arXiv subjects

Kai Wang

Publications and source records attributed to Kai Wang.

At least 361 records · Page 20Linked to original sources

arXiv2025

A Multimodal Benchmark Dataset and Model for Crop Disease Diagnosis

While conversational generative AI has shown considerable potential in enhancing decision-making for agricultural professionals, its exploration has predominantly been anchored in text-based interactions. The evolution of multimodal conversational AI, leveraging vast amounts of image-text data from diverse sources, marks a significant stride forward. However, the application of such advanced vision-language models in the agricultural domain, particularly for crop disease diagnosis, remains underexplored. In this work, we present the crop disease domain multimodal (CDDM) dataset, a pioneering resource designed to advance the field of agricultural research through the application of multimodal learning techniques. The dataset comprises 137,000 images of various crop diseases, accompanied by 1 million question-answer pairs that span a broad spectrum of agricultural knowledge, from disease identification to management practices. By integrating visual and textual data, CDDM facilitates the development of sophisticated question-answering systems capable of providing precise, useful advice to farmers and agricultural professionals. We demonstrate the utility of the dataset by finetuning state-of-the-art multimodal models, showcasing significant improvements in crop disease diagnosis. Specifically, we employed a novel finetuning strategy that utilizes low-rank adaptation (LoRA) to finetune the visual encoder, adapter and language model simultaneously. Our contributions include not only the dataset but also a finetuning strategy and a benchmark to stimulate further research in agricultural technology, aiming to bridge the gap between advanced AI techniques and practical agricultural applications. The dataset is available at https: //github.com/UnicomAI/UnicomBenchmark/tree/main/CDDMBench.

cs.CV↗

arXiv2025

Symplectic tracking through curved three dimensional fields by a method of generating functions

Symplectic integrator plays a pivotal role in the long-term tracking of charged particles within accelerators. To get symplectic maps in accurate simulation of single-particle trajectories, two key components are addressed: precise analytical expressions for arbitrary electromagnetic fields and a robust treatment of the equations of motion. In a source-free region, the electromagnetic fields can be decomposed into harmonic functions, applicable to both scalar and vector potentials, encompassing both straight and curved reference trajectories. These harmonics are constructed according to the boundary surface's field data due to uniqueness theorem. Finding generating functions to meet the Hamilton-Jacobi equation via a perturbative ansatz, we derive symplectic maps that are independent of the expansion order. This method yields a direct mapping from initial to final coordinates in a large step, bypassing the transition of intermediate coordinates. Our developed particle-tracking algorithm translates the field harmonics into beam trajectories, offering a high-resolution integration method in accelerator simulations.

physics.acc-ph↗

arXiv2025

Electron spin dynamics guide cell motility

Diverse organisms exploit the geomagnetic field (GMF) for migration. Migrating birds employ an intrinsically quantum mechanical mechanism for detecting the geomagnetic field: absorption of a blue photon generates a radical pair whose two electrons precess at different rates in the magnetic field, thereby sensitizing cells to the direction of the GMF. In this work, using an in vitro injury model, we discovered a quantum-based mechanism of cellular migration. Specifically, we show that migrating cells detect the GMF via an optically activated, electron spin-based mechanism. Cell injury provokes acute emission of blue photons, and these photons sensitize muscle progenitor cells to the magnetic field. We show that the magnetosensitivity of muscle progenitor cells is (a) activated by blue light, but not by green or red light, and (b) disrupted by the application of an oscillatory field at the frequency corresponding to the energy of the electron-spin/magnetic field interaction. A comprehensive analysis of protein expression reveals that the ability of blue photons to promote cell motility is mediated by activation of calmodulin calcium sensors. Collectively, these data suggest that cells possess a light-dependent magnetic compass driven by electron spin dynamics.

physics.bio-ph↗

arXiv2025

Deformation theory and Koszul duality for Rota-Baxter systems

This paper investigates Rota-Baxter systems in the sense of Brzeziński from the perspective of operad theory. The minimal model of the Rota-Baxter system operad is constructed, equivalently a concrete construction of its Koszul dual homotopy cooperad is given. The concept of homotopy Rota-Baxter systems and the $L_\infty$-algebra that governs deformations of a Rota-Baxter system are derived from the Koszul dual homotopy cooperad. The notion of infinity-Yang-Baxter pairs is introduced, which is a higher-order generalization of the traditional Yang-Baxter pairs. It is shown that a homotopy Rota-Baxter system structure on the endomorphism algebra of a graded space is equivalent to an associative infinity-Yang-Baxter pair on this graded algebra, thereby generalizing the classical correspondence between Yang-Baxter pairs and Rota-Baxter systems.

math.RA↗

arXiv2025

Digital Player: Evaluating Large Language Models based Human-like Agent in Games

With the rapid advancement of Large Language Models (LLMs), LLM-based autonomous agents have shown the potential to function as digital employees, such as digital analysts, teachers, and programmers. In this paper, we develop an application-level testbed based on the open-source strategy game "Unciv", which has millions of active players, to enable researchers to build a "data flywheel" for studying human-like agents in the "digital players" task. This "Civilization"-like game features expansive decision-making spaces along with rich linguistic interactions such as diplomatic negotiations and acts of deception, posing significant challenges for LLM-based agents in terms of numerical reasoning and long-term planning. Another challenge for "digital players" is to generate human-like responses for social interaction, collaboration, and negotiation with human players. The open-source project can be found at https:/github.com/fuxiAIlab/CivAgent.

cs.LG↗

arXiv2025

Real-Time Video Generation with Pyramid Attention Broadcast

We present Pyramid Attention Broadcast (PAB), a real-time, high quality and training-free approach for DiT-based video generation. Our method is founded on the observation that attention difference in the diffusion process exhibits a U-shaped pattern, indicating significant redundancy. We mitigate this by broadcasting attention outputs to subsequent steps in a pyramid style. It applies different broadcast strategies to each attention based on their variance for best efficiency. We further introduce broadcast sequence parallel for more efficient distributed inference. PAB demonstrates up to 10.5x speedup across three models compared to baselines, achieving real-time generation for up to 720p videos. We anticipate that our simple yet effective method will serve as a robust baseline and facilitate future research and application for video generation.

cs.CV↗

arXiv2025

Enhance-A-Video: Better Generated Video for Free

DiT-based video generation has achieved remarkable results, but research into enhancing existing models remains relatively unexplored. In this work, we introduce a training-free approach to enhance the coherence and quality of DiT-based generated videos, named Enhance-A-Video. The core idea is enhancing the cross-frame correlations based on non-diagonal temporal attention distributions. Thanks to its simple design, our approach can be easily applied to most DiT-based video generation frameworks without any retraining or fine-tuning. Across various DiT-based video generation models, our approach demonstrates promising improvements in both temporal consistency and visual quality. We hope this research can inspire future explorations in video generation enhancement.

cs.CV↗

arXiv2025

BPS invariants from framed links

In this article, we investigate the BPS invariants associated with framed links. We extend the relationship between the algebraic curve (i.e. dual $A$-polynomial) and the BPS invariants of a knot investigated in \cite{GKS} to the case of a framed knot. With the help of the framing change formula for the dual $A$-polynomial of a framed knot, we give several explicit formulas for the extremal $A$-polynomials and the BPS invariants of framed knots. As to the framed links, we present several numerical calculations for the Ooguri-Vafa invariants and BPS invariants for framed Whitehead links and Borromean rings and verify the integrality property for them.

math.GT↗

arXiv2025

Testing galaxy formation models with the stellar mass-halo mass relations for star-forming and quiescent galaxies

The tight relationship between the stellar mass and halo mass of galaxies is one of the most fundamental scaling relations in galaxy formation and evolution. It has become a critical constraint for galaxy formation models. Over the past decade, growing evidence has convincingly shown that the stellar mass-halo mass relations (SHMRs) for star-forming and quiescent central galaxies differ significantly: at a given stellar mass, the average host halo mass of quiescent centrals is more massive than that of the star-forming centrals. Despite the importance of this feature, its scientific implications have not yet been fully recognized or thoroughly explored in the field. In this work, we demonstrate that the semi-analytical model L-GALAXIES successfully reproduces these observational results, whereas three state-of-the-art hydrodynamic galaxy formation simulations (TNG, Illustris, and EAGLE) do not. Consequently, in L-GALAXIES, star-forming central galaxies are more efficient at converting baryons into stars than quiescent central galaxies at a given halo mass, while the other models predict similar efficiencies for both populations. Further analysis reveals that these fundamental discrepancies stem from distinct evolutionary paths on the stellar mass-halo mass plane. We show that the observed SHMRs for star-forming and quiescent galaxies support galaxy formation models in which quenching only weakly correlates with halo assembly histories, and in which the stellar mass of star-forming galaxies can increase significantly since cosmic noon. In contrast, models in which quenching strongly prefers to happen in early-formed halos are not very favored. Additionally, we find that galaxy downsizing is present in L-GALAXIES and TNG, but absent in Illustris and EAGLE.

astro-ph.GA↗

arXiv2025

What's in a Query: Polarity-Aware Distribution-Based Fair Ranking

Machine learning-driven rankings, where individuals (or items) are ranked in response to a query, mediate search exposure or attention in a variety of safety-critical settings. Thus, it is important to ensure that such rankings are fair. Under the goal of equal opportunity, attention allocated to an individual on a ranking interface should be proportional to their relevance across search queries. In this work, we examine amortized fair ranking -- where relevance and attention are cumulated over a sequence of user queries to make fair ranking more feasible in practice. Unlike prior methods that operate on expected amortized attention for each individual, we define new divergence-based measures for attention distribution-based fairness in ranking (DistFaiR), characterizing unfairness as the divergence between the distribution of attention and relevance corresponding to an individual over time. This allows us to propose new definitions of unfairness, which are more reliable at test time. Second, we prove that group fairness is upper-bounded by individual fairness under this definition for a useful class of divergence measures, and experimentally show that maximizing individual fairness through an integer linear programming-based optimization is often beneficial to group fairness. Lastly, we find that prior research in amortized fair ranking ignores critical information about queries, potentially leading to a fairwashing risk in practice by making rankings appear more fair than they actually are.

cs.LG↗

arXiv2025

Lumina-Video: Efficient and Flexible Video Generation with Multi-scale Next-DiT

Recent advancements have established Diffusion Transformers (DiTs) as a dominant framework in generative modeling. Building on this success, Lumina-Next achieves exceptional performance in the generation of photorealistic images with Next-DiT. However, its potential for video generation remains largely untapped, with significant challenges in modeling the spatiotemporal complexity inherent to video data. To address this, we introduce Lumina-Video, a framework that leverages the strengths of Next-DiT while introducing tailored solutions for video synthesis. Lumina-Video incorporates a Multi-scale Next-DiT architecture, which jointly learns multiple patchifications to enhance both efficiency and flexibility. By incorporating the motion score as an explicit condition, Lumina-Video also enables direct control of generated videos' dynamic degree. Combined with a progressive training scheme with increasingly higher resolution and FPS, and a multi-source training scheme with mixed natural and synthetic data, Lumina-Video achieves remarkable aesthetic quality and motion smoothness at high training and inference efficiency. We additionally propose Lumina-V2A, a video-to-audio model based on Next-DiT, to create synchronized sounds for generated videos. Codes are released at https://www.github.com/Alpha-VLLM/Lumina-Video.

cs.CV↗

arXiv2025

Recurrent Diffusion for Large-Scale Parameter Generation

Parameter generation has long struggled to match the scale of today large vision and language models, curbing its broader utility. In this paper, we introduce Recurrent Diffusion for Large Scale Parameter Generation (RPG), a novel framework that generates full neural network parameters up to hundreds of millions on a single GPU. Our approach first partitions a networks parameters into non-overlapping tokens, each corresponding to a distinct portion of the model. A recurrent mechanism then learns the inter token relationships, producing prototypes which serve as conditions for a diffusion process that ultimately synthesizes the full parameters. Across a spectrum of architectures and tasks including ResNets, ConvNeXts and ViTs on ImageNet 1K and COCO, and even LoRA based LLMs RPG achieves performance on par with fully trained networks while avoiding excessive memory overhead. Notably, it generalizes beyond its training set to generate valid parameters for previously unseen tasks, highlighting its flexibility in dynamic and open ended scenarios. By overcoming the longstanding memory and scalability barriers, RPG serves as a critical advance in AI generating AI, potentially enabling efficient weight generation at scales previously deemed infeasible.

cs.LG↗

arXiv2025

Assessing Open-world Forgetting in Generative Image Model Customization

Recent advances in diffusion models have significantly enhanced image generation capabilities. However, customizing these models with new classes often leads to unintended consequences that compromise their reliability. We introduce the concept of open-world forgetting to characterize the vast scope of these unintended alterations. Our work presents the first systematic investigation into open-world forgetting in diffusion models, focusing on semantic and appearance drift of representations. Using zero-shot classification, we demonstrate that even minor model adaptations can lead to significant semantic drift affecting areas far beyond newly introduced concepts, with accuracy drops of up to 60% on previously learned concepts. Our analysis of appearance drift reveals substantial changes in texture and color distributions of generated content. To address these issues, we propose a functional regularization strategy that effectively preserves original capabilities while accommodating new concepts. Through extensive experiments across multiple datasets and evaluation metrics, we demonstrate that our approach significantly reduces both semantic and appearance drift. Our study highlights the importance of considering open-world forgetting in future research on model customization and finetuning methods.

cs.CV↗

arXiv2025

One-Prompt-One-Story: Free-Lunch Consistent Text-to-Image Generation Using a Single Prompt

Text-to-image generation models can create high-quality images from input prompts. However, they struggle to support the consistent generation of identity-preserving requirements for storytelling. Existing approaches to this problem typically require extensive training in large datasets or additional modifications to the original model architectures. This limits their applicability across different domains and diverse diffusion model configurations. In this paper, we first observe the inherent capability of language models, coined context consistency, to comprehend identity through context with a single prompt. Drawing inspiration from the inherent context consistency, we propose a novel training-free method for consistent text-to-image (T2I) generation, termed "One-Prompt-One-Story" (1Prompt1Story). Our approach 1Prompt1Story concatenates all prompts into a single input for T2I diffusion models, initially preserving character identities. We then refine the generation process using two novel techniques: Singular-Value Reweighting and Identity-Preserving Cross-Attention, ensuring better alignment with the input description for each frame. In our experiments, we compare our method against various existing consistent T2I generation approaches to demonstrate its effectiveness through quantitative metrics and qualitative assessments. Code is available at https://github.com/byliutao/1Prompt1Story.

cs.CV↗

arXiv2025

Common Neighborhood Estimation over Bipartite Graphs under Local Differential Privacy

Bipartite graphs, formed by two vertex layers, arise as a natural fit for modeling the relationships between two groups of entities. In bipartite graphs, common neighborhood computation between two vertices on the same vertex layer is a basic operator, which is easily solvable in general settings. However, it inevitably involves releasing the neighborhood information of vertices, posing a significant privacy risk for users in real-world applications. To protect edge privacy in bipartite graphs, in this paper, we study the problem of estimating the number of common neighbors of two vertices on the same layer under edge local differential privacy (edge LDP). The problem is challenging in the context of edge LDP since each vertex on the opposite layer of the query vertices can potentially be a common neighbor. To obtain efficient and accurate estimates, we propose a multiple-round framework that significantly reduces the candidate pool of common neighbors and enables the query vertices to construct unbiased estimators locally. Furthermore, we improve data utility by incorporating the estimators built from the neighbors of both query vertices and devise privacy budget allocation optimizations. These improve the estimator's robustness and consistency, particularly against query vertices with imbalanced degrees. Extensive experiments on 15 datasets validate the effectiveness and efficiency of our proposed techniques.

cs.DB↗

arXiv2025

SAMScore: A Content Structural Similarity Metric for Image Translation Evaluation

Image translation has wide applications, such as style transfer and modality conversion, usually aiming to generate images having both high degrees of realism and faithfulness. These problems remain difficult, especially when it is important to preserve content structures. Traditional image-level similarity metrics are of limited use, since the content structures of an image are high-level, and not strongly governed by pixel-wise faithfulness to an original image. To fill this gap, we introduce SAMScore, a generic content structural similarity metric for evaluating the faithfulness of image translation models. SAMScore is based on the recent high-performance Segment Anything Model (SAM), which allows content similarity comparisons with standout accuracy. We applied SAMScore on 19 image translation tasks, and found that it is able to outperform all other competitive metrics on all tasks. We envision that SAMScore will prove to be a valuable tool that will help to drive the vibrant field of image translation, by allowing for more precise evaluations of new and evolving translation models. The code is available at https://github.com/Kent0n-Li/SAMScore.

cs.CV↗

arXiv2025

From Halos to Galaxies. X: Decoding Galaxy SEDs with Physical Priors and Accurate Star Formation History Reconstruction

The spectral energy distribution (SED) of galaxies is essential for deriving fundamental properties like stellar mass and star formation history (SFH). However, conventional methods, including both parametric and non-parametric approaches, often fail to accurately recover the observed cosmic star formation rate (SFR) density due to oversimplified or unrealistic assumptions about SFH and their inability to account for the complex SFH variations across different galaxy populations. To address this issue, we introduce a novel approach that improves galaxy broadband SED analysis by incorporating physical priors derived from hydrodynamical simulations. Tests using IllustrisTNG simulations demonstrate that our method can reliably determine galaxy physical properties from broadband photometry, including stellar mass within 0.05 dex, current SFR within 0.3 dex, and fractional stellar formation time within 0.2 dex, with a negligible fraction of catastrophic failures. When applied to the Sloan Digital Sky Survey (SDSS) main photometric galaxy sample with spectroscopic redshift, our estimates of stellar mass and SFR are consistent with the widely used MPA-JHU and GSWLC catalogs. Notably, using the derived SFHs of individual SDSS galaxies, we estimate the cosmic SFR density and stellar mass density with remarkable consistency to direct observations up to $z \sim 6$. This demonstrates a significant advancement in deriving SFHs from SEDs that closely align with observational data. Consequently, our method can reliably recover observed spectral indices such as $\rm D_{\rm n}(4000)$ and $\rm Hδ_{\rm A}$ by synthesizing the full spectra of galaxies using the estimated SFHs and metal enrichment histories, relying solely on broadband photometry as input. Furthermore, this method is extremely computationally efficient compared to conventional approaches.

astro-ph.GA↗

arXiv2025

Ferroelectricity in undoped HfO2 down to one-unit-cell on Si substrate

Hafnium oxide (HfO2), particularly at low-dimensional scales, exhibits extensive promising applications in ultrahigh density devices like low-power logic and non-volatile memory devices due to its compatibility with current semiconductor technology1-5. However, achieving ferroelectricity (FE) at ultimate scale especially in undoped HfO2 remains challenging as the non-centrosymmetric FE phase, so-called O-III (space group: Pca21) is metastable and FE has a strong tendency of depolarization with the decrease in thickness6. Up to now, this phase has usually stabilized via doping with other elements7-9. But the minimum film thickness is still limited to 1 nm, about 2-unit-cell, to keep FE8. Thinner and undoped films, conducive to further miniature device size and avoid contamination during deposition process, have been a challenge to fabricate on Si substrates. Herein, we report the robust FE observed in undoped HfO2 ultrathin films directly grown on Si substrate via atomic layer deposition (ALD) and post-heat treat in vacuum. The so-fabricated ferroelectric O-III phase contains about 4.48 at% oxygen vacancy, is robust even monoclinic phase (space group: P21/c) coexists. The spontaneous and switchable polarization is remarkably stable, still surviving even in films down to 0.5 nm (one-unit-cell). Our results show the robust FE O-III phase can be obtained in films down to one-unit-cell in thickness on Si, providing a practical way to fabricating this important material in thickness limit.

cond-mat.mtrl-sci↗