arXiv subjects

Jian Wang

Publications and source records attributed to Jian Wang.

At least 253 records · Page 14Linked to original sources

arXiv2026

Tailoring Dynamical Quantum Phase Transitions via Double-Mode Squeezing Manipulation

We propose a protocol to tailor dynamical quantum phase transitions (DQPTs) by double-mode squeezing onto the initial state in the XY chain. The effect of squeezing depends critically on the system's symmetry and parameters. When the squeezing operator breaks particle-hole symmetry (PHS), DQPTs become highly tunable, allowing one to either induce transitions within a single phase or suppress them. Remarkably, when PHS is preserved and the squeezing strength reaches $r=π/4$, a universal class of DQPTs emerges, independent of the quench path. This universality is characterized by two key features: (i) the collapse of all Fisher zeros onto the real-time axis, and (ii) the saturation of intermode entanglement to its maximum in each $(k,-k)$ modes. Moreover, the critical momenta governing the DQPTs coincide exactly with the modes attaining the maximal entanglement. At this universal point, the dynamical phase vanishes, leading to a purely geometric evolution marked by $π$-jumps in the Pancharatnam geometric phase. Our work establishes initial-state squeezing as a versatile tool for tailoring far-from-equilibrium criticality and reveals a direct link between entanglement saturation and universal nonanalytic dynamics.

quant-ph↗

arXiv2026

AHA! Animating Human Avatars in Diverse Scenes with Gaussian Splatting

We present a novel framework for animating humans in 3D scenes using 3D Gaussian Splatting (3DGS), a neural scene representation that has recently achieved state-of-the-art photorealistic results for novel-view synthesis but remains under-explored for human-scene animation and interaction. Unlike existing animation pipelines that use meshes or point clouds as the underlying 3D representation, our approach introduces the use of 3DGS as the 3D representation for animating humans in scenes. By representing humans and scenes as Gaussians, our approach allows geometry-consistent free-viewpoint rendering of humans interacting with 3D scenes. Our key insight is that rendering can be decoupled from motion synthesis, and each sub-problem can be addressed independently without the need for paired human-scene data. Central to our method is a Gaussian-aligned motion module that synthesizes motion without explicit scene geometry, using opacity-based cues and projected Gaussian structures to guide human placement and pose alignment. To ensure natural interactions, we further propose a human-scene Gaussian refinement optimization that enforces realistic contact and navigation. We evaluate our approach on scenes from Scannet++ and the SuperSplat library, and on avatars reconstructed from sparse and dense multi-view human capture. Finally, we demonstrate that our framework enables novel applications such as geometry-consistent free-viewpoint rendering of edited monocular RGB videos with newly animated humans, showcasing the unique advantages of 3DGS for monocular video-based human animation. To assess the full quality of our results, we encourage readers to view the supplementary material available at https://miraymen.github.io/aha/ .

cs.CV↗

arXiv2026

Animated 3DGS Avatars in Diverse Scenes with Consistent Lighting and Shadows

We present a method for consistent lighting and shadows when animated 3D Gaussian Splatting (3DGS) avatars interact with 3DGS scenes or with dynamic objects inserted into otherwise static scenes. Our key contribution is Deep Gaussian Shadow Maps (DGSM), a modern analogue of the classical shadow mapping algorithm tailored to the volumetric 3DGS representation. Building on the classic deep shadow mapping idea, we show that 3DGS admits closed form light accumulation along light rays, enabling volumetric shadow computation without meshing. For each estimated light, we tabulate transmittance over concentric radial shells and store them in octahedral atlases, which modern GPUs can sample in real time per query to attenuate affected scene Gaussians and thus cast and receive shadows consistently. To relight moving avatars, we approximate the local environment illumination with HDRI probes represented in a spherical harmonic (SH) basis and apply a fast per Gaussian radiance transfer, avoiding explicit BRDF estimation or offline optimization. We demonstrate environment consistent lighting for avatars from AvatarX and ActorsHQ, composited into ScanNet++, DL3DV, and SuperSplat scenes, and show interactions with inserted objects. Across single and multi avatar settings, DGSM and SH relighting operate fully in the volumetric 3DGS representation, yielding coherent shadows and relighting while avoiding meshing.

cs.CV↗

arXiv2026

A note on the maximum diversity of intersecting families in the symmetric group

Let $\mathcal{S}_n$ be the symmetric group on the set $[n]:=\{1,2,\ldots,n\}$. A family $\mathcal{F}\subset \mathcal{S}_n$ is called intersecting if for every $σ,π\in \mathcal{F}$ there exists some $i\in [n]$ such that $σ(i)=π(i)$. Deza and Frankl proved that the largest intersecting family of permutations is the full star, that is, the collection of all permutations with a fixed position. The diversity of an intersecting family $\mathcal{F}$ is defined as the minimum number of permutations in $\mathcal{F}$, whose deletion results in a star. In the present paper, by applying the spread approximation method developed recently by Kupavskii and Zakharov, we prove that for $n\geq 500$ the diversity of an intersecting subfamily of $\mathcal{S}_n$ is at most $(n-3)(n-3)!$, which is best possible.

math.CO↗

arXiv2025

TTC: Transformer-based TDE Classifier for the Wide Field Survey Telescope (WFST)

We propose the Transformer-based Tidal disruption events (TDE) Classifier (\texttt{TTC}), specifically designed to operate effectively with both real-time alert streams and archival data of the Wide Field Survey Telescope (WFST). It aims to minimize the reliance on external catalogs and find TDE candidates from pure light curves, which is more suitable for finding TDEs in faint and distant galaxies. \texttt{TTC} consists of two key modules that can work independently: (1) A light curve parametric fitting module and (2) a Transformer (\texttt{Mgformer})-based classification network. The training of the latter module and evaluation for each module utilize a light curve dataset of 7413 spectroscopically classified transients from the Zwicky Transient Facility (ZTF). The \texttt{Mgformer}-based module is superior in performance and flexibility. Its representative recall and precision values are 0.79 and 0.76, respectively, and can be modified by adjusting the threshold. It can also efficiently find TDE candidates within 30 days from the first detection. For comparison, the parametric fitting module yields values of 0.72 and 0.40, respectively, while it is $>$10 times faster in average speed. Hence, the setup of modules allows a trade-off between performance and time, as well as precision and recall. \texttt{TTC} has successfully picked out all spectroscopically identified TDEs among ZTF transients in a real-time classification test, and selected $\sim$20 TDE candidates in the deep field survey data of WFST. The discovery rate will greatly increase once the differential database for the wide field survey is ready.

astro-ph.IM↗

arXiv2025

Convergence rates for the $p$-Wasserstein distance of the empirical measures of an ergodic Markov process

Let $X:=(X_t)_{t\geq 0}$ be an ergodic Markov process on $\real^d$, and $p>0$. We derive upper bounds of the $p$-Wasserstein distance between the invariant measure and the empirical measures of the Markov process $X$. For this we assume, e.g.\ that the transition semigroup of $X$ is exponentially contractive in terms of the $1$-Wasserstein distance, or that the iterated Poincaré inequality holds together with certain moment conditions on the invariant measure. Typical examples include diffusions and underdamped Langevin dynamics.

math.PR↗

arXiv2025

Perplexity-Aware Data Scaling Law: Perplexity Landscapes Predict Performance for Continual Pre-training

Continual Pre-training (CPT) serves as a fundamental approach for adapting foundation models to domain-specific applications. Scaling laws for pre-training define a power-law relationship between dataset size and the test loss of an LLM. However, the marginal gains from simply increasing data for CPT diminish rapidly, yielding suboptimal data utilization and inefficient training. To address this challenge, we propose a novel perplexity-aware data scaling law to establish a predictive relationship between the perplexity landscape of domain-specific data and the test loss. Our approach leverages the perplexity derived from the pre-trained model on domain data as a proxy for estimating the knowledge gap, effectively quantifying the informational perplexity landscape of candidate training samples. By fitting this scaling law across diverse perplexity regimes, we enable adaptive selection of high-utility data subsets, prioritizing content that maximizes knowledge absorption while minimizing redundancy and noise. Extensive experiments demonstrate that our method consistently identifies near-optimal training subsets and achieves superior performance on both medical and general-domain benchmarks.

cs.LG↗

arXiv2025

FlashVLM: Text-Guided Visual Token Selection for Large Multimodal Models

Large vision-language models (VLMs) typically process hundreds or thousands of visual tokens per image or video frame, incurring quadratic attention cost and substantial redundancy. Existing token reduction methods often ignore the textual query or rely on deep attention maps, whose instability under aggressive pruning leads to degraded semantic alignment. We propose FlashVLM, a text guided visual token selection framework that dynamically adapts visual inputs to the query. Instead of relying on noisy attention weights, FlashVLM computes an explicit cross modal similarity between projected image tokens and normalized text embeddings in the language model space. This extrinsic relevance is fused with intrinsic visual saliency using log domain weighting and temperature controlled sharpening. In addition, a diversity preserving partition retains a minimal yet representative set of background tokens to maintain global context. Under identical token budgets and evaluation protocols, FlashVLM achieves beyond lossless compression, slightly surpassing the unpruned baseline while pruning up to 77.8 percent of visual tokens on LLaVA 1.5, and maintaining 92.8 percent accuracy even under 94.4 percent compression. Extensive experiments on 14 image and video benchmarks demonstrate that FlashVLM delivers state of the art efficiency performance trade offs while maintaining strong robustness and generalization across mainstream VLMs.

cs.CV↗

arXiv2025

LLM-CAS: Dynamic Neuron Perturbation for Real-Time Hallucination Correction

Large language models (LLMs) often generate hallucinated content that lacks factual or contextual grounding, limiting their reliability in critical applications. Existing approaches such as supervised fine-tuning and reinforcement learning from human feedback are data intensive and computationally expensive, while static parameter editing methods struggle with context dependent errors and catastrophic forgetting. We propose LLM-CAS, a framework that formulates real-time hallucination correction as a hierarchical reinforcement learning problem. LLM-CAS trains an agent to learn a policy that dynamically selects temporary neuron perturbations during inference based on the current context. Unlike prior dynamic approaches that rely on heuristic or predefined adjustments, this policy driven mechanism enables adaptive and fine grained correction without permanent parameter modification. Experiments across multiple language models demonstrate that LLM-CAS consistently improves factual accuracy, achieving gains of 10.98 percentage points on StoryCloze, 2.71 points on TriviaQA, and 2.06 points on the MC1 score of TruthfulQA. These results outperform both static editing methods such as ITI and CAA and the dynamic SADI framework. Overall, LLM-CAS provides an efficient and context aware solution for improving the reliability of LLMs, with promising potential for future multimodal extensions.

cs.CL↗

arXiv2025

OptScale: Probabilistic Optimality for Inference-time Scaling

Inference-time scaling has emerged as a powerful technique for enhancing the reasoning performance of Large Language Models (LLMs). However, existing approaches often rely on heuristic strategies for parallel sampling, lacking a principled foundation. To address this gap, we propose a probabilistic framework that formalizes the optimality of inference-time scaling under the assumption that parallel samples are independently and identically distributed (i.i.d.), and where the Best-of-$N$ selection strategy follows a probability distribution that can be estimated. Within this framework, we derive a theoretical lower bound on the required number of samples to achieve a target performance level, providing the first principled guidance for compute-efficient scaling. Leveraging this insight, we develop \textsc{OptScale}, a practical algorithm that dynamically determines the optimal number of sampled responses. \textsc{OptScale} employs a language model-based predictor to estimate probabilistic prior parameters, enabling the decision of the minimal number of samples needed that satisfy predefined performance thresholds and confidence levels. Extensive experiments on representative reasoning benchmarks (including MATH-500, GSM8K, AIME, and AMC) demonstrate that \textsc{OptScale} significantly reduces sampling overhead while remaining better or on par with state-of-the-art reasoning performance. Our work offers both a theoretical foundation and a practical solution for principled inference-time scaling, addressing a critical gap in the efficient deployment of LLMs for complex reasoning.

cs.LG↗

arXiv2025

GAPS: A Clinically Grounded, Automated Benchmark for Evaluating AI Clinicians

Current benchmarks for AI clinician systems, often based on multiple-choice exams or manual rubrics, fail to capture the depth, robustness, and safety required for real-world clinical practice. To address this, we introduce the GAPS framework, a multidimensional paradigm for evaluating Grounding (cognitive depth), Adequacy (answer completeness), Perturbation (robustness), and Safety. Critically, we developed a fully automated, guideline-anchored pipeline to construct a GAPS-aligned benchmark end-to-end, overcoming the scalability and subjectivity limitations of prior work. Our pipeline assembles an evidence neighborhood, creates dual graph and tree representations, and automatically generates questions across G-levels. Rubrics are synthesized by a DeepResearch agent that mimics GRADE-consistent, PICO-driven evidence review in a ReAct loop. Scoring is performed by an ensemble of large language model (LLM) judges. Validation confirmed our automated questions are high-quality and align with clinician judgment (90% agreement, Cohen's Kappa 0.77). Evaluating state-of-the-art models on the benchmark revealed key failure modes: performance degrades sharply with increased reasoning depth (G-axis), models struggle with answer completeness (A-axis), and they are highly vulnerable to adversarial perturbations (P-axis) as well as certain safety issues (S-axis). This automated, clinically-grounded approach provides a reproducible and scalable method for rigorously evaluating AI clinician systems and guiding their development toward safer, more reliable clinical practice. The benchmark dataset GAPS-NSCLC-preview and evaluation code are publicly available at https://huggingface.co/datasets/AQ-MedAI/GAPS-NSCLC-preview and https://github.com/AQ-MedAI/MedicalAiBenchEval.

cs.CL↗

arXiv2025

TalkVerse: Democratizing Minute-Long Audio-Driven Video Generation

We introduce TalkVerse, a large-scale, open corpus for single-person, audio-driven talking video generation designed to enable fair, reproducible comparison across methods. While current state-of-the-art systems rely on closed data or compute-heavy models, TalkVerse offers 2.3 million high-resolution (720p/1080p) audio-video synchronized clips totaling 6.3k hours. These are curated from over 60k hours of video via a transparent pipeline that includes scene-cut detection, aesthetic assessment, strict audio-visual synchronization checks, and comprehensive annotations including 2D skeletons and structured visual/audio-style captions. Leveraging TalkVerse, we present a reproducible 5B DiT baseline built on Wan2.2-5B. By utilizing a video VAE with a high downsampling ratio and a sliding window mechanism with motion-frame context, our model achieves minute-long generation with low drift. It delivers comparable lip-sync and visual quality to the 14B Wan-S2V model but with 10$\times$ lower inference cost. To enhance storytelling in long videos, we integrate an MLLM director to rewrite prompts based on audio and visual cues. Furthermore, our model supports zero-shot video dubbing via controlled latent noise injection. We open-source the dataset, training recipes, and 5B checkpoints to lower barriers for research in audio-driven human video generation. Project Page: https://zhenzhiwang.github.io/talkverse/

cs.CV↗

arXiv2025

Andr{á}sfai--Erdős--Sós theorem under max-degree constraints

We establish the following strengthening of the celebrated Andr{á}sfai--Erdős--Sós theorem: If $G$ is an $n$-vertex $K_{r+1}$-free graph whose minimum degree $δ(G)$ and maximum degree $Δ(G)$ satisfy \begin{align*} δ(G) > \min \left\{ \frac{3r-4}{3r-2}n-\frac{Δ(G)}{3r-2},~n-\frac{Δ(G)+1}{r-1} \right\}, \end{align*} then $G$ is $r$-partite. This bound is tight for all feasible values of $Δ(G)$. We also obtain an analogous tight result for graphs with large odd girth. Our proof does not rely on the Andr{á}sfai--Erdős--Sós theorem itself, and therefore yields an alternative proof of this classical result.

math.CO↗

arXiv2025

MM-CoT:A Benchmark for Probing Visual Chain-of-Thought Reasoning in Multimodal Models

The ability to perform Chain-of-Thought (CoT) reasoning marks a major milestone for multimodal models (MMs), enabling them to solve complex visual reasoning problems. Yet a critical question remains: is such reasoning genuinely grounded in visual evidence and logically coherent? Existing benchmarks emphasize generation but neglect verification, i.e., the capacity to assess whether a reasoning chain is both visually consistent and logically valid. To fill this gap, we introduce MM-CoT, a diagnostic benchmark specifically designed to probe the visual grounding and logical coherence of CoT reasoning in MMs. Instead of generating free-form explanations, models must select the sole event chain that satisfies two orthogonal constraints: (i) visual consistency, ensuring all steps are anchored in observable evidence, and (ii) logical coherence, ensuring causal and commonsense validity. Adversarial distractors are engineered to violate one of these constraints, exposing distinct reasoning failures. We evaluate leading vision-language models on MM-CoT and find that even the most advanced systems struggle, revealing a sharp discrepancy between generative fluency and true reasoning fidelity. MM-CoT shows low correlation with existing benchmarks, confirming that it measures a unique combination of visual grounding and logical reasoning. This benchmark provides a foundation for developing future models that reason not just plausibly, but faithfully and coherently within the visual world.

cs.CV↗

arXiv2025

Observation of Stable Bimeron Transport Driven by Spoof Surface Acoustic Waves on Chiral Metastructures

Topological quasiparticles, such as merons and bimerons, are characterized by non-trivial textures that exhibit remarkably robust transport against deformation, offering significant potential for information processing. While these phenomena have been explored in various systems, acoustic realizations remain challenging. Here, we report that acoustic meron topological textures were successfully realized using designed Archimedeanlike square spiral metastructures via the excitation of spoof surface acoustic waves (SSAWs). By applying mirror-symmetric combinatorial operations to the unit structures, we further construct composite chiral metastructures that enable both one-dimensional and two-dimensional stable transport of acoustic bimerons. It is further revealed that bimeron transport originates from the locked opposite phase differences of SSAWs, induced by the handedness of the cavity resonant modes. The intrinsic robustness of the meron textures against structural defects is confirmed through the calculation of their topological charge. Our findings establish stable acoustic bimeron transport as a topologically resilient foundation for future acoustic information processing and storage technologies.

cond-mat.mtrl-sci↗

arXiv2025

Channel-Aware Multi-Domain Feature Extraction for Automatic Modulation Recognition in MIMO Systems

Automatic modulation recognition (AMR) is a key technology in non-cooperative communication systems, aiming to identify the modulation scheme from signals without prior information. Deep learning (DL)-based methods have gained wide attention due to their excellent performance, but research mainly focuses on single-input single-output (SISO) systems, with limited exploration for multiple-input multiple-output (MIMO) systems. The confounding effects of multi-antenna channels can interfere with the statistical properties of MIMO signals, making identification particularly challenging. To overcome these limitations, we propose a Channel-Aware Multi-Domain feature extraction (CAMD) framework for AMR in MIMO systems. Our CAMD framework reconstructs the transmitted signal through an efficient channel compensation module and achieves a more robust representation capability against channel interference by extracting and integrating multi-domain features, including intra-antenna temporal correlations and inter-antenna channel correlations. We have verified our method on the widely-used dataset, MIMOSig-Ref, with complex mobile channel environments. Extensive experiments confirm the performance advantages of CAMD over previous state-of-the-art methods.

eess.SP↗

arXiv2025

Microclimatic variation in tropical canopies: A glimpse into the processes of community assembly in epiphytic bryophyte communities

Epiphytic communities offer an original framework to disentangle the contributions of environmental filters, biotic interactions and dispersal limitations to community structure at fine spatial scales. We determine here whether variations in light, microclimatic conditions and host tree size affect the variation in species composition and phylogenetic structure of epiphytic bryophyte communities, and hence, assess the contribution of environmental filtering, phylogenetic constraints and competition to community assembly.A canopy crane giving access to 1.1 ha of tropical rainforest in Yunnan (China) was employed to record hourly light and microclimatic conditions from 54 dataloggers and epiphytic bryophyte communities from 408 plots. Generalized Dissimilarity Modelling was implemented to analyse the relationship between taxonomic and phylogenetic turnover among epiphytic communities, host-tree characteristics and microclimatic variation.Within-tree vertical turnover of bryophyte communities was significantly about 30% higher than horizontal turnover among-trees. Thus, the sharp vertical variations in microclimatic conditions from tree base to canopy are more important than differences in age, reflecting the likelihood of colonization, area, and habitat conditions between young and old trees, in shaping the composition of epiphytic bryophyte communities.

q-bio.PE↗

arXiv2025

Defects4C: Benchmarking Large Language Model Repair Capability with C/C++ Bugs

Automated Program Repair (APR) plays a critical role in enhancing the quality and reliability of software systems. While substantial progress has been made in Java-based APR, largely facilitated by benchmarks like Defects4J, there remains a significant gap in research on C/C++ program repair, despite the widespread use of C/C++ and the prevalence of associated vulnerabilities. This gap is primarily due to the lack of high-quality, open-source benchmarks tailored for C/C++. To address this issue, we introduce Defects4C, a comprehensive and executable benchmark specifically designed for C/C++ program repair. Our dataset is constructed from real-world C/C++ repositories and includes a large collection of bug-relevant commits (9M in total), 248 high-quality buggy functions, and 102 vulnerable functions, all paired with test cases for reproduction. These resources enable rigorous evaluation of repair techniques and support the retraining of learning-based approaches for enhanced performance. Using Defects4C, we conduct a comprehensive empirical study evaluating the effectiveness of 24 state-of-the-art large language models (LLMs) in repairing C/C++ faults. Our findings offer valuable insights into the strengths and limitations of current LLM-based APR techniques in this domain, highlighting both the need for more robust methods and the critical role of Defects4C in advancing future research

cs.SE↗