Peer Revenue Play credits. Real papers.

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 8 September

Machine Learning

Difficulty-Adaptive Tree-Structured Policy Optimization for Expanding Reasoning Coverage in RLVR

Reinforcement Learning with Verifiable Rewards (RLVR) has been central to the recent success of Large Reasoning Models. However, while RLVR significantly improves single-sample accuracy, it often fails to expand the model's intrinsic reasoning coverage (pass@k) due to limited exploration during training. To address this, we optimize the structural design of train-time rollouts to enhance pass@k. Our analysis identifies three key design principles: (1) difficulty-adaptive rollout can play an important role in expanding pass@k, beyond serving as an efficiency heuristic; (2) tree-based rollout outperforms parallel sampling in discovering correct answers; and (3) sentence-entropy-guided forking overcomes the localization phenomenon of token-level branching to maximize semantic diversity. Building on these insights, we propose DATPO (Difficulty-Adaptive Sentence-entropy-guided Tree-structured Policy Optimization). DATPO integrates difficulty-adaptive tree search with a sibling-diversity advantage term, explicitly promoting semantic diversity to expand reasoning coverage during training. Experiments on mathematical reasoning benchmarks demonstrate that DATPO outperforms baselines especially in pass@k, which directly translates to superior test-time scaling performance.
Read the full abstract

Cosmology and Nongalactic Astrophysics

$\texttt{LensFactory.jl}$: A general-purpose strong lens modeling package

We present LensFactory.jl, an efficient, open-source, general-purpose strong lens modeling package written in Julia. It supports single- and multi-plane lensing with a wide range of analytic lens profiles, and its modular design, built on Julia's multiple dispatch, makes it straightforward to add new profiles or inference methods, while native multi-threading parallelizes the computationally expensive steps, enabling full cluster-scale lens reconstructions within hours on a desktop-class machine. We validate the mass-reconstruction pipeline of LensFactory.jl on two simulated galaxy clusters (Ares and Hera), recovering the convergence and magnification maps well where the lensing constraints lie, and on the JWST-observed cluster SMACS J0723.3-7327, obtaining a strong lens model (image-plane RMS of 0.46”) consistent with previously published results. In addition, we introduce singularity maps as a complementary diagnostic for comparing lens models, probing differences in the higher-order derivatives of the lensing potential, which remain largely unconstrained by standard observables such as image positions, time delays, and flux ratios. LensFactory.jl is accompanied by a set of worked examples covering all core functionality.
Read the full abstract

High Energy Physics (Theory)

Symplectic Chern-Simons theories dual to Abelian rank-$0$ theories

We propose that the 3d symplectic Chern-Simons theory $USp(2n)$ at level $k=n+\frac12+\ell$ with a single chiral multiplet in the fundamental representation is dual, for every pair of positive integers $(n,\ell)$, to the Abelian rank-$0$ theory $\mathcal{T}_{\ell,n}$ of Creutzig, Garner and Kim. The fundamental is pseudo-real, so the level is half-integral, there is no non-Abelian flavor symmetry, and the supersymmetry is enhanced to $\mathcal{N}=4$. Both sides have vanishing Higgs and Coulomb branches. We match the counting of supersymmetric vacua under real mass deformations for all $(n,\ell)$, and the supersymmetric indices in a number of cases. We then study in detail the case $\ell=1$, the minimal $USp(2n)$ Chern-Simons theory at level $n+\frac32$, whose dual is a $U(1)^{n}$ theory with a tadpole level matrix and a bare monopole superpotential.
Read the full abstract

Combinatorics

Monotone Diameters of Lattice Polytopes

An influential 1989 result of Naddef shows that the diameters of $0/1$-polytopes are at most their dimension. This was extended shortly after by Kleinschmidt and Onn to any lattice polytope in $[0,k]^{d}$, where they showed a bound of at most $dk$. Naddef's argument easily extends to the monotone setting motivated by the simplex method, where one requires paths to increase with respect to a linear objective function. However, the Kleinschmidt-Onn argument does not. In fact, no argument in the 30 years since has managed to fill that gap. Prior to this work, it remained open whether the monotone diameter is bounded by a polynomial in $d$ and $k$ with no lower bounds suggesting any separation between the worst-case diameter and worst-case monotone diameter. Linear upper bounds hold for $k=1$ and $k=2$. However, we exhibit a sharp threshold for this question at $k = 3$ by constructing for each $d \geq 1$ a lattice polytope in $[0,3]^{6d}$ with monotone diameter at least $2^{d}-1$. In particular, the polynomial bound does not hold. Furthermore, we show that Naddef's result does not extend to the unbounded setting by exhibiting a family of unbounded polyhedra with $0/1$-vertices and diameter exponential in their dimension.
Read the full abstract

Computation and Language

Combating Instruction Conflict via Energy-Driven Latent Conflict Detection

Large Language Models (LLMs) are increasingly deployed with hierarchical instructions, yet they remain vulnerable to conflicts in which user directives override system-level constraints. Existing defense mechanisms predominantly focus on static input inspection and therefore fail to detect Response Drift, a phenomenon in which the model's final response violates system-level constraints despite seemingly compliant inputs. To bridge this gap, we introduce ELCD, a response-level latent conflict detector for post-generation, pre-delivery verification. Given the full generated output, ELCD constructs a composite hidden-state representation by concatenating the final-token embedding with the mean-pooled response embedding. It then optimizes a pairwise margin ranking objective to separate compliant and drifting responses in latent space. Extensive experiments across five mainstream LLMs ranging from 1.5B to 14B parameters demonstrate that ELCD significantly outperforms competitive baselines. Notably, it improves the PR-AUC on Llama-2-7B by approximately 30 percentage points and reduces the False Positive Rate at 95% TPR (FPR95) on Mistral-7B to 2.67%. These results suggest that ELCD provides a promising approach for latent instruction-conflict detection in open-weight or self-hosted LLM deployments.
Read the full abstract

All authors Mingyu Ma, Yuxin Wu, Jingbo Wang, Tianxiao Huang, Leixin Sun, Xiaochuan Shi

Probability

Polynomial mixing for the 3D damped cubic nonlinear Schrödinger equation with degenerate noise

We prove polynomial mixing for the defocusing damped cubic stochastic nonlinear Schrödinger equation on the three-dimensional torus under saturating smooth finite rank Brownian forcing. The mixing rate is measured in the $p$-Wasserstein metric induced by the $H^1$ distance for every $1\le p<\infty$. We also obtain sharp geometric characterizations of saturation. The proof is based on a polynomial mixing criterion built on a stable–compact decomposition of the exact solution differences with polynomial moment control of the logarithmic path amplification. Dense Malliavin range allows the compact defect to be compensated by finite-dimensional Cameron–Martin shifts, producing a block multiplier with negative mean logarithm. A logarithmic transportation gauge, combined with a renewal–reset coupling scheme, then yields mixing at every prescribed polynomial order in the weaker $L^2$ distance. A stationary regularity gain to $H^{2-}$ then enables us to upgrade the convergence to $H^1$.
Read the full abstract

High Energy Physics (Phenomenology)

The transverse-spin asymmetry $A^{\sin\phi_{S_\Lambda}}_{UUT}$ in $\Lambda$ production in SIDIS within the collinear framework

We investigate the transverse-spin asymmetry $A^{\sin\phi_{S_\Lambda}}_{UUT}$ for transversely polarized $\Lambda$ production in semi-inclusive deep inelastic scattering with an unpolarized electron beam and an unpolarized nucleon target. The spin-dependent asymmetry is sensitive to the twist-3 T-odd quark-gluon-quark fragmentation function $\tilde{D}_{T}$, which describes the correlation between an unpolarized fragmenting quark and the transverse polarization of the produced hadron. We calculate $\tilde{D}_{T}$ of the $\Lambda$ hyperon within a diquark model and investigate its contribution to the asymmetry. We determine the model parameters by fitting the fragmentation functions $D_1$ and $G_1$ to the DSV parametrization. Using the numerical results of the model and the available parametrization of the unpolarized parton distribution function $f_1$, we predict the asymmetry in the electroproduction of the $\Lambda$ hyperon in the kinematic regions of the EIC and the EicC. We also examine the impact of including the strange-quark contribution on the transverse-spin asymmetry. In our phenomenological analysis, we include the QCD evolution of the parton distribution function $f_1$ and the fragmentation function $\tilde{D}_{T}$. The results show that the asymmetry is sizable and can be significantly modified by the strange-quark contribution.
Read the full abstract

Analysis of PDEs

An Inverse Random Source Problem for the Moore-Gibson-Thompson Equation Driven by Fractional Brownian Motion

In this paper, we consider an inverse random source problem for the stochastic Moore-Gibson-Thompson equation driven by fractional Brownian motion with Hurst index $H \in (0, 1)$ of the form $f_1(x)g_1(t)\dot{B}^H(t)+f_2(x)g_2(t)$. Given the random source, existence and uniqueness of mild solutions are verified. For the inverse problem, the uniqueness of recovering the strength $f_i(x)$ if the time functions $g_i$ are known and $g_i(t)$ if the spatial functions $f_i$ are known when $H \in(0,1)$ from the boundary flux on a special nonempty open subset is proved.
Read the full abstract

Machine Learning

SUN: Reaching for Novelty in Reinforcement Learning

Exploration in reinforcement learning (RL) remains a fundamental challenge. Recent goal-conditioned RL strategies (which select goals to encourage broader state coverage) have shown promising results, but none scores a goal by novelty and reachability jointly: the two signals are traded off by hand, applied in sequence, or one is neglected outright. In this paper, we introduce a reachability-aware goal-selection framework that explicitly integrates these two aspects, and that can be seamlessly incorporated into any off-policy RL algorithm. To this aim, we propose SUccessor-to-Novelty (SUN), an indicator derived from successor value functions to identify goals that are both novel and reachable. We prove that SUN is the value of a count-scaled goal-reaching reward, bounds short-horizon hitting probabilities, and rejects unreachable goals. We further present an adaptive goal-selection strategy that leverages these properties, along with an accurate yet lightweight pseudocount to avoid the overhead of classic methods. We back up all our claims with thorough benchmarks: SUN consistently outperforms state-of-the-art methods in standard and novel environments with unreachable or hard-to-reach states, irreversible transitions, obstacles, mazes, and unbounded spaces. Code at https://github.com/sparisi/rl-anvil
Read the full abstract

All authors Wenyan Yang, Arsenii Mustafin, Dominik Baumann, Joni Pajarinen, Simone Parisi

General Relativity and Quantum Cosmology

Black holes with torsion hair in cubic Holst-type Poincaré gauge gravity: from singular to regular geometries

Motivated by the singularity theorems of Poincaré Gauge (PG) theory, we investigate extensions of the Holst quadratic model by introducing cubic order invariants constructed from the curvature and torsion tensors into the gravitational action. Such models are characterised by a kinetic structure that is governed by a pseudoscalar mode, whereas the remaining irreducible modes of torsion contribute through nonlinear interactions that can have important implications for the space-time geometry. In particular, in line with other well-known models of PG theory, the Birkhoff theorem does not hold in general, allowing for new exact static and spherically symmetric black hole solutions with dynamical torsion. Across the different torsion sectors, corresponding to the irreducible modes and parity components of the torsion field involved in the analysis, we find both singular and regular configurations. Among the singular solutions, we obtain Kiselev-like and Boulware-Deser-like geometries, as well as new geometries with distinct algebraic and Lambert $W$ metric corrections. In addition, we find regular black holes with both primary and secondary torsion hair, which evade the singularity theorems through violations of the causal convergence conditions induced by the nonlinear torsion interactions. Therefore, we show that Holst-type PG models can support a rich variety of black hole geometries, while providing explicit mechanisms for evading the singularity theorems of PG theory.
Read the full abstract

Analysis of PDEs

Modulated Gibbs Measure and Mean-Field Limit for 3D Vlasov–Poisson–Fokker–Planck Equations

We introduce a modulated Gibbs measure for the usual tensorized initial data for stochastic Newton's systems with singular repulsive interactions. Using the uniform-in-$N$ partition-function estimates of Wang–Zhao \cite{wang2026uniform}, we show that the modulated Gibbs measure and its tensorized reference are $O(N^{-1})$-close in normalized relative entropy. A stability argument then transfers propagation of chaos from modulated Gibbs measures to tensorized data and other initial laws at the same entropy scale. Combined with the weighted BBGKY estimates of Bresch–Jabin–Soler \cite{bresch2025new}, this yields the first cutoff-free mean-field limit and propagation of chaos for the three-dimensional Vlasov–Poisson–Fokker–Planck (VPFP) equation on an $N$-independent short-time interval.
Read the full abstract

Quantum Physics

Causal-Class Hierarchies in Coherence-Constrained Channel Transformation

Higher-order quantum transformations allow multiple channel uses to be combined through different causal architectures, from parallel and fixed-order sequential networks to general higher-order processes. Whether this causal freedom improves channel transformation when the higher-order operation is also constrained by a resource theory remains largely unexplored. We study this question in the dynamical resource theory of coherence using a unified semidefinite-programming framework. For two qubit amplitude-damping channels and the identity target, we prove a strict causal hierarchy at every nontrivial damping strength under both maximally incoherent superchannels (MISC) and dephasing-covariant incoherent superchannels (DISC). In contrast, mixed-Pauli channels admit a common teleportation simulation that transfers the channel dependence to Bell-diagonal program states prepared in parallel. The remaining processing can then be absorbed into a single quantum operational, so parallel, fixed-order sequential, and general higher-order strategies achieve the same optimal error for any target. These results identify free program-state parallelisation as a structural obstruction to causal enhancement.
Read the full abstract

All authors Lin Zhu, Benchi Zhao, Xuanqiang Zhao, Ranyiliu Chen, Xin Wang, Shenggen Zheng

Robotics

CASD: Chunk-Aligned Semantic Distillation for Multi-StageRobot Manipulation

An action chunk can span several stages of a manipulation task, yet a label for its first step describes only the current stage. We introduce Chunk-Aligned Semantic Distillation (CASD), which derives semantic targets for entire action chunks. An offline vision–language model segments demonstrations into described stages. Their occupancy within each action chunk determines a weighted semantic target, including transitions between stages. A CASD generator learns to predict this target from the current observation, robot state, and task instruction. We then freeze the generator and train a policy conditioned on its predictions. The semantic branch runs once per policy query, without online VLM calls or reasoning-trace decoding. Teacher matching on annotated LIBERO training episodes is above chance for both single-stage and boundary-crossing chunks. We evaluate three Fast-WAM variants and a DreamZero integration across four benchmarks, including distribution shifts on LIBERO-Plus. Compared with published references, IDM+CASD reaches 98.9% versus 98.0% average success on LIBERO, while Uncond falls below its reference. Joint+CASD reaches 93.0% versus 90.6% on RoboTwin 2.0, and DreamZero+CASD reaches a 47.9% four-category MolmoSpaces manipulation average versus 40.7%. Performance varies across backbone integrations.
Read the full abstract

Computation and Language

Navigating the digital spectrum: Assessing political bias, stability, and downstream fairness in Large Language Models

Large Language Models are increasingly deployed as information intermediaries, yet measuring their political behavior remains fragile because questionnaire results mix model dispositions with measurement artifacts and response-elicitation biases. We introduce a robust Political Compass Test evaluation framework that samples 300 configurations across an eight-dimensional perturbation space varying language, framing, instructions, answer format, option order, and persona wording. We evaluate eight Gemma 3 and Qwen 3 models across 14 languages and three quantization levels, obtaining design-averaged political coordinates with quantified uncertainty. Most models lean Libertarian-Left on average, but instruction phrasing, language, and answer format significantly affect recovered coordinates. Cross-lingual differences primarily reflect coordinate drift rather than distinct cultural reasoning. Reverse-engineering the test also exposes axis-weighting imbalances and the collapse of degenerate responses toward the center, so near-origin estimates for the smallest models can reflect weak signal rather than centrism. Free-text reasoning and chat-then-classify elicitation alter recovered coordinates, and larger models show clearer persona separation, with a specific failure of the Authoritarian-Left persona to move most models in the intended social direction. In downstream tasks, persona effects are modest relative to model size and target group for hate-speech detection, while base and centrist prompts give the highest agreement for topic-level sentiment. Political role prompting therefore has measurable but task- and dataset-specific downstream effects.
Read the full abstract

All authors Luka Debevc, Nishan Chatterjee, Antoine Doucet, Senja Pollak, Matej Martinc

Computer Vision

From Where to How: Continuous 4D Interaction Forecasting from Egocentric Video

Egocentric 4D interaction forecasting aims to anticipate both where future interactions will occur in 3D and how the human body will move to realize them, providing an important capability for assistive robotics and human-computer interaction. Existing methods struggle to translate semantic understanding into precise continuous 3D localization and to balance motion diversity with structural consistency in pose forecasting. More fundamentally, these tasks are often modeled separately, leaving the continuous geometric and temporal correspondence between interaction locations and body motion insufficiently captured. To address these challenges, we introduce Coherent4D, a large-scale egocentric dataset for continuous 4D interaction forecasting, comprising approximately 233K samples across three domains. Each sample pairs a sequence of future 3D interaction locations with corresponding full-body poses, aligned in time and expressed in a shared coordinate system. We also provide evaluation metrics in continuous space. Building on this formulation, we propose HIGFlow, a Hand Interaction Guided Residual Flow framework that models forecasting as a cascaded where-to-how process. HIGFlow first forecasts continuous future interaction locations by combining semantic grounding with short-horizon visual dynamics, and then uses the predicted location sequence to condition a deterministic motion anchor and residual Flow Matching for diverse yet structurally consistent full-body motion forecasting. Extensive experiments across all three domains demonstrate consistent improvements over representative baselines on both location and pose forecasting, while ablations validate the contributions of the proposed components. The project page is available at https://corrineqiu.github.io/from-where-to-how/.
Read the full abstract

All authors Qiaohui Chu, Haoyu Zhang, Meng Liu, Haoxiang Shi, Dongmei Jiang, Liqiang Nie

Quantum Physics

Finite-size Entanglement Certification via Third-Order Local Randomized Measurements

Certifying entanglement in high-dimensional systems usually requires full state tomography, whose cost grows rapidly with the system dimension. Local randomized measurements offer a scalable alternative, but existing tests based on second-order correlations access only limited information about the state. Here, we derive a finite-size entanglement certificate that extends local randomized measurements to third order. The additional third-order information reveals entanglement that remains undetected at second order, while a dimension-independent concentration bound provides rigorous control of finite-sample errors. Our result opens a practical route to extracting stronger entanglement information from experimental platforms without the dimension-dependent overhead of state tomography.
Read the full abstract

All authors Giovanni Scala, Corrado Cosimo Mattiacci, Dorota Pietryka, Gniewomir Sarbicki

Machine Learning

Suan: Rectifying Direct Preference Safety Alignment in Large Language Models

Integrating robust safety guardrails into Large Language Models (LLMs) is essential for delivering helpful yet harmless responses. While proprietary systems exhibit reliable safety controls, their underlying methodologies and trade-offs remain largely undisclosed. Achieving comparable security in open-weight models remains a persistent challenge, as post-trained variants frequently suffer from over-refusal and degraded general quality. To overcome these drawbacks, we introduce Suan, a novel preference optimization algorithm. Unlike existing methods, we formulate the optimization objective directly at the gradient level, bypassing the standard variational derivation. As a result, we obtain more interpretable and robust training dynamics. Extensive evaluations across a diverse suite of competitive baselines and benchmarks demonstrate that Suan achieves superior safety alignment while fully preserving response utility.
Read the full abstract

Algebraic Geometry

Singularities of totally invariant hypersurfaces of endomorphisms

Let $f\colon X\to X$ be a polarised endomorphism of a complex projective manifold, and let $D\subset X$ be a reduced totally invariant hypersurface. We prove that every irreducible component of the singular locus of $D$ has codimension one in $D$. In particular, if $D$ is normal, then it is smooth. As an application we prove the linearity of totally invariant divisors for endomorphisms of $\mathbb P^4$.
Read the full abstract

Materials Science

Pulse-Burst Excitation Reveals Time-Dose Reciprocity Breakdown in Mixed-Halide Perovskites

Time-dose reciprocity, commonly associated with the Bunsen-Roscoe law, states that the response of a photosensitive system depends only on the total exposure dose, regardless of how that energy is delivered over time. Light-sensitive processes in mixed-halide perovskites, such as photoinduced halide segregation, often exhibit threshold-like behavior that may violate this principle and enable material-state control by photon timing. We test this using pulse-burst excitation, which introduces an additional temporal control dimension beyond conventional parameters such as pulse fluence, repetition rate, and average power. By redistributing the same photon dose over microsecond-to-millisecond timescales, we create distinct nonequilibrium excitation conditions and show that mixed-halide perovskites can evolve into different metastable states, revealing a breakdown of time-dose reciprocity in the combined processes of halide segregation and remixing. This additional temporal degree of freedom not only enables control of the material state but also provides a new experimental framework for disentangling the competing processes underlying photoinduced halide redistribution. Our findings establish photon timing as a control parameter for perovskite photochemistry and open additional opportunities for optical memory and neuromorphic photonic applications.
Read the full abstract

All authors Alexandr Marunchenko, Shivam Singh, Daniel Lizotte, Bhaskar De, Yana Vaynzof, Ivan G. Scheblykin

Signal Processing

CrossRAFT: Cross-Domain Complex-Valued Feature Extraction for Ultrasound Motion Estimation

Accurate multi-dimensional motion estimation is fundamental to broad biomedical ultrasound applications, primarily performed using real-valued radio-frequency (RF), analytic, and in-phase and quadrature (IQ) signals. Real-valued RF and analytic signals offer high subsample accuracy but are computationally intensive. Regardless of the signal domain used, traditional algorithms suffer from interframe decorrelation, resolution tradeoffs, and a lack of a carrier and low sampling in the lateral direction. Emerging deep learning (DL) models demonstrate improved performance and faster inference. However, existing architectures are real-valued and disrupt inherent phase coupling crucial for accurate motion estimation by treating complex components as independent channels. To address this, we propose CrossRAFT, an end-to-end trainable network designed for subsample motion estimation directly from complex-valued ultrasound signals. Built on the RAFT architecture, CrossRAFT features complex-valued encoders and a custom correlation module to explicitly extract and use phase information. We evaluated CrossRAFT across real RF, analytic, and IQ signal presentations on simulated phantoms, synthetic echocardiographic, and private in vivo echocardiographic datasets. Experimental results show that CrossRAFT estimates displacements accurately under complex tissue motion and in the challenging lateral direction. Compared to real RF-based RAFT, analytic-based CrossRAFT reduces vertical and horizontal displacement errors by 69% and 51% on the synthetic echocardiographic data, respectively. These findings demonstrate that leveraging phase information via complex-valued neural networks provides a powerful framework for ultrasound motion estimation, showing potential for clinical elastography and functional imaging.
Read the full abstract

All authors Yang Leng, Yuchen Tang, Kai-Hang Yiu, Yik-Chung Wu, Wei-Ning Lee

Numerical Analysis

A Fully Dual Domain Decomposition Method for Extreme Learning Machines Using GenEO Spectral Coarse Spaces

Extreme learning machines (ELMs) are a class of machine learning algorithms that randomly intialize and subsequently fix the hidden layer parameters, and solve for the last layer coefficients using a least squares method. ELMs have been successfully applied to the solution of various partial differential equations, but the cost of solving large least squares problems limits their use in large scale problems or when high accuracy is desired. To address this limitation, domain decomposition methods (DDMs) have been developed for ELMs, reducing training times via parallel computation. Yet these methods, even those incorporating a coarse space, have struggled with scalability. This paper introduces a novel nonoverlapping DDM for ELMs that treats and enforces continuity and flux conditions equally. Additive Schwarz preconditioners and GenEO (Generalized Eigenvalues problems in the Overlaps) coarse spaces are constructed to accelerate the method. Numerical experiments show the new method outperforms previous DDMs for ELMs in both iteration count and accuracy, especially for the Stokes and wave equations. Furthermore, weak scalability has been observed in up to 4,096 subdomains when employing GenEO coarse spaces.
Read the full abstract

Signal Processing

Efficient Deep Learning Adaptation for Cross-Environment RSS-Based Indoor Localization

Received signal strength (RSS)-based indoor localization has attracted increasing attention due to its low cost and compatibility with existing wireless infrastructures. However, RSS measurements are highly sensitive to environmental variations, making it challenging for deep learning-based localization models to generalize across different physical configurations. Driven by these challenges, this paper proposes a scenario-adaptive RSS localization framework based on backbone reuse and efficient adaptation. The proposed architecture consists of a lightweight extractor and a shared backbone, where the extractor projects heterogeneous RSS inputs into a unified feature space and the backbone captures transferable localization knowledge. During adaptation, the backbone trained from the source dataset is reused, while the extractor is optimized to adapt to the new dataset configuration. A short optimization stage is further introduced to slightly refine the backbone with a lower learning rate. Experimental results on four datasets demonstrate that the proposed training strategy enables faster convergence and improved adaptation compared with training from scratch. In addition, the proposed framework reduces training time under different dataset configurations, verifying its effectiveness and efficiency for adaptive RSS-based indoor localization.
Read the full abstract

Nuclear Theory

The Interplay of Symmetry Energy Uncertainties, Nonlinear $\sigma-\delta$ Coupling, and Dark Matter in Neutron Star Macroscopic Properties

The poorly constrained density dependence of the nuclear symmetry energy introduces significant uncertainties in the equation of state (EOS) of dense nuclear matter and, consequently, in neutron-star properties. We systematically investigate how uncertainties in the symmetry energy $E_{\rm sym}$ and its slope $L$ at saturation density $n_0$ affect NS properties within the relativistic mean-field (RMF) framework, including the effects of nonlinear $\sigma$-$\delta$ coupling and possible admixture of dark matter(DM) . For fixed $E_{\rm sym}(n_0)$ and $L(n_0)$, we find that the $\sigma$-$\delta$ coupling with $g_{\sigma\delta}=-0.004$ induces an abnormal softening of the EOS, which simultaneously increasing the maximum NS mass and reducing the stellar radius and tidal deformability. A similar behavior is found in the presence of DM with Fermi momentum $k_F^{\rm DM}=50~\mathrm{MeV}$ at $E_{\rm sym}(n_0)=36~\mathrm{MeV}$. In this case, the RMF EOS satisfies the tidal-deformability constraint from GW170817. We further find that the surface curvature of NSs is strongly correlated with the stiffness of $E_{\rm sym}$, with softer symmetry energy corresponding to larger surface curvature. Our results also indicate that the behavior of $E_{\rm sym}$ around $n_0$ is mainly governed by isoscalar rather than isovector parameters.
Read the full abstract

All authors Zhaohui Feng, Xiaoxuan Zhai, Shuangxuan Chen, Defu Hou

Computer Vision

SynthRCT: Scalable Conditional Deformation Synthesis for Synthetic Repeat CT Generation

In proton therapy, plans are typically optimized on a single planning CT, making robustness evaluation essential under anatomical changes. However, current scenarios often rely on simplified perturbations that poorly capture complex, patient-specific variability. We propose SynthRCT, a scalable conditional generative framework for 3D anatomical deformation synthesis. Based on a conditional variational autoencoder, SynthRCT learns a latent deformation space and decodes sampled latent codes into local stationary velocity fields conditioned on an input anatomy. Local fields are assembled into coherent full-volume transformations, enabling memory-scalable generation for large field-of-view CT data. We validate the approach on respiratory 4DCT data with multiple breathing-phase anatomies per subject. SynthRCT enables patient-specific sampling of plausible anatomical transformations beyond predefined robustness scenarios. Code available at: https://github.com/TomasGuija/SynthRCT.
Read the full abstract

Robotics

MFVINS: Multiple Fisheye Camera-Based Visual Inertial System

A simultaneous localization and mapping (SLAM) method using a monocular camera and a low-cost inertial measurement unit (IMU) sensor is an effective way to fulfill a low-cost sensor configuration. Using this sensor configuration, visual-inertial system (VINS) focuses on fusing data from a camera and an IMU sensor to estimate the six degrees-of-freedom (DOF) of the sensor pose. Typically, VINS uses only a single camera as visual input, which lead to problems such as error accumulation due to occlusion, various illumination, and textureless environments. In this paper, we propose a new multiple fisheye camera-based visual-inertial system called MFVINS. We present an IMU-aided FAST feature tracker for multiple cameras that enables efficient extraction and robust matching of local features. Then, the proposed method filters out outliers caused by fisheye distortion on the normalized image plane. Subsequently, a new reprojection error with physical validity constraints is proposed for bundle adjustment using learning-based depth estimation. The proposed method is applied to various scenarios, and its effectiveness is demonstrated by comparing previous VINS methods. In particular, MFVINS is implemented in real-time process to leverage the advantages of using multiple cameras – robustness against occlusion and textureless regions – while reducing the computational burden.
Read the full abstract

All authors Eunseong Jang, YuJin Chung, Sang Jun Lee, Jihyun Yoon, HyungGi Jo

Discrete Mathematics

Generalized Graph Search Trees

Graph search algorithms and their corresponding graph search trees are commonly used in algorithmic graph theory. In recent years, the recognition problem of these graph search trees has received significant attention. So far, the research has focused on two types of search trees: first-in trees that behave like BFS-trees and last-in trees that behave like DFS-trees. The search tree paradigms differ from each other by the parent a vertex is connected to. In first-in trees, it is the first visited neighbor, while in last-in trees it is the last neighbor visited before that vertex. Here, we will generalize these concepts of graph search trees by allowing every preceding neighbor of a vertex to be the parent. We study the complexity of the recognition problem of these generalized graph search trees. We present NP-completeness proofs for most searches. We also show that the problem is trivial for Generic Search and polynomial-time solvable for several searches on bipartite graphs and chordal graphs. We also study the question how fixing the start vertex influences the complexity of the problem.
Read the full abstract

Combinatorics

Latin Squares with Few Transversals

Let $t(n)$ denote the minimum number of transversals in a Latin square of odd order $n$. Improving upon a recent bound of Dai, Divoux and Kelly, we prove that for every $n$ such that $n \equiv 3 \pmod 6$, \[ t(n) \leq \left( \left(1+o(1)\right) \frac{2n}{3e^2}\right)^n . \] Our proof is based on a family of $3 \times 3$ block Latin squares whose transversals are constrained to either lie entirely in the diagonal blocks or avoid them altogether.
Read the full abstract

Cryptography and Security

Navigating the Latent Manifold: Proactive Concept Drift Adaptation for Resilient NIDS

Network intrusion detection systems (NIDS) are critical for cybersecurity, safeguarding services and data from potential attacks. However, existing AI-based NIDS often assume static data distributions and fail to handle concept drift, leading to degraded performance and increased false positives in dynamic network environments. To address this issue, we propose DriftXpert, a novel NIDS for drift-adaptive detection. Specifically, we propose a decoupled two-stage offline adaptive framework. In Phase 1, we introduce an unsupervised anomaly metric based on latent manifold deviation. By performing outlier analysis within the latent space, the framework achieves high-sensitivity detection of network traffic concept drift. In Phase 2, to mitigate catastrophic forgetting under non-stationary distributions, we design a representation consistency alignment strategy. This strategy constrains the feature mapping between the legacy model and the drifted distribution, ensuring the model captures emerging attack characteristics while retaining discriminative power over known patterns. Furthermore, we incorporate cross-epoch neuron weight aggregation and selective freezing mechanisms to enable fine-grained knowledge transfer in the parameter space, effectively balancing model plasticity and stability. Extensive experiments on public datasets demonstrate that DriftXpert effectively adapts to drifted data without catastrophic forgetting. Furthermore, real-world evaluations on enterprise network further confirm its robustness and practical applicability, contributing to improved security protection for millions of users.
Read the full abstract

All authors Chao Zha, Zifeng Kang, Tian Liu, Dakun Shen, Ruyun Zhang

Machine Learning

Leveraging contextual events on structure-aware next activity prediction

Predictive process monitoring aims at forecasting various aspects of running processes. Among the different tasks, next activity prediction represents the most extensively investigated. However, only a limited number of existing approaches explicitly encode contextual information, i.e., the environmental conditions in which the process is executed, typically modeled through event log attributes or aggregated measures. In this paper, an approach based on the concept of Instance Graphs is introduced. To incorporate contextual process instances, several encoding strategies are proposed and evaluated by measuring their impact on prediction performance. For each encoding strategy, a set of prefix-Instance Graphs is generated and subsequently provided as input to a Graph Neural Network for the classification task. The proposed approach is evaluated on multiple real-world event logs, and the experimental results demonstrate that incorporating contextual process instances benefits prediction performance.
Read the full abstract

Earth and Planetary Astrophysics

XUE. ProDiMo models of internally and externally irradiated planet-forming disks around 0.3-4.0 solar mass stars (The IRIS project I)

Most stars and planets form in massive star-forming regions, where disks are exposed to external far-ultraviolet (FUV) radiation from nearby O- and B-type stars. The combined effects of stellar irradiation and external FUV fields on terrestrial planet-forming regions (< 10 au) across stellar masses remain unclear. We investigate how internal UV and X-ray irradiation and external FUV fields affect mid-infrared (mid-IR) gas emission and the atmospheric carbon-to-oxygen (C/O) ratios inferred in T Tauri and Herbig Ae/Be disks. We compute disk structures with ProDiMo and synthetic spectra with FLiTs, convolved to a representative JWST/MIRI-MRS resolution (R$\sim$2680). (1) We present the Internal and external irRadIation of diSks (IRIS) grid: four model sets spanning stellar masses of 0.3-4.0 solar masses, including stellar X-ray flares and an external FUV field of 1e4 G0 (Habing units). (2) We predict increasing flux densities with stellar mass for key atomic and molecular mid-IR tracers. (3) External FUV irradiation enhances CH3+ and H2 emission, whereas FUV-induced disk truncation yields inner disk chemistry resembling that of disks irradiated only by their host stars. (4) Mid-IR H2O, CO2, and C2H2 line ratios imply carbon-rich compositions (C/O$\sim$1-10) in the warm emitting layers of T Tauri and Herbig Ae/Be disks, primarily reflecting stellar irradiation, with little sensitivity to external FUV irradiation. The IRIS grid provides a flexible framework for interpreting JWST and future Extremely Large Telescope (ELT) infrared disk observations across a broad range of stellar properties and irradiation conditions. Future models should include FUV-driven photoevaporative winds, X-ray radiative transfer, and time-dependent X-ray irradiation.
Read the full abstract

All authors Jenny Frediani, Konstantin V. Getman, Peter Woitke, Bayron Portilla-Revelo, Eric D. Feigelson, Christian Rab, Arjan Bik, María Claudia Ramírez-Tannus, Rens Waters, Thomas Henning, Inga Kamp, Germán Chaparro, Sebastián Hernández, Simón Rodríguez, Pablo Cuartas-Restrepo, Andrew Winter, Thomas J. Haworth, Thomas Preibisch, Sierk E. van Terwisga, Veronica Roccatagliata, Alexis Brandeker