arXiv subjects

Xin Zhang

Publications and source records attributed to Xin Zhang.

At least 325 records · Page 18Linked to original sources

arXiv2026

Loupe: A Generalizable and Adaptive Framework for Image Forgery Detection

The proliferation of generative models has raised serious concerns about visual content forgery. Existing deepfake detection methods primarily target either image-level classification or pixel-wise localization. While some achieve high accuracy, they often suffer from limited generalization across manipulation types or rely on complex architectures. In this paper, we propose Loupe, a lightweight yet effective framework for joint deepfake detection and localization. Loupe integrates a patch-aware classifier and a segmentation module with conditional queries, allowing simultaneous global authenticity classification and fine-grained mask prediction. To enhance robustness against distribution shifts of test set, Loupe introduces a pseudo-label-guided test-time adaptation mechanism by leveraging patch-level predictions to supervise the segmentation head. Extensive experiments on the DDL dataset demonstrate that Loupe achieves state-of-the-art performance, securing the first place in the IJCAI 2025 Deepfake Detection and Localization Challenge with an overall score of 0.846. Our results validate the effectiveness of the proposed patch-level fusion and conditional query design in improving both classification accuracy and spatial localization under diverse forgery patterns. The code is available at https://github.com/Kamichanw/Loupe.

cs.CV↗

arXiv2026

On regional treatment effect assessment using robust MAP priors

Bayesian dynamic borrowing has become an increasingly important tool for evaluating the consistency of regional treatment effects which is a key requirement for local regulatory approval of a new drug. It helps increase the precision of regional treatment effect estimate when regional and global data are similar, while guarding against potential bias when they differ. In practice, the two-component mixture prior, of which one mixture component utilizes the power prior to incorporate external data, is widely used. It allows convenient prior specification, analytical posterior computation, and fast evaluation of operating characteristics. Though the robust meta-analytical-predictive (MAP) prior is broadly used with multiple external data sources, it remains underutilized for regional treatment effect assessment (typically only one external data source is available) due to its inherit complexity in prior specification and posterior computation. In this article, we illustrate the applicability of the robust MAP prior in the regional treatment effect assessment by developing a closed-form approximation for its posterior distribution while leveraging its relationship with the power prior. The proposed methodology substantially reduces the computational burden of identifying prior parameters for desired operating characteristics. Moreover, we have demonstrated that the MAP prior is an attractive choice to construct the informative component of the mixture prior compared to the power prior. The advantage can be explained through a Bayesian hypothesis testing perspective. Using a real-world example, we illustrate how our proposed method enables efficient and transparent development of a Bayesian dynamic borrowing design to show regional consistency.

stat.ME↗

arXiv2026

Gravitational wave standard sirens: A brief review of cosmological parameter estimation

Gravitational wave (GW) observations are expected to serve as a powerful and independent probe of the expansion history of the universe. By providing direct and calibration-free measurements of luminosity distances through waveform analysis, GWs provide a fundamentally different and potentially more robust approach to measuring cosmic-scale distances compared to traditional electromagnetic (EM) observations, which is known as the standard siren method. In this review, we present an overview of recent developments in GW standard siren cosmology, including up-to-date $H_0$ constraints, and prospects for constraining cosmological parameters using future GW detections. A central focus of this review is the unique ability of GW observations to break cosmological parameter degeneracies inherent in the EM observations. We also briefly highlight the impact of systematic uncertainties, such as detector calibration, weak lensing, peculiar velocities, and host-galaxy catalog completeness, and corresponding potential mitigation strategies, which currently limit the constraint precision of cosmological parameters. Looking forward, we highlight the importance of combining GW standard sirens with other emerging late-universe cosmological probes such as fast radio bursts, 21 cm intensity mapping, and strong gravitational lensing to forge a precise cosmological probe for exploring the late universe.

astro-ph.CO↗

arXiv2025

TeleWorld: Towards Dynamic Multimodal Synthesis with a 4D World Model

World models aim to endow AI systems with the ability to represent, generate, and interact with dynamic environments in a coherent and temporally consistent manner. While recent video generation models have demonstrated impressive visual quality, they remain limited in real-time interaction, long-horizon consistency, and persistent memory of dynamic scenes, hindering their evolution into practical world models. In this report, we present TeleWorld, a real-time multimodal 4D world modeling framework that unifies video generation, dynamic scene reconstruction, and long-term world memory within a closed-loop system. TeleWorld introduces a novel generation-reconstruction-guidance paradigm, where generated video streams are continuously reconstructed into a dynamic 4D spatio-temporal representation, which in turn guides subsequent generation to maintain spatial, temporal, and physical consistency. To support long-horizon generation with low latency, we employ an autoregressive diffusion-based video model enhanced with Macro-from-Micro Planning (MMPL)--a hierarchical planning method that reduces error accumulation from frame-level to segment-level-alongside efficient Distribution Matching Distillation (DMD), enabling real-time synthesis under practical computational budgets. Our approach achieves seamless integration of dynamic object modeling and static scene representation within a unified 4D framework, advancing world models toward practical, interactive, and computationally accessible systems. Extensive experiments demonstrate that TeleWorld achieves strong performance in both static and dynamic world understanding, long-term consistency, and real-time generation efficiency, positioning it as a practical step toward interactive, memory-enabled world models for multimodal generation and embodied intelligence.

cs.CV↗

arXiv2025

A Stepwise-Enhanced Reasoning Framework for Large Language Models Based on External Subgraph Generation

Large Language Models (LLMs) have achieved strong performance across a wide range of natural language processing tasks in recent years, including machine translation, text generation, and question answering. As their applications extend to increasingly complex scenarios, however, LLMs continue to face challenges in tasks that require deep reasoning and logical inference. In particular, models trained on large scale textual corpora may incorporate noisy or irrelevant information during generation, which can lead to incorrect predictions or outputs that are inconsistent with factual knowledge. To address this limitation, we propose a stepwise reasoning enhancement framework for LLMs based on external subgraph generation, termed SGR. The proposed framework dynamically constructs query relevant subgraphs from external knowledge bases and leverages their semantic structure to guide the reasoning process. By performing reasoning in a step by step manner over structured subgraphs, SGR reduces the influence of noisy information and improves reasoning accuracy. Specifically, the framework first generates an external subgraph tailored to the input query, then guides the model to conduct multi step reasoning grounded in the subgraph, and finally integrates multiple reasoning paths to produce the final answer. Experimental results on multiple benchmark datasets demonstrate that SGR consistently outperforms strong baselines, indicating its effectiveness in enhancing the reasoning capabilities of LLMs.

cs.CL↗

arXiv2025

MiMo-Audio: Audio Language Models are Few-Shot Learners

Existing audio language models typically rely on task-specific fine-tuning to accomplish particular audio tasks. In contrast, humans are able to generalize to new audio tasks with only a few examples or simple instructions. GPT-3 has shown that scaling next-token prediction pretraining enables strong generalization capabilities in text, and we believe this paradigm is equally applicable to the audio domain. By scaling MiMo-Audio's pretraining data to over one hundred million of hours, we observe the emergence of few-shot learning capabilities across a diverse set of audio tasks. We develop a systematic evaluation of these capabilities and find that MiMo-Audio-7B-Base achieves SOTA performance on both speech intelligence and audio understanding benchmarks among open-source models. Beyond standard metrics, MiMo-Audio-7B-Base generalizes to tasks absent from its training data, such as voice conversion, style transfer, and speech editing. MiMo-Audio-7B-Base also demonstrates powerful speech continuation capabilities, capable of generating highly realistic talk shows, recitations, livestreaming and debates. At the post-training stage, we curate a diverse instruction-tuning corpus and introduce thinking mechanisms into both audio understanding and generation. MiMo-Audio-7B-Instruct achieves open-source SOTA on audio understanding benchmarks (MMSU, MMAU, MMAR, MMAU-Pro), spoken dialogue benchmarks (Big Bench Audio, MultiChallenge Audio) and instruct-TTS evaluations, approaching or surpassing closed-source models. Model checkpoints and full evaluation suite are available at https://github.com/XiaomiMiMo/MiMo-Audio.

cs.CL↗

arXiv2025

Proper conflict-free choosability of planar graphs

A proper conflict-free coloring of a graph is a proper vertex coloring wherein each non-isolated vertex's open neighborhood contains at least one color appearing exactly once. For a non-negative integer $k$, a graph $G$ is said to be proper conflict-free (degree+$k$)-choosable if given any list assignment $L$ for $G$ where $|L(v)| = d(v) + k$ holds for every vertex $v \in V(G)$, there exists a proper conflict-free coloring $ϕ$ of $G$ such that $ϕ(v) \in L(v)$ for all $v \in V(G)$. Recently, Kashima, Škrekovski, and Xu proposed two related conjectures on proper conflict-free choosability: the first asserts the existence of an absolute constant $k$ such that every graph is proper conflict-free (degree+$k$)-choosable, while the second strengthens this claim by restricting to connected graphs other than the cycle of length 5 and reducing the constant to $k=2$. In this paper, we confirm the second conjecture for three graph classes: $K_4$-minor-free graphs with maximum degree at most 4, outer-1-planar graphs with maximum degree at most 4, and planar graphs with girth at least 12; we also confirm the first conjecture for these same graph classes, in addition to all outer-1-planar graphs (without degree constraints). Moreover, we prove that planar graphs with girth at least 12 and outer-1-planar graphs are proper conflict-free $6$-choosable.

math.CO↗

arXiv2025

Fast algorithm for $S$-packing coloring of Halin graphs

Motivated by frequency assignment problems in wireless broadcast networks, Goddard, Hedetniemi, Hedetniemi, Harris, and Rall introduced the notion of $S$-packing coloring in 2008. Given a non-decreasing sequence $S = (s_1, s_2, \ldots, s_k)$ of positive integers, an $S$-packing coloring of a graph $G$ is a partition of its vertex set into $k$ subsets $\{V_1, V_2, \ldots, V_k\}$ such that for each $1 \leq i \leq k$, the distance between any two distinct vertices $u, v \in V_i$ is at least $s_i + 1$. In this paper, we study the $S$-packing coloring problem for Halin graphs with maximum degree $Δ\leq 5$. Specifically, we present a linear-time algorithm that constructs a $(1,1,2,2,2)$-packing coloring for any Halin graph satisfying $Δ\leq 5$. It is worth noting that there are Halin graphs that are not $(1,2,2,2)$-packing colorable.

math.CO↗

arXiv2025

Exact rainbow numbers of cycle-related graphs in multi-hubbed wheels

The rainbow number ${\rm rb}(G, H)$ is the minimum number of colors $k$ for which any edge-coloring of $G$ with at least $k$ colors guarantees a rainbow subgraph isomorphic to $H$. The rainbow number has many applications in diverse fields such as wireless communication networks, cryptography, bioinformatics, and social network analysis. In this paper, we determine the exact rainbow number $\mathrm{rb}(G, H)$ where $G$ is a multi-hubbed wheel graph $W_d(s)$, defined as the join of $s$ isolated vertices and a cycle $C_d$ of length $d$ (i.e., $W_d(s) = \overline{K_s} + C_d$), and $H = θ_{t,\ell}$ represents a cycle $C_t$ of length $t$ with $0 \leq \ell \leq t-3$ chords emanating from a common vertex, by establishing \[ {\rm rb}(W_{d}(s), θ_{t,\ell}) = \begin{cases} \left\lfloor \dfrac{2t - 5}{t - 2}d \right\rfloor + 1, & \text{if } \ell=t-3,~s = 1 \text{ and } t\ge 4, \\[10pt] \left\lfloor \dfrac{3t-10}{t - 3}d \right\rfloor + 1, & \text{if } \ell=t-3,~s = 2\text{ and } t\ge 6,\\[10pt] \left\lfloor \dfrac{(s + 1)t - (3s + 4)}{t - 3}d \right\rfloor + 1, & \text{if } \ell=t-3,~s \geq 3\text{ and } t\ge 7,\\[10pt] \left\lfloor \dfrac{2t - 7}{t - 3}d \right\rfloor + 1, & \text{if } s = 1 \text{ and } t\ge \max\{5,\ell+4\}, \end{cases} \] when $d\geq 3t-5$, with all bounds for the parameter $t$ presented here being tight. This addresses the problems proposed by Jakhar, Budden, and Moun (2025), which involve investigating the rainbow numbers of large cycles and large chorded cycles in wheel graphs (specifically corresponding to the cases in our framework where $s=1$ and $\ell\in \{0,1\}$). Furthermore, it completely determines the rainbow numbers of cycles of arbitrary length in large wheel graphs, thereby generalizing a result of Lan, Shi, and Song (2019).

math.CO↗

arXiv2025

The Multi-Messenger Astroparticle Physics: the First Constraint on Light Millicharged Dark Matter via Time-Delay Analysis of GRB GW170817A

The multi-messenger astroparticle physics provides a new approach to probe the new physics beyond the Standard Model. We propose to probe the light dark matter which can interact with electromagnetic interaction. To be concrete, we derive the new constraint on the millicharged dark matter from the multi-messenger observations of GW170817. In the neutron star merger event GW170817, the first detection of a gamma-ray burst (GRB) delayed by approximately 1.7 seconds relative to the gravitational wave emission was observed. Utilizing this delay, we constrain the parameter space of the millicharged dark matter within the large-scale structure of the Universe. For dark matter mass below $10^{-15}$ eV, the parameter $ε$ is constrained to be less than $10^{-14}$, representing the most stringent limits achieved to date.

hep-ph↗

arXiv2025

DrVoice: Parallel Speech-Text Voice Conversation Model via Dual-Resolution Speech Representations

Recent studies on end-to-end (E2E) speech generation with large language models (LLMs) have attracted significant community attention, with multiple works extending text-based LLMs to generate discrete speech tokens. Existing E2E approaches primarily fall into two categories: (1) Methods that generate discrete speech tokens independently without incorporating them into the LLM's autoregressive process, resulting in text generation being unaware of concurrent speech synthesis. (2) Models that generate interleaved or parallel speech-text tokens through joint autoregressive modeling, enabling mutual modality awareness during generation. This paper presents DrVoice, a parallel speech-text voice conversation model based on joint autoregressive modeling, featuring dual-resolution speech representations. Notably, while current methods utilize mainly 12.5Hz input audio representation, our proposed dual-resolution mechanism reduces the input frequency for the LLM to 5Hz, significantly reducing computational cost and alleviating the frequency discrepancy between speech and text tokens and in turn better exploiting LLMs' capabilities. Experimental results demonstrate that DrVoice-7B establishes new state-of-the-art (SOTA) on prominent speech benchmarks including OpenAudioBench, VoiceBench, UltraEval-Audio and Big Bench Audio, making it a leading open-source speech foundation model in ~7B models.

cs.CL↗

arXiv2025

Cosmological search for sterile neutrinos after DESI 2024

The question of whether the massive sterile neutrinos exist remains a crucial unresolved issue in both particle physics and cosmology. We explore the cosmological constraints on the massive sterile neutrinos using the latest observational data, including the baryon acoustic oscillations data from DESI, the cosmic microwave background data from Planck satellite and ACT, and the 5-year Type Ia supernova data and the 3-year weak-lensing data from DES. We search for the massive sterile neutrinos within the $Λ$CDM, $w$CDM, and $w_0w_a$CDM models. Our analysis shows that when considering massive sterile neutrinos within the $w_0w_a\rm CDM$ model, the combined datasets allow us to infer a non-zero sterile neutrino mass at approximately $2σ$ confidence level. Specifically, in the $w_0w_a$CDM+Sterile model, the effective mass of sterile neutrinos and the effective number of relativistic species are constrained to be $m_{ν,\ \mathrm{sterile}}^{\mathrm{eff}} = 0.50^{+0.33}_{-0.27} \, \mathrm{eV}$ and $N_\mathrm{eff} = 3.076^{+0.011}_{-0.017}$, respectively. However, the $Λ$CDM+Sterile and $w$CDM+Sterile models could not provide evidence supporting the existence of massive sterile neutrinos.

astro-ph.CO↗

arXiv2025

MCI: Multi-Channel Imager on the Chinese Space Station Survey Telescope

The Multi-Channel Imager (MCI) is a powerful near-ultraviolet (NUV) and visible imager onboard the Chinese Space Station Survey Telescope (CSST). The MCI provides three imaging channels, which are the NUV channel, the Blue channel and the Red channel, with the wavelength range of 255-430 nm, 430-700 nm, and 700-1000 nm, respectively. MCI's three channels can target the same field simultaneously, which is unique compared to other imagers onboard the Hubble Space Telescope (HST) or the James Webb Space Telescope (JWST). Each channel employs a CCD focal plane of 9216 x 9232 pixels and $\sim$7\arcmin.5 x 7\arcmin.5 field of view (FOV), which are about $\gtrsim 4$ times greater than the FOVs of HST imagers. The MCI's three channels feature unprecedented sensitivities and field of views complement the NUV and visible capabilities of the CSST for high-precision photometry and weak-signal detection, which would help build a new standard-star system and the deepest UV-Optical exposures for CSST. Rich filter sets of MCI would help explore other sciences such as local emission line mapping, high-z Ly$α$ emitters searching, etc. Here we present key design features, results of current ground tests, and suggested observing strategies of the MCI.

astro-ph.IM↗

arXiv2025

Constraining primordial black hole abundance with Insight-HXMT

Primordial black holes (PBHs) are a major candidate for dark matter and they have been extensively constrained across most mass ranges. However, PBHs in the mass range of $10^{17}$ - $10^{21}$ g remain a viable explanation for all dark matter. In this work, we use observational data from the Hard X-ray Modulation Telescope (Insight-HXMT) to refine constraints on PBHs within the mass range of $2\times10^{16}$ - $5\times10^{17}$ g. Our analysis explores three scenarios: directly using observational data, incorporating the astrophysical background model (ABM), and employing the power-law spectrum with an exponential cutoff. Our results indicate that although Insight-HXMT does not have an advantage in the first two scenarios, when considering the power-law model, its exceptional sensitivity in the hard X-ray regime and sufficiently high upper energy limit significantly strengthen the constraints on PBHs with masses greater than $10^{17}$ g compared to previous limits. Furthermore, the exclusion limit for PBHs as dark matter has reached $4\times10^{17}$ g, which is comparable to the current threshold.

astro-ph.CO↗

arXiv2025

Survey and Experiments on Mental Disorder Detection via Social Media: From Large Language Models and RAG to Agents

Mental disorders represent a critical global health challenge, and social media is increasingly viewed as a vital resource for real-time digital phenotyping and intervention. To leverage this data, large language models (LLMs) have been introduced, offering stronger semantic understanding and reasoning than traditional deep learning, thereby enhancing the explainability of detection results. Despite the growing prominence of LLMs in this field, there is a scarcity of scholarly works that systematically synthesize how advanced enhancement techniques, specifically Retrieval-Augmented Generation (RAG) and Agentic systems, can be utilized to address these reliability and reasoning limitations. Here, we systematically survey the evolving landscape of LLM-based methods for social media mental disorder analysis, spanning standard pre-trained language models, RAG to mitigate hallucinations and contextual gaps, and agentic systems for autonomous reasoning and multi-step intervention. We organize existing work by technical paradigm and clinical target, extending beyond common internalizing disorders to include psychotic disorders and externalizing behaviors. Additionally, the paper comprehensively evaluates the performance of LLMs, including the impact of RAG, across various tasks. This work establishes a unified benchmark for the field, paving the way for the development of trustworthy, autonomous AI systems that can deliver precise and explainable mental health support.

cs.CL↗

arXiv2025

Revisiting the phenomenologically emergent dark energy model: is non-zero equation of state of dark matter favored by DESI DR2?

The nature of dark matter remains one of the most fundamental and unresolved questions in modern cosmology. In most cosmological models, dark matter is typically modeled as pressureless dust with an equation of state (EoS) parameter $w_{\rm dm} = 0$. However, there is no fundamental theoretical reason to exclude the possibility of a non-zero dark matter EoS parameter. In this work, we explore the possibility of a non-zero dark matter EoS within the phenomenologically emergent dark energy (PEDE) model, given its simplicity and proven ability to alleviate the Hubble tension. We perform observational constraints by using the latest baryon acoustic oscillation data from DESI DR2, the cosmic microwave background (CMB) data from Planck, and the type Ia supernova data from DESY5 and PantheonPlus. From our analysis, we observe that a negative dark matter EoS parameter is preferred in all scenarios. Specifically, the CMB+DESI+DESY5 data yields $w_{\mathrm{dm}} = -0.00093 \pm 0.00032$, deviating from zero at approximately the $3σ$ level. However, this deviation is likely driven by unidentified systematics or inconsistencies in the DESY5 data, with the deviation decreasing to $2σ$ when using PantheonPlus data. Meanwhile, a negative $w_{\rm dm}$ would increase the Hubble tension due to the positive degeneracy between $w_{\rm dm}$ and $H_0$. Furthermore, Bayesian evidence suggests that the $Λ$CDM model is strongly preferred over the PEDE+$w_{\rm dm}$ model. These analyses illustrate that it is not possible to both support a non-cold dark matter component within the PEDE model and alleviate the Hubble tension simultaneously.

astro-ph.CO↗

arXiv2025

Measuring neutrino masses with joint JWST and DESI DR2 data

Early JWST observations reveal an unexpectedly abundant population of high-redshift candidate massive galaxies at $z \gtrsim 7$, and recent DESI measurements show a preference for dynamical dark energy, which together present a significant challenge to the standard $Λ$CDM cosmology. In this work, we jointly analyze high-redshift galaxy data from JWST, baryon acoustic oscillations data from DESI DR2, and cosmic microwave background (CMB) data from Planck and ACT, measuring the total neutrino mass $\sum m_ν$. We consider three dark energy models ($Λ$CDM, $w$CDM, and $w_0w_a$CDM) and three mass hierarchies. Our results indicate that in the $w_0w_a$CDM model, adding JWST data to CMB+DESI tightens the upper limit of $\sum m_ν$ by about $5.8\%-10.2\%$, and we obtain $\sum m_ν < 0.167~\mathrm{eV}$ ($2σ$) in the normal hierarchy (NH) case. Furthermore, JWST also offers indicative lower limits on star formation efficiency parameter of $f_{*,10} \gtrsim 0.146-0.161$. Bayesian evidence weakly favors the $w_0w_a$CDM+$\sum m_ν$(NH) model relative to the $Λ$CDM+$\sum m_ν$(NH) model using CMB+DESI+JWST data. These results suggest that the joint analysis of high-redshift JWST data and low-redshift DESI data provides compelling constraints on neutrino mass and merits further investigation.

astro-ph.CO↗

arXiv2025

CMB delensing with deep learning

The cosmic microwave background (CMB) stands as a pivotal source for studying weak gravitational lensing. While the lensed CMB aids in constraining cosmological parameters, it simultaneously smooths the original CMB's features. The angular power spectrum of the unlensed CMB showcases sharper acoustic peaks and more pronounced damping tails, enhancing the precision of inferring cosmological parameters that influence these aspects. Although delensing diminishes the $B$-mode power spectrum, it facilitates the pursuit of primordial gravitational waves and enables a lower variance reconstruction of lensing and additional sources of secondary CMB anisotropies. In this work, we explore the potential of deep learning techniques, specifically the U-Net++ algorithm, to play a pivotal role in CMB delensing. We analyze three fields, namely $T$, $Q$, and $U$ sky maps, present the angular power spectra of the CMB delensed $TT$, $EE$, $BB$, and $TE$, and compare them with the unlensed CMB angular power spectra. Our findings reveal that the angular power spectrum of the lensed CMB, processed by U-Net++, closely aligns with that of the unlensed CMB. Thus, U-Net++ based CMB delensing proves to be effective in mitigating the impacts of weak gravitational lensing, paving the way for enhancing the CMB delensing power spectrum in forthcoming CMB experiments. The code utilized for this analysis is available on GitHub.

astro-ph.CO↗