arXiv preprints from January 1, 2026 through September 23, 2026 — 14:37:30 EST
Recent advances in large language models (LLMs) enable agentic systems trained with reinforcement learning (RL) over multi-turn interaction trajectories, but practical deployment is bottlenecked by rapidly growing textual histories that inflate token budgets and memory usage. We introduce AgentOCR, a framework that exploits the...
Magnetic Resonance Imaging (MRI) provides detailed tissue information, but its clinical application is limited by long acquisition time, high cost, and restricted resolution. Image translation has recently gained attention as a strategy to address these limitations. Although Pix2Pix has been widely applied in medical image...
We consider Kramers-Fokker-Planck operators with general degenerate coefficients. We prove semiclassical hypocoercivity estimates for a large class of such operators. Then, we manage to prove Eyring-Kramers formulas for the bottom of the spectrum of some particular degenerate operators in the semiclassical regime, and quantify the...
This paper re-examines the empirical Phillips curve (PC) model and its usefulness in the context of medium-term inflation forecasting. A latent variable Phillips curve hypothesis is formulated and tested using 3,968 randomly generated factor combinations. Evidence from US core PCE inflation between Q1 1983 and Q1 2025 suggests that...
We investigate generalized Laurent multiple orthogonal polynomials on the unit circle satisfying simultaneous orthogonality conditions with respect to $r$ probability measures or linear functionals on the unit circle. We show that these polynomials can be characterized as solutions of a general two-point Hermite--Padé approximation...
We provide the first lattice-QCD estimate of the crossover line down to $T\simeq108$~MeV. We introduce a new method that combines the Lee-Yang edge in the complex plane of baryon chemical potential $μ_B$ with universal chiral scaling to determine the $μ_B$ dependence of the QCD chiral critical and pseudo-critical temperatures. By...
Thermal feedback is critical to a range of Virtual Reality (VR) applications, such as firefighting training or thermal comfort simulation. Previous studies showed that adding congruent thermal feedback positively influences User eXperience (UX). However, existing work did not compare different levels of thermal feedback quality and...
The Euclid satellite will measure spectroscopic redshifts for tens of millions of emission-line galaxies. In the context of Stage-IV surveys, the 3-dimensional clustering of galaxies plays a key role in providing cosmological constraints. In this paper, we conduct a model comparison for the multipole moments of the galaxy 2-point...
Over the past three decades, defocus has consistently provided groundbreaking depth information in scene images. However, accurately estimating depth from 2D images continues to be a persistent and fundamental challenge in the field of 3D recovery. Heuristic approaches involve with the ill-posed problem for inferring the spatial...
Video-language models (VLMs) achieve strong multimodal understanding but remain prone to hallucinations, especially when reasoning about actions and temporal order. Existing mitigation strategies, such as textual filtering or random video perturbations, often fail to address the root cause: over-reliance on language priors rather than...
Multimodal Large Language Models (MLLMs) have demonstrated impressive progress in single-image grounding and general multi-image understanding. Recently, some methods begin to address multi-image grounding. However, they are constrained by single-target localization and limited types of practical tasks, due to the lack of unified...
Monocular shape-from-polarization (SfP) leverages the intrinsic relationship between light polarization properties and surface geometry to recover surface normals from single-view polarized images, providing a compact and robust approach for three-dimensional (3D) reconstruction. Despite its potential, existing monocular SfP methods...
The demand for high-resolution, non-invasive imaging continues to drive innovation in magnetic resonance imaging (MRI), but long acquisition times remain a major practical limitation. Although deep learning-based reconstruction methods have enabled accelerated imaging, their predominant supervised paradigm relies on fully-sampled...
We present a new measurement of the Hubble constant ($H_0$) resulting from the first joint analysis of standard sirens with weak gravitational lensing and galaxy clustering observables comprising three two-point correlation functions (3$\times$2pt). For the 3$\times$2pt component of the analysis, we use data from the Dark Energy...
Modern radio telescope surveys, capable of detecting billions of galaxies in wide-field surveys, have made manual morphological classification impracticable. This applies in particular when the Square Kilometre Array Observatory (SKAO) becomes operable in 2027, which is expected to close an important gap in our understanding of the...
In this paper, we introduce two new forms of the dual Hartwig-Spindelb{ö}ck decomposition and employ them to derive explicit representations for several classes of dual generalized inverses. Building on these representations, we further explore and characterize the relationships and properties of these inverses, investigate the dual...
Chiral matter acting as a spin-selective device in biased electron transport is attracting attention for the quantum-technological design of miniaturized electronics. To date, however, experimental reports on spin selectivity are not conclusive. The magnetoresistance in electron transport measurements observed for chiral materials on...
The perovskite-inspired bismuth halide semiconductor Cs3Bi2Br9 is widely investigated as photoactive material for light-conversion applications. However, charge generation and separation are inherently limited by its modest sunlight absorption and strong exciton binding energy, respectively. Here, we demonstrate that both the light...
As large language models (LLMs) transition from general knowledge retrieval to complex scientific discovery, their evaluation standards must also incorporate the rigorous norms of scientific inquiry. Existing benchmarks exhibit a critical blind spot: general instruction-following metrics focus on superficial formatting, while...
Enterprises are confronted with an unprecedented escalation in cybersecurity vulnerabilities, with thousands of new CVEs disclosed each month. Conventional prioritization frameworks such as CVSS offer static severity metrics that fail to account for exploit probability, compliance urgency, and operational impact, resulting in...
The absolute calibration of period-luminosity (PL) relations of Cepheids in the Milky Way (MW) and its nearby galaxies has been a cornerstone in determining extragalactic distances and the current local expansion rate of the Universe. However, the universality of PL relations is still debated; particularly, the metallicity effect on...
Dense retrieval in multilingual settings often searches over mixed-language collections, yet multilingual embeddings encode language identity alongside semantics. This language signal can inflate similarity for same-language pairs and crowd out relevant evidence written in other languages. We propose LANGSAE EDITING, a post-hoc sparse...
LLM agents have emerged as powerful systems for tackling multi-turn tasks by interleaving internal reasoning and external tool interactions. Agentic Reinforcement Learning has recently drawn significant research attention as a critical post-training paradigm to further refine these capabilities. In this paper, we present AT$^2$PO...
Judge Decoding accelerates LLM inference by relaxing the strict verification of Speculative Decoding, yet it typically relies on expensive and noisy supervision. In this work, we revisit this paradigm from first principles, revealing that the ``criticality'' scores learned via costly supervision are intrinsically encoded in the...
We study how syntactic and semantic information is encoded in inner layer representations of Large Language Models (LLMs), focusing on the very large DeepSeek-V3. We find that, by averaging hidden-representation vectors of sentences sharing syntactic structure or meaning, we obtain vectors that capture a significant proportion of the...