January 1, 2026 through September 23, 2026 — 12:36:35 EST
On-policy distillation (OPD) adapts diffusion models by querying a teacher along trajectories generated by the current student, but how it should behave under classifier-free guidance (CFG), a default component of modern diffusion systems, remains poorly understood. Existing OPD methods naturally extend velocity matching to the...
Computer vision models have become highly effective for medical applications, yet their black-box nature continues to undermine clinician trust. In clinical workflows, chest X-ray classifiers are increasingly paired with Vision-Language Models (VLMs) to generate natural-language explanations. However, these systems add linguistic...
We introduce MicroZoom, a generative framework for gigapixel image synthesis at the microscopic scale. Given a standard photograph and a sparse set of consumer-grade microscope close-ups, MicroZoom synthesizes a seamless, gigapixel-resolution image grounded in the material character of the real references, enabling exploratory...
Passive linear optics is a restricted model of quantum computation, with complexity-theoretic evidence of quantum advantage for sampling tasks and low losses that make it attractive for near-term algorithms. In qubit architectures, a body of work has revealed a close connection between barren plateaus and classical simulability....
Background. Pediatric musculoskeletal trauma represents up to 18% of pediatric ED visits, yet diagnosis still depends on ionizing radiography. Cumulative low-dose radiation in early life raises lifetime leukemia and brain malignancy risk, motivating radiation-free triage alternatives. Objective. To synthesize evidence for a hybrid...
The Deep Galerkin Method (DGM) and Physics Informed Neural Networks (PINNs) have become widely-used methods for solving partial differential equations (PDEs) in the rapidly growing field of scientific machine learning. In these methods, a neural network is trained to approximate the PDE solution by using (stochastic) gradient descent...
The cycle double cover conjecture of Szekeres and Seymour, the proof of which was recently announced by OpenAI, states that every bridgeless graph has a collection of cycles covering every edge exactly twice. We study the counting version of this statement for cubic graphs, where we count circuit double covers --- collections of...
Optimal control problems arise in a wide range of scientific disciplines, but the corresponding optimization algorithms often display a strong dependence on hyperparameters that significantly influence performance and convergence. For the optimal implementation of quantum algorithms, these challenges are further amplified by...
We revisit the derivation of the diffraction integral, which is an approximate evaluation of the Kirchhoff integral, widely used in the literature for phenomenological applications in gravitational lensing. We propose a systematic approach to evaluate the Kirchhoff integral within a beyond-eikonal expansion, carefully tracking all...
Variational quantum algorithms offer a promising route to combinatorial optimization, but their applicability is limited by the challenge of encoding large-scale problems within restricted qubit resources. In this work, we introduce a scalable variational framework based on Pauli correlation encoding (PCE) and apply it to electric...
With the growth of gaming, animation, and virtual reality industries, the demand for efficient generation of stylized 3D assets is rapidly increasing. However, existing approaches still struggle to jointly preserve style fidelity, geometric consistency, and generation efficiency, as most of them still rely on indirect 2D-to-3D...
Multi-turn long-horizon planning is critical for foundation model agents, yet how to fundamentally improve it remains unclear. Existing models are trained on uncontrollable and opaque Internet data, making it difficult to identify how planning ability is acquired, shaped, and integrated. To address this challenge, we introduce a...
Let $G$ be a finite group and let $V$ be a finite-dimensional $G$-module over a field $k$. We construct explicit counterexamples in characteristic $2$ to several questions and conjectures of Wehlau concerning Noether numbers. For the $2$-group $G=D_8$, we exhibit a submodule $U\subseteq V$, with $\dim_kU=5$ and $\dim_kV=6$, such that...
We study doubly reflected backward stochastic differential equations with jumps and two completely separated right-continuous with left limits barriers in a filtration generated by an inhomogeneous Levy process. We establish existence and uniqueness results under a stochastic Lipschitz condition on the driver by means of a...
For a (molecular) graph $G$ and any real number $α\ne 0$ , the zero-order general Randić index , denote by $^0R_α$, is defined by the following equation: \begin{align*} {^0R_α} (G) =\sum_{v\in G}d_G (v) ^α (α\in \mathbb{R}-\left\{0\right\}) . \end{align*} The deficiency of $G$, denoted by $def(G)$, is equal to the cardinality of...
In recent work on course allocation, Rodríguez and Manlove consider the complexity of finding a stable assignment under four notions of stability, including two coalitional notions. In one case, which they call pair-size stability, they show that a stable assignment always exists and they provide a polynomial-time algorithm to find...
For a finite system with reference measure $p$ and positive weight $\ell$, the variational free energy $F(Q;p,\ell)=D(Q\,\|\,p)-E_Q[\log\ell]$ satisfies the exact identity $\log Z(p,\ell)=-F(Q;p,\ell)+D(Q\,\|\,π)$, where $Z$ is the partition function and $π$ the associated Gibbs measure. For a uniform reference measure this is the...
Single-cell light microscopy images have become an important data source for characterizing cell phenotypes, but their complexity and heterogeneity pose challenges to high-throughput automated analysis. Existing representation learning methods mostly rely on task-oriented modeling, which is limited by specific datasets and predefined...
Additional food sources and predator competition are both known to impact the dynamics of predator-prey models. The Bazykin model of predator competition, with Holling type II functional response, possesses a rich bifurcation structure consisting of a focus-type degenerate Bogdanov-Takens bifurcation of codimension 3, and a degenerate...
Direct feedback alignment (DFA) trains hidden layers with fixed random projections of the output error, avoiding the transposed-weight backward pass of backpropagation (BP). We study a failure mode of DFA training that is distinct from feedback quality: the local weight update is calculated by an outer product, so anisotropy can enter...
Automated market makers (AMMs) for prediction markets descend from market scoring rules, where a mechanism operator subsidizes a market to aggregate beliefs about uncertain events. The existing literature has focused on bounding the total worst-case loss to the subsidizer, but has not addressed how that loss is distributed across...
We formulate an obstruction and residual-completion theory for Newman-Janis-type deformations of black hole seeds, where Newman-Janis-type includes both the original complex-coordinate Newman-Janis algorithm and modified prescriptions such as non-complexification variants. The Newman-Janis algorithm is treated as an off-shell map from...
The ultrametric skeleton theorem associates with every compact metric probability space $(X,d,μ)$ a compact subset $S\subseteq X$ of ultrametric distortion $O(1/\varepsilon)$ and a probability measure $ν$ supported on $S$ such that $ν(B_d(x,r))\leqμ(B_d(x,C_\varepsilon r))^{1-\varepsilon}$. We prove a stronger version with a...
Ecological communities are shaped by the interplay between trophic interactions and infectious disease, yet how spatially mediated interactions influence disease-driven ecosystem dynamics remains poorly understood. Here, we develop a gravity-based eco-epidemiological framework for a tri-trophic food chain in which trophic interaction...
We study entanglement entropy in Deep Inelastic Scattering (DIS) using the dipole formulation of the high-energy limit of QCD. We argue that a reduced density matrix arises in low $x$ DIS due to a trace over unobserved color degrees of freedom and we obtain entanglement entropy in terms of dipole multiplicities, directly from the von...