Quantum Physics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 3 September
Self-sustained oscillations combine finite-amplitude stabilization with a neutral phase degree of freedom. We develop this bifurcation-based framework for driven-dissipative collective spins and show that the microscopic structure of the U(1)-covariant dissipation selects the background attractor, while the explicit U(1)-breaking channel governs its local bifurcation response. In the thermodynamic-limit mean-field dynamics, a single linear U(1)-covariant jump produces only polar fixed-point backgrounds, whereas nonlinear covariant dissipation provides amplitude-dependent saturation and stabilizes a finite-latitude self-sustained-oscillator manifold through a supercritical Hopf bifurcation. Under coherent U(1) breaking, exact resonance leads to a reversible double-zero degeneracy with vanishing critical frequency rather than a standard Hopf onset. Finite detuning unfolds this singularity into a genuine finite-frequency Hopf boundary, which exists only on the self-sustained-oscillator side and can be either supercritical or subcritical. By contrast, a single linear dissipative U(1)-breaking jump cannot generate a standard Hopf instability: when its phase-pinning invariant vanishes the azimuthal direction remains neutral, whereas otherwise the phase-locked fixed points have a purely real Jacobian spectrum. These results establish a general design principle: nonlinear covariant dissipation selects the selfsustained background, while the structure of the symmetry-breaking channel determines whether the resulting local response is double-zero, genuinely Hopf, or non-Hopf.
Probability
Epidemics with avoidance and isolation on $\mathbb{Z}^d$
The contact process with avoidance is a generalization of the classical contact process (SIS epidemic) that introduces a mechanism for healthy individuals to avoid their infected neighbors. Let $G$ be a directed graph. At each time $t$, each vertex is either healthy or infected and each edge is either active or inactive. Each infected vertex infects each healthy neighbor across each active edge at rate $\lambda$ and recovers at rate $1$, while each active edge pointing from an infected vertex to a healthy vertex becomes inactive at rate $\alpha$. An inactive edge becomes active when its tail vertex recovers. This model has been previously studied on $\mathbb{Z}$, the $n$-cycle $\mathbb{Z}_n$, and the $n$-star graph; here we extend the study of this model to lattices $\mathbb{Z}^d$, $d \geq 2$. We show that for every $d \geq 2$ and fixed $\alpha > 0$, there exist constants $\lambda(\alpha,d)^-$ and $\lambda(\alpha,d)^+$ such that for all $\lambda < \lambda(\alpha,d)^-$ the infection dies out almost surely while for all $\lambda > \lambda(\alpha,d)^+$ the infection persists indefinitely with positive probability. Furthermore, we show that both $\lambda(\alpha,d)^-$ and $\lambda(\alpha,d)^+$ scale like $1/d$ as $d \rightarrow \infty$ and that there exists a constant $C(\alpha)$ such that when $\lambda > C(\alpha)/d$ the process has a nontrivial invariant measure for $d$ sufficiently large. Our methods and most of our results also apply to the SIRS model and a related model in which infected vertices enter an isolated state at rate $\alpha$ and transition from both isolated and infected to healthy at rate $1$.
Read the full abstract
Machine Learning
Time Without Timesteps: Simulating Coupled Dynamical Systems via Self-Consistency
Numerical simulation of dynamical systems is usually organized as a causal march through time: each state is computed from the previous one. We explore a different formulation for coupled systems. For each subsystem type we train a neural surrogate mapping a full driving trajectory and initial condition directly to a full output trajectory; following classical waveform relaxation, coupled systems are assembled by enforcing self-consistency among these trajectories: simulation becomes a fixed-point problem over complete trajectories rather than a stepwise rollout. On coupled van der Pol oscillators and Hodgkin-Huxley neuron networks, sequential depth becomes the number of solver iterations: 4-10 Newton iterations where the reference integrator takes 1500 steps. The gradient likewise loses its time recursion: it becomes a linear system solved by GMRES at memory independent of solver depth. A single scalar measured from the learned operator, the spectral radius of its Jacobian, predicts in advance where the coupled solve will converge; past that boundary, unrolled backpropagation diverges and a Neumann adjoint fails, while the implicit gradient remains correct to 0.04%. We report where the approach succeeds and where surrogate error degrades it.
Read the full abstract
Sound
Neural Music Enhancement with Dual Time-Frequency Spectral Representations for Prediction and Discrimination
Non-professional music recordings shared online often suffer from background noise and reverberation, degrading perceived quality and limiting reuse. This paper proposes DSME, a music enhancement model based on dual time-frequency spectral representations. Within a generative adversarial framework, DSME uses short-time Fourier transform (STFT) spectra for generation and constant-Q transform (CQT) spectra for discrimination. Leveraging STFT's fixed window, invertibility, and predictability, the generator estimates clean amplitude-phase spectra from degraded inputs and reconstructs waveforms via inverse STFT. Exploiting CQT's log-frequency, variable-window structure aligned with musical octaves, we design an octave-segmented CQT discriminator. We also introduce a chroma-spectrum loss to emphasize pitch and harmonic consistency. Experiments show DSME outperforms baselines in objective and subjective tests, validating the effectiveness of the dual-spectrum approach.
Read the full abstract
Information Theory
Finite-SNR Closed Form for the Square Marčenko-Pastur MIMO Capacity
For the canonical independent and identically distributed (IID) single-user multiple-input multiple-output (MIMO) channel with full channel state information at the transmitter (CSIT), the asymptotic capacity per receive antenna is a waterfilling integral over the Marčenko-Pastur law, whose shape depends only on the ratio of transmit to receive antennas. Closed forms are known for every ratio other than the square one, and there only above a finite signal-to-noise ratio (SNR). The square ratio behaves differently because the Marčenko-Pastur support reaches the origin. The waterfilling cutoff then stays strictly inside the support at every finite SNR, and the water level has been available only numerically. This letter closes that gap with a classical trigonometric parametrization of the square law, which removes the edge singularity and replaces the moving cutoff by a single angle. Both the SNR and the capacity become explicit functions of that angle, so the capacity curve is traced by sweeping the angle rather than by solving a scalar constraint at each operating point. The same parametrization delivers the low-SNR behavior in closed form, where the slope of capacity in SNR equals the upper edge of the limiting spectrum.
Read the full abstract
Machine Learning (Statistics)
ALRA: Adaptive Local Relational Alignment for Logit-Based Pre-training Distillation of Autoregressive Language Models
Logit-based knowledge distillation for autoregressive language models usually aligns teacher and student next-token distributions over the entire vocabulary. However, this global objective overlooks relative preferences among likely token alternatives. Existing local approaches often select candidate tokens from either the teacher or the student alone. Teacher-only selection can miss tokens that the student considers likely, while student-only selection can rely on an inaccurate ranking early in training. We propose Adaptive Local Relational Alignment (ALRA), a position-specific framework combining student proposals with teacher guidance. At each valid prediction position, the student proposes likely tokens, while the teacher's most probable token is included as an anchor. ALRA adjusts the number of selected tokens according to how broadly the teacher distributes probability within this candidate set relative to the current batch. Adaptive Local Divergence retains the mass-matching term and separately matches the relative token distributions within the selected and remaining vocabulary regions. Unlike the exact full-vocabulary decomposition, it replaces the teacher-mass coefficients of the two conditional terms with unit coefficients, preventing either term from being downweighted solely because its region has low teacher probability. Student-Weighted Pairwise Relational Alignment emphasizes high-probability token pairs with small student probability gaps and gives less weight to unlikely or clearly separated pairs. Experiments on The Pile with randomly initialized 200M- and 500M-parameter students across nine zero-shot benchmarks yield average accuracies of 36.62% and 37.40%. ALRA exceeds the strongest competing distillation baseline by 0.94 and 0.83 percentage points and improves over pre-training without distillation by 2.31 and 2.91 points, respectively.
Read the full abstract
Plasma Physics
Minute-Scale High-Fidelity Gyrokinetic Simulations with Portability from Laptop to Supercomputer
Global gyrokinetic particle simulations remain computationally expensive, as they demand both adequate marker statistics and three-dimensional field solvers. In this work, we present a hybrid spectral method within the particle-in-Fourier (PIF) framework and implement it in the electrostatic model of GTC. Charge scatter and field gather are performed between particles and fields on a two-dimensional poloidal mesh, while the corresponding Poisson solver is discretized using radial finite differences and poloidal $m$-harmonics. Truncated spectral transforms are employed to connect multiple representations for fields, avoiding costly particle-grid operations for each individual $m$-harmonic within the particle loop. Benchmarks against conventional particle-in-cell (PIC) simulations successfully reproduce single-$n$ ion temperature gradient (ITG) mode structures and dispersion relations, as well as multi-$n$ nonlinear ITG transport and its regulation by zonal flows. Compared to conventional PIC, the proposed method reduces the effective problem size by more than a factor of 48 and achieves a speedup of over two orders of magnitude for single-$n$ cases. A 2000-step single-$n$ simulation with approximately 2 million markers completes in 78.2 seconds on a laptop GPU, while multi-$n$ turbulence simulation also completes within minutes. Furthermore, the elimination of toroidal particle-shift communication yields promising preliminary scaling performance on multiple NVIDIA A100 GPUs. The numerical scheme is broadly applicable for accelerating particle simulations on platforms ranging from laptops to supercomputers.
Read the full abstract
Number Theory
Resolving a conjecture on permutation polynomials over $\mathbb{F}_{2^n}$
Let $\delta\in\mathbb{F}_{2^n}$ satisfy $\operatorname{Tr}_{\mathbb{F}_{2^n}/\mathbb{F}_2}(\delta)=1$. We study the permutation behavior of $$ f(x) = \left(\frac{1}{x^2+x+\delta}\right)^{2^k}+x $$ over $\mathbb{F}_{2^n}$. Helleseth and Zinoviev proved that $f(x)$ is a permutation for $k=0,1$, and remarked that numerical evidence suggests that no other cases occur. In this paper, we confirm their assertion by proving that, for $0\leq k<n$, $f(x)$ is a permutation of $\mathbb{F}_{2^n}$ if and only if $k=0$ or $k=1$.
Read the full abstract
Neural and Evolutionary Computing
Efficient Constant Optimization for Symbolic Regression with GPU-Accelerated Tree-Based Genetic Programming
Constant optimization refines the numerical coefficients of candidate expressions in tree-based genetic programming for symbolic regression. But its per-generation cost has led modern GPU-accelerated frameworks to omit it or restrict it to lightweight forms. We present a GPU-resident, batched Levenberg–Marquardt solver that optimizes constants across a structurally heterogeneous population of expression trees using a fixed number of population-wide CUDA launches per iteration. Reverse-mode automatic differentiation assembles the per-tree Jacobian in one backward sweep, making the dominant per-iteration cost independent of the number of constants per tree, and a double-precision delivery guard guarantees that returned constants are never worse than their initial values. On early-generation populations, the solver sustains up to $5.1{\times}10^{5}$ trees per second on an NVIDIA A100; at a GPU-saturated benchmark configuration it delivers roughly $9.9{\times}$ the throughput of Operon running on a 64-core EPYC 7763, while matching fp64-reference quality. Integrated in-process into EvoGP, the solver enables end-to-end search to recover governing equations on $10$ of $18$ constructed problems versus 0 for stock EvoGP. Our code is at https://github.com/TensorConv/CuSR.
Read the full abstract
Systems and Control
Taylor-Informed Predictive Cost Adaptive Control for Quadrotors with Online Gravity-Trim Adaptation
This paper develops Taylor-informed predictive cost adaptive control (PCAC) for quadrotors with online gravity-trim adaptation. First-, second-, and third-order expansions of the nonlinear dynamics about nominal hover define sparse sampled-data dictionaries for row-wise recursive least-squares identification with variable-rate forgetting. At each step, the identified predictor is linearized at the current state, and its Jacobian is fixed over the prediction horizon. The identified vertical dynamics also estimate the vehicle mass and gravity-trim input, eliminating fixed nominal gravity compensation. Simulations with an abrupt payload change and aggressive helix tracking show that the higher-order predictors improve prediction and tracking while preserving the standard PCAC formulation.
Read the full abstract
Machine Learning
From Zero to Hero: An Open LLM Ecosystem for Armenian
Pretraining data for Armenian, a morphologically rich and low-resource language, is scarce, and no open Armenian LLM has been released with the data and recipe needed to reproduce it. To address this gap, we curate and release two datasets. ArmWeb is an extensively validated corpus of 4.37M Armenian news documents. ArmSTEM is a parallel English-Armenian collection of 373K math and science problems with step-by-step solutions, translated into Armenian and verified through both answer-preserving LLM judgment and human evaluation. Continued pretraining of Gemma-4-E4B on these datasets yields arm-gemma-e4b, which outperforms every existing open Armenian model as well as its unadapted base, and is the first open Armenian LLM with complete training data and recipe. Our ablations show that news-only continued pretraining improves fluency while eroding knowledge, a pattern we also observe in existing Armenian models, and that a small share of verified translated STEM data reverses the loss. We further find that the largest public Armenian corpora overlap web-derived evaluation panels heavily, including a train/test self-overlap inside FineWeb-2. We openly release all data, models, and code.
Read the full abstract
Computer Vision
P-CORE: Self-Supervised Surface Consistency for Point-Based Neural Editing
Advances in neural rendering have enabled high-fidelity multi-view reconstruction of 3D scenes. However, free-form non-rigid shape editing remains a significant challenge. Point-based neural representations are highly desirable for multi-view reconstruction because they lack fixed connectivity, which does not constrain the learned surface topology to that of the initialization. Yet this same property causes point-based representations to struggle with holes and surface discontinuities under large deformations. To address this, we propose a novel self-supervised method to enable point-based representations to adapt to large deformations without requiring ground truth multi-view images of deformed geometry. The key idea is to generate random deformations and to ensure consistency in the predicted surface before and after deformation. In particular, the surface prediction from the deformed point cloud should be the same as the deformation applied to the surface prediction from the original point cloud. We incorporate our approach into attention-based point representations, which differ from splatting-based point representations in their use of a learned interpolation kernel between points as opposed to a Gaussian kernel around each point. This learned interpolation kernel can learn to adapt to large deformations, without requiring addition or removal of points. We show that our framework significantly enhances its robustness to large deformations. Experiments on synthetic geometry editing benchmarks (Neural Editor, Objaverse) demonstrate that our approach outperforms existing point-based methods in zero-shot editing and significantly reduces artifacts. Furthermore, qualitative results on the DTU and Mip-NeRF 360 datasets demonstrate our method's effectiveness on real-world scenes.
Read the full abstract
Systems and Control
Taylor-Informed Indirect Adaptive Predictive Control Using Jacobian-Frozen Affine Predictors
This paper develops a Taylor-informed indirect adaptive predictive control framework for nonlinear sampled-data systems using Jacobian-frozen affine predictors. A finite Taylor expansion approximates the sampled nonlinear dynamics, and recursive least squares (RLS) identifies its polynomial coefficients online. At each sampling instant, the Jacobian of the identified map is evaluated at the current operating point and frozen over the prediction horizon, yielding an affine predictor for model predictive control. In contrast to generic nonlinear feature dictionaries, the implemented polynomial dictionary is a forward-Euler/Taylor-structure-informed reduced dictionary. Exact joint-odd symmetry eliminates even-total-degree monomials, whereas additional forward-Euler-informed pruning constitutes a deliberate model reduction. Numerical simulations on an unstable nonlinear benchmark compare different Taylor degrees. The results show that higher-order models improve tracking accuracy as the operating point moves farther from the expansion point while maintaining comparable control effort. The complete MATLAB implementation is publicly available to facilitate reproducibility.
Read the full abstract
High Energy Physics (Theory)
Conical Defects in JT Gravity from BF Theory: Quantization, Fusion, and Weighted Moduli Spaces
We construct a gauge-invariant BF representative of a conical defect in Jackiw–Teitelboim gravity, with its elliptic direction fixed by the reconstruction of the gravitational variables. In the gravitational sector, the BF observable reproduces the metric defect insertion and its distributional curvature source, and fixes the holonomy around the defect to an elliptic conjugacy class. For a single defect on the disk, the position integral combines with the residual \(\mathrm{PSL}(2,\mathbb R)\) quotient, leaving a fixed elliptic sector. The exact boundary quantization of this sector gives the elementary kernel and one-defect disk amplitude directly, without using a matrix model. The same local BF operator describes both sharp and blunt defects. Their difference appears only when several defect positions are integrated. BF source composition makes the fractional deficits additive. Requiring a fused source to remain in the gravitational cone sector gives the admissibility condition for defect collisions, while positivity of the hyperbolic area on each component gives the corresponding stability condition. These conditions coincide with those for Hassett weighted curves. Combining these BF results with the known Hassett compactification and conical Weil–Petersson geometry yields the fixed-order genus-zero contact terms and recovers the cluster expansion of deformed JT gravity. Thus sharp and blunt defects differ through the compactified geometry of their relative positions, rather than through the elementary BF operator.
Read the full abstract
Combinatorics
Counterexamples to two conjectures on the diameter of clique-free graphs
Erdős et al. (JCT-B, 1989) conjectured that, for integers $r\ge 2$ and $\delta\ge 2$ with $3r-1\mid\delta$, every connected $K_{2r+1}$-free graph of order $n$ and minimum degree $\delta$ has diameter at most $ \frac{3r-1}{r}\cdot \frac{n}{\delta}+O(1)$. Czabarka et al. (JCT-B, 2021) later proposed the following generalization: for every $k\ge 3$ and $\delta\ge\left\lceil\frac{3k}{2}\right\rceil-1$, every connected $K_{k+1}$-free graph of order $n$ and minimum degree at least $\delta$ has diameter at most $(3-\frac{2}{k})\cdot\frac{n}{\delta}+O(1)$. We disprove the latter conjecture, including its $k$-colorable version, for every $k\ge 7$ and sufficiently large $\delta$. When $k=2r\ge 8$ and $3r-1\mid\delta$, our construction also disproves the conjecture of Erdős et al. (JCT-B, 1989).
Read the full abstract
Systems and Control
Traffic Congestion Control for ARZ Model with an Arbitrarily Large Input Delay
This paper addresses the stabilization problem for Aw-Rascle-Zhang (ARZ) traffic model in the presence of an arbitrarily large input delay. The linearized ARZ model is a $2 \times 2$ hyperbolic partial differential equation (PDE) system with proximal reflection, which introduces significant analytical challenges when combined with input delays. To tackle this problem, we propose a backstepping-based boundary controller capable of stabilizing the linearized ARZ model under these conditions. The input delay is modeled as a transport PDE, which reformulates the entire system into a $3 \times 3$ hyperbolic PDE system. A backstepping transformation is designed to map the original system into a stable target system, enabling the design of a delay-compensated controller. A key technical contribution of this work is that for hyperbolic PDEs with delays, we develop a characteristic-region-wise construction for kernel functions subject to two boundary constraints and close the proof via successive approximation. Another contribution is that we utilize the small-gain theorem for input-to-state stability (ISS) of hyperbolic PDEs. Two simulations are provided to illustrate the effectiveness of the proposed delay-compensated controller: one compares it with a controller without compensation, and the other employs real traffic vehicle data to validate its effectiveness.
Read the full abstract
Physics and Society
Large-Language Models as a Cognitive Virus
Large-language models (LLMs) are rapidly becoming part of human culture, reshaping how information is produced, transmitted, and used. Here we propose that their diffusion can be understood through a viral analogy, with LLM use spreading through populations, becoming embedded in cognitive and cultural practices. We model transitions among uncoupled, coupled, and persistently dependent users, and show that the interplay between social transmission, recovery, and collective reinforcement can generate tipping points and technological lock-in. A central consequence is the possibility of runaway dynamics: once a critical threshold is crossed, small increases in adoption can trigger rapid population-level shifts toward persistent dependence, with abrupt losses in cognitive competence. The same framework, however, identifies conditions for cognitive immunization, based on reducing transmission and facilitating reversibility. Our results highlight how LLM adoption may involve nonlinear collective transitions with important consequences for cognitive autonomy.
Read the full abstract
Numerical Analysis
Learning Informative Prior with Infinite-Dimensional Continuous Normalizing Flow for Bayesian Inverse Problem
This paper addresses infinite-dimensional Bayesian inference for inverse problem of partial differential equations with model parameters in infinite-dimensional Hilbert space. To effectively incorporate prior information, we propose a novel continuous normalizing flows based infinite-dimensional model. Specifically, by introducing a well-defined neural ordinary differential equation in infinite-dimensional space, a simple reference measure can be transformed into a more complex measure which encodes the prior information. A corresponding theoretical framework is established to ensure the well-posedness of our proposed Bayesian prior in infinite-dimensional space. We also provide training methods of the prior for two distinct data settings, along with two sampling algorithms for the resulting Bayesian posterior. The proposed framework is applied to three representative inverse problems: the simple smooth inverse problem, inverse scattering problem, and the inverse heat conduction problem. Numerical experiments support the theoretical analysis and demonstrate the efficiency of the proposed algorithms.
Read the full abstract
Machine Learning
Gradients Know What Outcomes Don't: Unlocking Reinforcement Learning for LLM Reasoning with Gradient-Aligned Rewards
Reinforcement learning from verifiable rewards (RLVR) drives chain-of-thought reasoning in large language models, yet its binary outcome reward cannot distinguish among correct trajectories. Existing dense reward alternatives, from surface heuristics to process reward models, either ignore the expert solutions already present in training corpora or require expensive offline annotation. We propose Gradient-Aligned Reward (GAR), which operates in the policy's own gradient space: truncated backpropagation through the output projection layer extracts a compact gradient vector for each rollout, and cosine similarity with an expert-anchor gradient yields a dense, reasoning-aware reward with less than 9% wall-clock overhead. We prove that this cosine admits a multiplicative decomposition into prediction-error and activation-pattern factors, providing a concrete characterization of what the alignment signal measures. On Qwen3-4B and Qwen3-8B, GAR consistently improves over GRPO and other baselines on competition-level math benchmarks and transfers to GPQA Diamond and MMLU-Pro without domain-specific data. Code and data are available at https://github.com/LQgdwind/GAR.
Read the full abstract
Computer Vision
PointGT: Simultaneous Geometry and Texture Editing for Point-Based Representations
We present PointGT, a point-based 3D representation that enables simultaneous editing of object geometry and appearance. Existing reconstruction and view synthesis techniques produce volumetric 3D representations that are high-quality and photorealistic, but are difficult to edit. In particular, recent efforts to enable texture editing for 3D Gaussian Splatting representations are not compatible with geometry edits and deformations. Our method combines a point-based representation that is well-suited for geometry deformations with a learned UV mapping technique that enables high-resolution texture editing. We show that PointGT enables fine-grained editing of both geometry and texture in point-based neural representations with high rendering quality.
Read the full abstract
Artificial Intelligence
Fresh Memory, Stale Plans: Derivation Currency for Distributed LLM-Agent Memory
A large language model (LLM) agent that inherits a plan through shared memory can hold the latest requirement yet act on a plan derived from an older one: fresh memory, stale plan. Freshness checks miss this failure because they compare local copies with current state (observation currency) rather than the inputs the plan was derived from (derivation currency). Planfence makes derivation currency checkable after a handoff. Stored plans carry exact links to their recorded inputs; before a protected action, Planfence follows those links to an action-specific dependency frontier, asks each input's owner for its current head, refreshes what changed, and allows one replan before blocking. Application code supplies the links and declares the scope; no shared memory service is required. Holding the native S-Bus validator fixed, supplying inherited input versions raises detected handoff conflicts from 0/30 to 30/30: retained evidence is the missing ingredient. In 30 live five-agent workflows with a revision inserted after planning, a freshness-only executor acts on the stale plan every time, whereas Planfence, like a centralized-lineage baseline that requires a shared store, completes all 30 correctly. In matched replay under emulated LTE traces, Planfence's stall stays within 143-237ms per action across a 64$\times$ range of update rates while per-update synchronization grows from 52 to 1196ms; synchronization is cheaper only at the lowest tested rates. Scoping queries to the declared dependencies holds traffic at 8.1KiB per action, a tenth of all-key validation at 128 keys; at full scope the two tie.
Read the full abstract
Analysis of PDEs
Exterior Serrin Rigidity for the Homogeneous{k}-Hessian Equation
In this paper, we study the exterior overdetermined problems for the homogeneous k-Hessian equations $$\sigma_k(D^2u)=0\quad\text{in}~\mathbb{R}^n\setminus\overline{\Omega}$$ in three dimensional regimes. For $2\le k<n$ and smooth strictly star-shaped domain $\Omega\Subset\R^n$, we establish ball rigidity results in all three dimensional regimes. If $2\leq k<\frac{n}{2}$, the solution with boundary value $-1$ and limit zero at infinity is treated under strict $(k-1)$-convexity of $\partial\Omega$. If $k=\frac{n}{2}$, the solution with boundary value $0$ and logarithmic growth at infinity is considered under strict $(k-1)$-convexity of $\partial\Omega$; if $k>\frac{n}{2}$, the solution with boundary $1$ and fundamental power-growth at infinity is considered under strict $(k-1)$-convexity of $\partial\Omega$. In each case, for a positive constant $c$, $|Du|=c>0$ on $\partial\Omega$ forces $\Omega$ to be a Euclidean ball and determines the solution explicitly. The three arguments are organized by the single parameter $q=(n-k)/k$ and the scale function \[ F_q(t)=\frac{t^{1-q}-1}{1-q},\qquad F_1(t)=\log t. \] After normalization, $U=F_q(v)$ and $M_q[v]=vD^2v-qDv\otimes Dv$ satisfies $\sigma_k(M_q[v])=0$. A unified contact-point calculation gives the sharp bound $|Dv|\le b$;a boundary viscosity contact then yields $H_k\ge qbH_{k-1}$. The integral mechanism that closes the argument changes at the critical exponent: Rellich–Pohozaev identities are used for $q>1$, a scale-invariant Wronskian–Newton current for $q=1$, and a renormalized Newton–Jacobi mass for $q<1$. When $q\leq 1,$ we also obtain a global strict k-convex defining function by Minkowski gauge, this help us extend the exterior construction from convex domains to strictly star-shaped domains.
Read the full abstract
Information Retrieval
SciLENS: RL-Driven Autonomous Agents for Scientific Localized Evidence Navigation and Synthesis
Scientific literature synthesis agents increasingly rely on proprietary online services, limiting reproducibility, privacy, and offline deployment. To address this challenge, we introduce SciLENS Scientific Localized Evidence Navigation and Synthesis), a fully local autonomous agent framework operating on a dual-tier infrastructure indexing approximately 12 million academic records. SciLENS pioneers the integration of structural visualization as an actionable tool within the reasoning loop, enabling the agent to compress complex citation topologies into validated data-driven charts and thereby mitigate context exhaustion during macro-level synthesis. To train the agent without human annotation, we develop an automated data synthesis pipeline that extracts multi-hop subgraphs from a citation knowledge graph, verified by cross-model consensus among 20 frontier models. The agent is subsequently aligned through a reverse-decomposition rubric strategy that provides fine-grained process rewards for early planning and strict evidence grounding. Evaluations across six scientific benchmarks encompassing standard QA, citation accuracy, factual reasoning, and structural synthesis demonstrate that SciLENS significantly outperforms open-source baselines and achieves performance comparable to GPT-5.2 and Gemini-3.0-pro. Our source code and data are released at https://github.com/LQgdwind/SciLENS.
Read the full abstract
Machine Learning
A Large Open Multi-Energy Corpus of Soil Compaction Tests, with Machine-Learning Baselines
Every engineered fill is specified by a maximum dry density and an optimum moisture content. Each determination needs a full Proctor test. Published correlations rest on one to four hundred specimens, usually from one laboratory at one compactive energy, and are seldom released. This paper releases a corpus without those limits. It holds 2,854 laboratory compaction tests from six public sources, across 162 provenance groups and four Proctor energy levels, with fines from 1.5 to 100%. Every record is audited to the Proctor method its source names, and no energy is inferred. Screening on the zero-air-voids condition removed 11.8% of harmonised records, and 5.7% of those with a measured specific gravity. A material share of published compaction data is physically impossible. The optimum degree of saturation over the corpus is 0.815 at a coefficient of variation of 11%. That is a baseline, not a constant. Both parameters are then estimated from one classification suite and the compaction standard. A tabular foundation model reaches R2 0.824 for density and 0.784 for water content under random folds. It reaches 0.727 and 0.696 with folds drawn around provenance, and 0.520 and 0.614 with a whole source held out. Compactive energy is negligible marginally yet decisive conditionally. Density on the 66 modified-Proctor records is predicted at R2 0.740 with it and -0.651 without. Symbolic regression yields closed forms coupled through a phase relation. No predicted pair can then exceed the zero-air-voids line. The predictions are for screening, not acceptance.
Read the full abstract
Materials Science
Polaron Self-Trapping Rates from First Principles
Polaron formation, also known as self-trapping, is a process akin to nonradiative carrier capture at point defects or impurities. In this work, we develop the formalism to determine how long it takes to form a small hole or electron polaron from first principles. We employ an accurate, fully first-principles approach based on a Koopmans compliant hybrid functional. The self-trapping rate is the product of two components: the nonradiative capture coefficient, which we evaluate using a one-dimensional approximation, and the maximum density of polaron sites, whose physics we elucidate based on finite-size interactions present in supercells. We apply our methodology to several technologically relevant materials known to host hole polarons, Ga$_2$O$_3$, Al$_2$O$_3$, BeO, KBr, MgO, NaCl, SiO$_2$, SnO$_2$, TiO$_2$, and ZnO, and to an electron polaron in rutile TiO$_2$. We also study an emerging semiconductor, rutile GeO$_2$, where we find that polaron formation could hamper $p$-type conductivity. The calculated self-trapping lifetimes span 7 orders of magnitude, from $10^{-1}$ to $10^6$~ps, in agreement with experiments where available, and providing detailed insight into the dynamics of carrier localization and relaxation in solids.
Read the full abstract
Distributed Computing
Latency-Aware Orchestration for Multi-Agent LLM Workflows on Heterogeneous GPUs
Concurrent multi-agent workflows expose future dependencies and serving-state requirements while running on heterogeneous GPU pools with time-varying load, model residency, and resource availability. The logical workflow defines the required computation, whereas its physical scheduling units, model-lifecycle actions, resource ordering, and placement must be selected according to the observed pool state. We present a prediction-guided runtime that uses workflow forecasts to construct and optimize a physical execution graph. Predictor estimates device-specific activation latency, peak memory, and model-loading cost, then propagates these predictions through workflow dependencies to forecast activation readiness and future model demand. Constructor builds semantics-preserving fusion and model-lifecycle alternatives, while Scheduler jointly optimizes their selection, placement, and execution order based on the live pool state. Across a workload spanning three workflow scenarios on a heterogeneous GPU pool, our system reduces end-to-end makespan and overall p95 completion latency under burst arrivals by up to 36.8% and 25.9%, respectively, over state-of-the-art workflow schedulers. It also saves up to 24.63 GPU-s per completed session.
Read the full abstract
Computer Vision
Laplacian Frequency Hierarchies for Efficient 3D Gaussian Splatting Training
A key bottleneck in 3D Gaussian Splatting training is the continual growth of Gaussian primitives, which increases optimization cost and slows convergence, especially at high resolutions. We propose Laplacian Frequency Hierarchies, a simple yet efficient 3DGS scheme that combines Laplacian image decomposition with coarse-to-fine, frequency-staged training. After fitting lower-frequency structure, we archive the corresponding Gaussian field so that subsequent fields can optimize higher-frequency residuals without carrying the full primitive burden, and we compose the rendered components in the image domain via a Laplacian-style reconstruction at inference time. This design reduces the number of active Gaussians during training, thereby lowering optimization overhead and accelerating training. The proposed scheme is plug-and-play and orthogonal to prior 3DGS accelerations: it can be directly combined with strong backbones such as Taming-3DGS and FastGS to improve training speed with competitive reconstruction quality. It achieves average speedups of 1.73x and 1.21x at 1K setting, and 1.74x and 1.33x at 4K setting on Taming-3DGS and FastGS, with larger gains on more challenging scenes and increasingly pronounced benefits at higher resolutions.
Read the full abstract
Quantum Physics
Redesigning quantum theory
QBism understands quantum mechanics to be probability theory supplemented by additional nonclassical coherence conditions. In this dissertation, we develop these nonclassical coherence conditions from first principles, emphasizing the role of a well chosen reference measurement. After treating standard probability on subjective Bayesian lines, we demonstrate an equivalence between the QBist approach and the existing framework of generalized probabilistic theories. We show that the fundamental nonclassical coherence relation may almost always be taken to be a gentle modification of the law of total probability, and give a coherentist account of when an experimental scenario has a classical explanation. Finally, we show that when the reference measurement is chosen to correspond to a complex projective 3-design, the shape of quantum state space is implicit in the probabilities which characterize the reference measurement itself. Thus coherence with this single reference measurement, properly understood, implies coherence with all of finite dimensional quantum mechanics. We then attempt a modest reconstruction of quantum theory along these lines, one practical consequence of which is a method for self-testing complex projective $t$-designs for $t\ge 3$ in a theory-agnostic way.
Read the full abstract
Differential Geometry
Singularity Models of Finite-Time Kähler-Ricci Flows
We study the singularity type and models of the Kähler–Ricci flow on compact manifolds constructed from the 1-parameter foliation of a circle-bundle over a product of Kähler–Einstein manifolds $N := N_1 \times \cdots \times N_r$, with metric constructed using the ansatz considered in \cite{DW2011}, \cite{WW} et. al. In the earlier work \cite{FT} by the authors, we considered the “two-bolt” case where both ends of the foliation close with the “bolt” $N$. In this article, we continue our work on the more subtle “nut-bolt” and “two-nut” cases. The former has one end of the interval closes with a nut-type collapse (i.e. $N' := N_2 \times \cdots \times N_r$) and the other with a bolt (i.e. $N$). The compactification $\widehat{M}$ is then a $\mathbb{CP}^{m+1}$-bundle over $N'$. The “two-nut” case is one that both ends close with nut-type collapses, necessarily two of the $N_i$'s must be $\mathbb{CP}^{m_0}$ and $\mathbb{CP}^{m_\ell}$, and the compactification $\widehat{M}$ is a $\mathbb{CP}^{m_0+m_\ell+1}$-bundle over $\prod_{k\geq 3}N_k$. We proved that in all “two-bolt”, “nut-bolt” and “two-nut” caess the singularity must be of Type I. Furthermore, we study the pointed Cheeger-Gromov limit of the rescaled and dilated sequence of the flow in all of three cases, and prove that the limit model must be $(\Sigma^{m+1}, g_\Sigma(t)) \times (\mathbb{C}^{k}, \textrm{flat})$ with $m, k \geq 0$, where $\Sigma$ is one of the following: $\mathbb{CP}^{m+1}$, $\textrm{Tot}(\mathcal{L}^{\oplus(m+1)})$, or a projectivization $\mathbb{P}\big(\mathcal{O}^{\oplus(m_0+1)} \oplus \mathcal{L}^{\oplus(m_\ell+1)}\big)$ with $m_0 + m_\ell = m$, and $\mathcal{L}$ is a line bundle over the product of some of the $N_1, \cdots, N_r$ factors. The metric $g_\Sigma(t)$ is a Kähler-Ricci shrinker satisfying the circle-bundle ansatz.
Read the full abstract
Computation and Language
FPCO-Dialog: A Multi-Turn False-Premise Benchmark for Correction and Cooperation in Vision-Language Models
Vision-language models (VLMs) are increasingly deployed in multi-turn settings where users may describe visual content with incorrect assumptions. Yet existing evaluations rarely isolate how models respond when the same visually grounded false premise persists across dialogue turns. We introduce FPCO-Dialog, a benchmark for evaluating correction and cooperation behavior in VLMs under repeated false premises. FPCO-Dialog contains 1,080 images and 10,800 question turns, stratified by visual complexity, object category, and false-premise class, and uses a 10-turn protocol in which a correct dialogue prefix is followed by repeated false-premise referring expressions. We evaluate 20 commercial and open-source VLMs with a model-agnostic protocol and CorrTP@K, a correction-rate metric over false-premise turns, scored by two independent detectors. FPCO-Dialog reveals substantial and persistent cross-model differences in aggregate correction tendency, model-specific turn-wise dynamics, and systematic variation across false-premise types under the benchmark's substitution distribution. The dataset, evaluation protocol, model outputs, detector labels, and code are available.