arXiv subjects

Jingbo Sun

Publications and source records attributed to Jingbo Sun.

At least 19 recordsLinked to original sources

arXiv2026

The Intermediate-Mass Black Hole Reverberation Mapping Project: Scientific Overview and Sample Characteristics

Recent discoveries with the James Webb Space Telescope of massive black holes at high redshift have highlighted fundamental questions about black hole seed formation and the coevolution of black holes with their host galaxies. Because the initial seed population cannot yet be observed directly, nearby intermediate-mass black holes provide a complementary fossil record of black hole formation and early growth. Motivated by this opportunity, we present the Intermediate-Mass Black Hole Reverberation Mapping (IMBH-RM) project and construct a homogeneous Sloan Digital Sky Survey sample of active broad-line IMBHs by uniformly reanalyzing literature candidates with consistent spectral decomposition and black hole mass estimation. Our sample contains 192 reliable IMBH candidates at $z\lesssim0.3$ with $\log(M_{\rm BH}/M_\odot)<6$, including four particularly compelling sources with $\log(M_{\rm BH}/M_\odot)<5$. The primary goal of IMBH-RM is to obtain reliable black hole masses from direct measurements and characteristic sizes of the broad-line region and accretion disk for a carefully selected subsample. These measurements will provide robust low-mass anchors for calibrating single-epoch black hole mass estimates and extending black hole--galaxy scaling relations into the IMBH regime. By building a statistically meaningful reverberation-mapped sample spanning $10^4-10^6\,M_\odot$, we aim to constrain the local IMBH mass distribution and place observational constraints on competing black hole seed formation scenarios. The future Multi-Channel Imager aboard the Chinese Space-station Survey Telescope provides a particularly promising platform for achieving these goals.

astro-ph.GA↗

arXiv2026

Constraints on the Intranight Optical Variability of Intermediate-Mass Black Hole Candidates

Intermediate-mass black holes (IMBHs) provide a unique regime for studying accretion variability at the low-mass end of the black hole population, yet their intranight optical variability (INOV) remains poorly constrained. We present a systematic investigation of INOV in an optically selected sample of IMBH candidates using high-cadence observations from the Zwicky Transient Facility (ZTF). From a parent sample of 1,447 broad H$\alpha$-selected candidates, we identify 64 IMBH candidates (median $f_{\mathrm{AGN}}\sim0.06$) with 163 intranight monitoring sessions. Apparent INOV signals identified by conventional ZTF PSF-fit photometry are largely associated with seeing-dependent changes in the relative contributions of compact nuclear and extended host components, which can mimic intrinsic short-timescale variability. In contrast, no robust INOV is detected with difference-image analysis. An ensemble structure function spanning $\Delta t\sim0.003$--$1600$ days reveals long-term variability in a small subsample of sources, whereas intrinsic variability remains unresolved at intranight timescales. Monte Carlo simulations further show that ZTF-like single-night monitoring has a low INOV recovery probability ($\sim1.2%$) for the variability amplitudes inferred from the long-term analysis. The recovery probability is primarily controlled by source brightness, AGN contribution, intrinsic variability amplitude, and photometric precision. These results demonstrate that the absence of detected INOV does not imply the absence of rapid accretion variability, but can reflect the limited detectability of low-amplitude signals under current observing capabilities. Our findings highlight the importance of robust photometric methodologies for future high-cadence variability studies of low-mass accreting black holes.

astro-ph.GA↗

arXiv2026

Early Near-Infrared Excess and Rapid Disk-Corona Evolution in the Tidal Disruption Event 2024aepd

We present multi-wavelength observations of the tidal disruption event (TDE) 2024aepd, spanning primarily the first $\sim$300 days after discovery. The X-ray spectrum is initially dominated by a thermal disk component accompanied by a hard excess. From $\sim$178 days onward, the spectrum becomes power-law dominated and subsequently hardens, indicating the rapid emergence and strengthening of a hot corona. A prominent near-infrared (NIR) excess is detected as early as $\sim40$ days. Its nearly flat power-law spectrum strongly deviates from the Rayleigh-Jeans tail of the UV-optical blackbody. Although a conventional dust-echo origin cannot be completely ruled out, free-free emission from a reprocessing photospheric envelope provides a more plausible explanation. Moreover, the UV-optical-to-NIR break shifts to higher frequencies as the density-profile index remains nearly constant, implying evolving reprocessing conditions within a broadly unchanged density structure. Together with AT2019azh and TDE 2025abcr, TDE 2024aepd is the third TDE reported to exhibit an early-time NIR excess. A larger sample with early-time NIR coverage is needed to determine whether such excesses are common among TDEs.

astro-ph.HE↗

arXiv2026

Randomized neural operator for parametric PDEs with fast training and conformal uncertainty quantification

Repeatedly solving parametric PDEs is essential for uncertainty quantification, design optimization and inverse problems, but conventional neural operators require expensive non-convex training. We introduce PCA--RaNN, a randomized latent neural operator that combines PCA-based dimensionality reduction with fixed random features and a closed-form least-squares readout. It recasts latent operator learning as fixed-feature linear regression, reducing training time by one to three orders of magnitude across benchmarks while maintaining competitive accuracy. We introduce an energy-matched scaling rule and a lightweight two-parameter BFGS refinement to correct suboptimal feature scales. Ensemble averaging reduces predictive variance. On Burgers, Darcy, Navier--Stokes and backward heat equation benchmarks, PCA--RaNN provides a favorable speed--accuracy trade-off against operator-learning baselines. The ensemble supports split-conformal prediction intervals, and the linear readout enables rapid online adaptation via recursive least squares without retraining hidden features. This provides an efficient, uncertainty-aware surrogate for many-query scientific workflows.

cs.LG↗

arXiv2026

Beyond Query Memorization: Large Language Model Routing with Query Decomposition and Historical Matching

Optimizing the trade-off among predictive performance and computational cost is a central focus in the deployment of Large Language Models (LLMs). Current routing methods primarily rely on direct mapping from queries to models based on surface-level features, making them susceptible to the memorization trap and leading to poor generalizability on out-of-distribution (OOD) data. In this paper, we propose DecoR, a novel routing framework that recasts the routing task as a matching process of sifting similar queries from historical logs, effectively mitigating the memorization trap. To enhance matching accuracy, we introduce a query capability deconstruction method that decouples linguistic surface forms from task-intrinsic requirements, directing matching toward capability dimensions to ground decisions in essential task attributes. Furthermore, we develop CodaSet, a comprehensive benchmark for assessing routing generalization, where experimental results demonstrate that DecoR maintains superior accuracy while substantially lowering inference costs across both in-distribution and OOD settings. All the codes and data are available at https://github.com/lvbotenbest/DecoR.

cs.AI↗

arXiv2026

AutoSearch: Adaptive Search Depth for Efficient Agentic RAG via Reinforcement Learning

Agentic retrieval-augmented generation (RAG) systems enable large language models (LLMs) to solve complex tasks through multi-step interaction with external retrieval tools. However, such multi-step interaction often involves redundant search steps, incurring substantial computational cost and latency. Prior work limits search depth (i.e., the number of search steps) to reduce cost, but this often leads to underexploration of complex questions. To address this, we first investigate how search depth affects accuracy and find a minimal sufficient search depth that defines an accuracy-efficiency trade-off, jointly determined by question complexity and the agent's capability. Furthermore, we propose AutoSearch, a reinforcement learning (RL) framework that evaluates each search step via self-generated intermediate answers. By a self-answering mechanism, AutoSearch identifies the minimal sufficient search depth and promotes efficient search by rewarding its attainment while penalizing over-searching. In addition, reward mechanisms are introduced to stabilize search behavior and improve answer quality on complex questions. Extensive experiments on multiple benchmarks show that AutoSearch achieves a superior accuracy-efficiency trade-off, alleviating over-searching while preserving search quality.

cs.AI↗

arXiv2026

Saliency-Guided Representation with Consistency Policy Learning for Visual Unsupervised Reinforcement Learning

Zero-shot unsupervised reinforcement learning (URL) offers a promising direction for building generalist agents capable of generalizing to unseen tasks without additional supervision. Among existing approaches, successor representations (SR) have emerged as a prominent paradigm due to their effectiveness in structured, low-dimensional settings. However, SR methods struggle to scale to high-dimensional visual environments. Through empirical analysis, we identify two key limitations of SR in visual URL: (1) SR objectives often lead to suboptimal representations that attend to dynamics-irrelevant regions, resulting in inaccurate successor measures and degraded task generalization; and (2) these flawed representations hinder SR policies from modeling multi-modal skill-conditioned action distributions and ensuring skill controllability. To address these limitations, we propose Saliency-Guided Representation with Consistency Policy Learning (SRCP), a novel framework that improves zero-shot generalization of SR methods in visual URL. SRCP decouples representation learning from successor training by introducing a saliency-guided dynamics task to capture dynamics-relevant representations, thereby improving successor measure and task generalization. Moreover, it integrates a fast-sampling consistency policy with URL-specific classifier-free guidance and tailored training objectives to improve skill-conditioned policy modeling and controllability. Extensive experiments on 16 tasks across 4 datasets from the ExORL benchmark demonstrate that SRCP achieves state-of-the-art zero-shot generalization in visual URL and is compatible with various SR methods.

cs.CV↗

arXiv2026

Evidence for a Delayed UV Counterpart to X-ray Quasi-periodic Eruptions in Ansky

X-ray quasi-periodic eruptions (QPEs) represent a novel population of extreme, repeating nuclear transients whose physical origins remain debated. A defining characteristic of QPEs has been their exclusive detection in the X-ray band, with a notable absence of correlated multi-wavelength counterparts. Here we report the first detection of a recurrent UV response temporally coupled to the X-ray QPE signal in the source Ansky/ZTF19acnskyy. The UV emission displays coherent periodic modulations over five consecutive cycles, systematically lagging the X-ray eruptions by $0.96^{+0.38}_{-0.39}$ days, with a cross-correlation coefficient of $r_{\rm max} \sim 0.6$. We suggest that the detectability of this corresponding signal may be enabled by Ansky's unusually long recurrence timescale, which could reduce the temporal smearing of the UV response seen in more rapid QPEs. The observed delay may correspond to a diffusion timescale associated with heated blobs. However, we cannot exclude the possibility that the lag corresponds to the light-crossing time associated with X-ray irradiation that originates near the central black hole and propagates to the outer UV-emitting region. While numerous QPE models have been proposed, any viable model for Ansky must be able to simultaneously explain the presence of a UV counterpart, its measured time lag, and the previously observed steadily increasing recurrence period.

astro-ph.HE↗

arXiv2026

Randomized Neural Networks for Partial Differential Equation on Static and Evolving Surfaces

Surface partial differential equations arise in numerous scientific and engineering applications. Their numerical solution on static and evolving surfaces remains challenging due to geometric complexity and, for evolving geometries, the need for repeated mesh updates and geometry or solution transfer. While neural-network-based methods offer mesh-free discretizations, approaches based on nonconvex training can be costly and may fail to deliver high accuracy in practice. In this work, we develop a randomized neural network (RaNN) method for solving PDEs on both static and evolving surfaces: the hidden-layer parameters are randomly generated and kept fixed, and the output-layer coefficients are determined efficiently by solving a least-squares problem. For static surfaces, we present formulations for parametrized surfaces, implicit level-set surfaces, and point-cloud geometries, and provide a corresponding theoretical analysis for the parametrization-based formulation with interface compatibility. For evolving surfaces with topology preserved over time, we introduce a RaNN-based strategy that learns the surface evolution through a flow-map representation and then solves the surface PDE on a space--time collocation set, avoiding remeshing. Extensive numerical experiments demonstrate broad applicability and favorable accuracy--efficiency performance on representative benchmarks.

math.NA↗

arXiv2025

MSME: A Multi-Stage Multi-Expert Framework for Zero-Shot Stance Detection

LLM-based approaches have recently achieved impressive results in zero-shot stance detection. However, they still struggle in complex real-world scenarios, where stance understanding requires dynamic background knowledge, target definitions involve compound entities or events that must be explicitly linked to stance labels, and rhetorical devices such as irony often obscure the author's actual intent. To address these challenges, we propose MSME, a Multi-Stage, Multi-Expert framework for zero-shot stance detection. MSME consists of three stages: (1) Knowledge Preparation, where relevant background knowledge is retrieved and stance labels are clarified; (2) Expert Reasoning, involving three specialized modules-Knowledge Expert distills salient facts and reasons from a knowledge perspective, Label Expert refines stance labels and reasons accordingly, and Pragmatic Expert detects rhetorical cues such as irony to infer intent from a pragmatic angle; (3) Decision Aggregation, where a Meta-Judge integrates all expert analyses to produce the final stance prediction. Experiments on three public datasets show that MSME achieves state-of-the-art performance across the board.

cs.CL↗

arXiv2025

Photoinduced giant modulation of terahertz nonlinearity from metasurfaces

Active control of optical nonlinearity is essential for advancing next-generation electronics and photonics, including high-speed wireless communications, optical information processing, and nonlinear signal manipulation. However, achieving tunable nonlinearity at terahertz (THz) frequencies faces significant challenges due to the lack of materials that combine high nonlinear responses with strong sensitivity to external stimuli in this spectral regime. Here, we show giant modulation of THz nonlinearity by optically tailoring the valley degree of freedom in semiconductor-based metasurfaces. Mediated by the resonant behaviors of metasurfaces, photoexcited electrons transition into different valleys in the conduction band in response to the driving THz field, with the transition rate controlled by light intensity. Since THz nonlinearities vary significantly with electron dynamics in different valleys, various nonlinear effects-such as nonlinear transmission and generation-can be efficiently enhanced and modulated within a single metasurface using weak optical pumping. With optical energy as low as a few picojoules, we achieve on-off switching of THz third harmonic generation with a modulation depth exceeding 20000%, along with effective tunability of its nonperturbative behaviors. Our approach breaks new ground in active THz devices fully compatible with semiconductor industry standards, indicating a promising building block for ultrafast THz signal processing, all-optical computing, and nonlinear optical elements.

physics.optics↗

arXiv2025

Perception-Consistency Multimodal Large Language Models Reasoning via Caption-Regularized Policy Optimization

While multimodal large language models excel at tasks that integrate visual perception with symbolic reasoning, their performance is often undermined by a critical vulnerability: perception-induced errors that propagate through the reasoning chain. Current reinforcement learning (RL) fine-tuning methods, while enhancing reasoning abilities, largely fail to address the underlying misalignment between visual grounding and the subsequent reasoning process. To address this challenge, we propose \textbf{Caption-Regularized Policy Optimization (CapPO)}, a novel RL framework that explicitly enforces perceptual consistency during policy optimization. CapPO integrates two key mechanisms: (1) a caption-based consistency regularization, which minimizes the divergence between responses conditioned on raw images and those conditioned on captions, thereby anchoring reasoning to semantically faithful visual content; and (2) a KL-weighted advantage estimation scheme, which adaptively scales reinforcement signals to strengthen perceptually consistent trajectories while suppressing spurious correlations. Extensive experiments on five math-focused and five general reasoning benchmarks demonstrate that CapPO achieves competitive performance, yielding gains of +6.0% accuracy on math-related tasks and +2.4% on general reasoning tasks over the base Qwen2.5-VL-7B model. Moreover, ablation studies further confirm the effectiveness of each component, while error analysis reveals that CapPO significantly reduces perception-related mistakes compared with baselines. Overall, CapPO provides a simple yet effective framework for improving multimodal reasoning.

cs.MM↗

arXiv2025

LLMsPark: A Benchmark for Evaluating Large Language Models in Strategic Gaming Contexts

As large language models (LLMs) advance across diverse tasks, the need for comprehensive evaluation beyond single metrics becomes increasingly important. To fully assess LLM intelligence, it is crucial to examine their interactive dynamics and strategic behaviors. We present LLMsPark, a game theory-based evaluation platform that measures LLMs' decision-making strategies and social behaviors in classic game-theoretic settings, providing a multi-agent environment to explore strategic depth. Our system cross-evaluates 15 leading LLMs (both commercial and open-source) using leaderboard rankings and scoring mechanisms. Higher scores reflect stronger reasoning and strategic capabilities, revealing distinct behavioral patterns and performance differences across models. This work introduces a novel perspective for evaluating LLMs' strategic intelligence, enriching existing benchmarks and broadening their assessment in interactive, game-theoretic scenarios. The benchmark and rankings are publicly available at https://llmsparks.github.io/.

cs.CL↗

arXiv2025

Tidal disruption events in active galactic nuclei: on orbital inclination and Schwarzschild apsidal precession

Tidal disruption events (TDEs) in active galactic nuclei (AGNs) mark a regime where traditional vacuum models fail to capture the full dynamics, especially due to interaction between stellar debris and pre-existing accretion disks. We perform meshless hydrodynamic simulations incorporating both general relativistic (GR) effects and radiative cooling to study TDEs in AGNs with different orbital inclinations ($\theta_{\rm inc}$) of the disrupted star, ranging from projected prograde to retrograde orbits. We post-process the simulations to derive multi-wavelength light curves and identify several distinct features in the light curves, including a precursor flare from early debris-disk collision and a major flare driven by fallback. The dynamics of the stellar debris and accretion disk, and subsequently the light curve features, are strongly affected by $\theta_{\rm inc}$ and GR effects. Retrograde orbits ($\theta_{\rm inc}=135^\circ$) yield a more luminous, shorter major flare and a more prominent precursor than prograde ones ($\theta_{\rm inc}=22.5^\circ$). During fallback, prograde cases ($\theta_{\rm inc} = 22.5^\circ$, $45^\circ$) develop a central cavity with spirals in the inner region of the AGN disk, leading to transient UV/X-ray suppression accompanied by oscillations, while higher inclinations ($\theta_{\rm inc}=90^\circ$, $135^\circ$) form a gradually tilting inner disk, potentially causing UV/X-ray dips via geometric effects at certain viewing angles. Relativistic apsidal precession alters stream collisions, producing structural differences in the inner disk, outer disk, and debris compared to Newtonian cases, and drives quasi-periodic signals in prograde configurations. These results provide predictive diagnostics for identifying AGN TDEs and interpreting observed light-curve diversity.

astro-ph.HE↗

arXiv2025

Why the hyperbolic polaritons are hyperbolic?

Polaritons travelling along a hyperbolic medium's surface have recently sparked significant interest in nanophotonics for the unprecedented manipulation ability on light at the nanoscale in a planar way, promising potential nano-optical applications, especially in two-dimensional circuitry. Despite of being named hyperbolic polaritons, the hyperbolic nature has not been thoroughly revealed since an analytical description of the Iso-frequency contour is still elusive. In this work, we proposed an analytical form for describing the iso-frequency contour of the hyperbolic polaritons, showcasing their strictly hyperbolic nature. Such an analytical form is obtained based on the focusing behavior of the hyperbolic polaritons and verified by both the published data from commonly used hyperbolic media systems of the hyperbolic polaritons and our own experimental characterizations on a hyperbolic metamaterial film. By presenting a concise and intuitive physical image, this work may provide a groundbreaking methodology in developing novel hyperbolic polaritons based optical devices.

physics.optics↗

arXiv2025

The Intermediate-Mass Black Hole Reverberation Mapping Project: First Detection of Mid-Infrared Lags in Prototypical IMBHs in NGC 4395 and POX 52

The search for robust evidence of intermediate-mass black holes (IMBHs) is crucial for understanding black hole seeding process and the formation of supermassive black holes in the early Universe. NGC 4395 and POX 52 are two prototypical IMBH hosts, both exhibiting multi-line evidence of low-mass black hole activity. Here, we report the first detection of mid-infrared (MIR) lags in response to optical variability, with measurements of $3.0^{+2.4}_{-1.9}$ days for NGC 4395 and $35.2^{+14.2}_{-11.7}$ days for POX~52 at $3.4$ $\mu$m, respectively, using archival optical data and observations from the Wide-field Infrared Survey Explorer (WISE). This detection provides the first reverberation evidence of low-mass black hole activity in POX 52. The time lags of these two low-mass, low-luminosity active galactic nuclei (AGNs) generally follow the extent of the $R_{\rm dust}-L_{\rm 5100}$ relation found in higher-mass AGNs. Based on an empirical relation between the broad-line region and dusty torus size, we constrain the black hole mass of POX 52 to log($M_{\rm BH}$/$M_\odot$) = 5.5 $\pm$ 0.37 (systemic and statistical errors), confirming its IMBH nature. Furthermore, long-term optical continuum monitoring of POX 52 reveals a mild inter-band lag of $\lesssim$ 1 day. However, no significant intranight variability was detected during its one-night, high-cadence monitoring, which we attribute to the longer duty cycle of fast variability in POX 52 compared to that in NGC 4395.

astro-ph.GA↗

arXiv2025

Salience-Invariant Consistent Policy Learning for Generalization in Visual Reinforcement Learning

Generalizing policies to unseen scenarios remains a critical challenge in visual reinforcement learning, where agents often overfit to the specific visual observations of the training environment. In unseen environments, distracting pixels may lead agents to extract representations containing task-irrelevant information. As a result, agents may deviate from the optimal behaviors learned during training, thereby hindering visual generalization.To address this issue, we propose the Salience-Invariant Consistent Policy Learning (SCPL) algorithm, an efficient framework for zero-shot generalization. Our approach introduces a novel value consistency module alongside a dynamics module to effectively capture task-relevant representations. The value consistency module, guided by saliency, ensures the agent focuses on task-relevant pixels in both original and perturbed observations, while the dynamics module uses augmented data to help the encoder capture dynamic- and reward-relevant representations. Additionally, our theoretical analysis highlights the importance of policy consistency for generalization. To strengthen this, we introduce a policy consistency module with a KL divergence constraint to maintain consistent policies across original and perturbed observations.Extensive experiments on the DMC-GB, Robotic Manipulation, and CARLA benchmarks demonstrate that SCPL significantly outperforms state-of-the-art methods in terms of generalization. Notably, SCPL achieves average performance improvements of 14\%, 39\%, and 69\% in the challenging DMC video hard setting, the Robotic hard setting, and the CARLA benchmark, respectively.Project Page: https://sites.google.com/view/scpl-rl.

cs.AI↗

arXiv2025

A simplified method for full-wave simulation of metamaterials: utilizing near-field decoupling technology

Simulating the electromagnetic properties of large-scale, complex metamaterial structures demands significant time and memory resources. If these large-scale structures can be divided into smaller, simpler components, the overall cost of studying all the smaller structures could be much lower than directly simulating the entire structure. Unfortunately, decoupling complex structures has been challenging due to the unclear mechanisms of near-field coupling in metamaterials. In this paper, we identify that the key to understanding near-field coupling in metamaterials lies in evanescent wave interactions, which can be captured through full-wave simulations. Our findings suggest that by accounting for the influence of evanescent waves, it becomes possible to analytically decouple and then recouple structures, even when the types of metamaterial structures vary. Building on this insight, we successfully decomposed complex structures into multiple groups of simpler components. By studying these simpler components, the electromagnetic properties of the entire structure can be calculated analytically. This decoupling method dramatically reduces the computation time or memory required for research into the electromagnetic properties of metamaterials.

physics.optics↗