id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-06-27 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.20930 | Chi-Sheng Chen | Chi-Sheng Chen, Xinyu Zhang, Ya-Chuan Chen | Quantum Reinforcement Learning Trading Agent for Sector Rotation in the Taiwan Stock Market | null | null | null | null | quant-ph cs.LG q-fin.CP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a hybrid quantum-classical reinforcement learning framework for sector rotation in the Taiwan stock market. Our system employs Proximal Policy Optimization (PPO) as the backbone algorithm and integrates both classical architectures (LSTM, Transformer) and quantum-enhanced models (QNN, QRWKV, QASA) as policy and value networks. An automated feature engineering pipeline extracts financial indicators from capital share data to ensure consistent model input across all configurations. Empirical backtesting reveals a key finding: although quantum-enhanced models consistently achieve higher training rewards, they underperform classical models in real-world investment metrics such as cumulative return and Sharpe ratio. This discrepancy highlights a core challenge in applying reinforcement learning to financial domains -- namely, the mismatch between proxy reward signals and true investment objectives. Our analysis suggests that current reward designs may incentivize overfitting to short-term volatility rather than optimizing risk-adjusted returns. This issue is compounded by the inherent expressiveness and optimization instability of quantum circuits under Noisy Intermediate-Scale Quantum (NISQ) constraints. We discuss the implications of this reward-performance gap and propose directions for future improvement, including reward shaping, model regularization, and validation-based early stopping. Our work offers a reproducible benchmark and critical insights into the practical challenges of deploying quantum reinforcement learning in real-world finance. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:29:19 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Chi-Sheng",
""
],
[
"Zhang",
"Xinyu",
""
],
[
"Chen",
"Ya-Chuan",
""
]
] |
2506.20931 | Jiale Zhang | Chengcheng Zhu, Ye Li, Bosen Rao, Jiale Zhang, Yunlong Mao, Sheng Zhong | SPA: Towards More Stealth and Persistent Backdoor Attacks in Federated Learning | 18 pages | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated Learning (FL) has emerged as a leading paradigm for privacy-preserving distributed machine learning, yet the distributed nature of FL introduces unique security challenges, notably the threat of backdoor attacks. Existing backdoor strategies predominantly rely on end-to-end label supervision, which, despite their efficacy, often results in detectable feature disentanglement and limited persistence. In this work, we propose a novel and stealthy backdoor attack framework, named SPA, which fundamentally departs from traditional approaches by leveraging feature-space alignment rather than direct trigger-label association. Specifically, SPA reduces representational distances between backdoor trigger features and target class features, enabling the global model to misclassify trigger-embedded inputs with high stealth and persistence. We further introduce an adaptive, adversarial trigger optimization mechanism, utilizing boundary-search in the feature space to enhance attack longevity and effectiveness, even against defensive FL scenarios and non-IID data distributions. Extensive experiments on various FL benchmarks demonstrate that SPA consistently achieves high attack success rates with minimal impact on model utility, maintains robustness under challenging participation and data heterogeneity conditions, and exhibits persistent backdoor effects far exceeding those of conventional techniques. Our results call urgent attention to the evolving sophistication of backdoor threats in FL and emphasize the pressing need for advanced, feature-level defense techniques. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:33:14 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhu",
"Chengcheng",
""
],
[
"Li",
"Ye",
""
],
[
"Rao",
"Bosen",
""
],
[
"Zhang",
"Jiale",
""
],
[
"Mao",
"Yunlong",
""
],
[
"Zhong",
"Sheng",
""
]
] |
2506.20932 | Gleb Smirnov | Gleb Smirnov, Roman Vershynin | Thinning to improve two-sample discrepancy | 7 pages | null | null | null | math.PR cs.DS | http://creativecommons.org/licenses/by/4.0/ | The discrepancy between two independent samples \(X_1,\dots,X_n\) and \(Y_1,\dots,Y_n\) drawn from the same distribution on $\mathbb{R}^d$ typically has order \(O(\sqrt{n})\) even in one dimension. We give a simple online algorithm that reduces the discrepancy to \(O(\log^{2d} n)\) by discarding a small fraction of the points. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:39:14 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Smirnov",
"Gleb",
""
],
[
"Vershynin",
"Roman",
""
]
] |
2506.20933 | Erik Jahn | Erik Jahn, Frederick Eberhardt, Leonard J. Schulman | Lower Bounds on the Size of Markov Equivalence Classes | null | null | null | null | stat.ML cs.LG math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | Causal discovery algorithms typically recover causal graphs only up to their Markov equivalence classes unless additional parametric assumptions are made. The sizes of these equivalence classes reflect the limits of what can be learned about the underlying causal graph from purely observational data. Under the assumptions of acyclicity, causal sufficiency, and a uniform model prior, Markov equivalence classes are known to be small on average. In this paper, we show that this is no longer the case when any of these assumptions is relaxed. Specifically, we prove exponentially large lower bounds for the expected size of Markov equivalence classes in three settings: sparse random directed acyclic graphs, uniformly random acyclic directed mixed graphs, and uniformly random directed cyclic graphs. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:44:23 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Jahn",
"Erik",
""
],
[
"Eberhardt",
"Frederick",
""
],
[
"Schulman",
"Leonard J.",
""
]
] |
2506.20934 | T. M. Dunster | T. M. Dunster | Simplified Airy function Asymptotic expansions for Reverse Generalised Bessel Polynomials | null | null | null | null | math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Uniform asymptotic expansions are derived for reverse generalised Bessel polynomials of large degree $n$, real parameter $a$, and complex argument $z$, which are simpler than previously known results. The defining differential equation is analysed; for large $n$ and $\frac{3}{2} - n < a < \infty$, it possesses two turning points in the complex $z$ plane. Away from the turning points Liouville-Green expansions are obtained for the polynomials and two companion solutions of the differential equation, where asymptotic series appear in the exponent. Then representations involving Airy functions and two slowly-varying coefficient functions are constructed. Using the Liouville-Green representations, asymptotic expansions are obtained for the coefficient functions that involve coefficients that can be easily and explicitly computed recursively. In conjunction with a suitable re-expansion, or Cauchy's integral formula, near the turning point, the expansions are valid for $\frac{3}{2} - n < a \leq \mathcal{O}(n)$, uniformly for all unbounded complex values of $z$. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:46:41 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Dunster",
"T. M.",
""
]
] |
2506.20935 | Hsin-Hsiung Huang | Hsin-Hsiung Huang and Hayden Hampton | Forecasting Geopolitical Events with a Sparse Temporal Fusion Transformer and Gaussian Process Hybrid: A Case Study in Middle Eastern and U.S. Conflict Dynamics | null | null | null | null | stat.ML cs.LG stat.AP stat.CO | http://creativecommons.org/licenses/by/4.0/ | Forecasting geopolitical conflict from data sources like the Global Database of Events, Language, and Tone (GDELT) is a critical challenge for national security. The inherent sparsity, burstiness, and overdispersion of such data cause standard deep learning models, including the Temporal Fusion Transformer (TFT), to produce unreliable long-horizon predictions. We introduce STFT-VNNGP, a hybrid architecture that won the 2023 Algorithms for Threat Detection (ATD) competition by overcoming these limitations. Designed to bridge this gap, our model employs a two-stage process: first, a TFT captures complex temporal dynamics to generate multi-quantile forecasts. These quantiles then serve as informed inputs for a Variational Nearest Neighbor Gaussian Process (VNNGP), which performs principled spatiotemporal smoothing and uncertainty quantification. In a case study forecasting conflict dynamics in the Middle East and the U.S., STFT-VNNGP consistently outperforms a standalone TFT, showing a superior ability to predict the timing and magnitude of bursty event periods, particularly at long-range horizons. This work offers a robust framework for generating more reliable and actionable intelligence from challenging event data, with all code and workflows made publicly available to ensure reproducibility. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:53:25 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Huang",
"Hsin-Hsiung",
""
],
[
"Hampton",
"Hayden",
""
]
] |
2506.20936 | Hao Zhang | Hao Zhang, Haolan Xu, Chun Feng, Varun Jampani, and Narendra Ahuja | PhysRig: Differentiable Physics-Based Skinning and Rigging Framework for Realistic Articulated Object Modeling | Accepted by ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Skinning and rigging are fundamental components in animation, articulated object reconstruction, motion transfer, and 4D generation. Existing approaches predominantly rely on Linear Blend Skinning (LBS), due to its simplicity and differentiability. However, LBS introduces artifacts such as volume loss and unnatural deformations, and it fails to model elastic materials like soft tissues, fur, and flexible appendages (e.g., elephant trunks, ears, and fatty tissues). In this work, we propose PhysRig: a differentiable physics-based skinning and rigging framework that overcomes these limitations by embedding the rigid skeleton into a volumetric representation (e.g., a tetrahedral mesh), which is simulated as a deformable soft-body structure driven by the animated skeleton. Our method leverages continuum mechanics and discretizes the object as particles embedded in an Eulerian background grid to ensure differentiability with respect to both material properties and skeletal motion. Additionally, we introduce material prototypes, significantly reducing the learning space while maintaining high expressiveness. To evaluate our framework, we construct a comprehensive synthetic dataset using meshes from Objaverse, The Amazing Animals Zoo, and MixaMo, covering diverse object categories and motion patterns. Our method consistently outperforms traditional LBS-based approaches, generating more realistic and physically plausible results. Furthermore, we demonstrate the applicability of our framework in the pose transfer task highlighting its versatility for articulated object modeling. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:58:09 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhang",
"Hao",
""
],
[
"Xu",
"Haolan",
""
],
[
"Feng",
"Chun",
""
],
[
"Jampani",
"Varun",
""
],
[
"Ahuja",
"Narendra",
""
]
] |
2506.20937 | Christopher Glein | Christopher R. Glein and Ngoc Truong | Phosphates Reveal High pH Ocean Water on Enceladus | Accepted by Icarus, 23 pages, 4 figures, 1 table | Icarus, 441, 116717 (2025) | 10.1016/j.icarus.2025.116717 | null | astro-ph.EP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Enceladus offers our best opportunity for exploring the chemistry of an ocean on another world. Here, we perform geochemical modeling to show how the distribution of phosphate species found in ice grains from Enceladus's plume provides a very straightforward constraint on the pH of the host solution. The ratio of HPO$_4$/PO$_4$ species serves as a pH indicator. We find evidence of moderately alkaline water (pH 10.1-11.6)--significantly more alkaline than current estimates (~8-9) of the pH of Enceladus's ocean. Nevertheless, the pH range from phosphates is consistent with the CO$_2$/H$_2$O ratio measured in the plume if CO$_2$ exsolves from ocean water according to its equilibrium solubility. A simple energy balance can be used to quantify volatile fractionation during gas transport inside Enceladus's tiger stripes; we deduce that ~83% of water vapor is removed as ice during transport between the liquid-vapor interface and where gases exit the subsurface. We also explore how CO$_2$ degassing may lead to an increase in the apparent pH of ocean water. We generate maps of allowed combinations of pH and dissolved inorganic carbon concentration of the source water for a wide range of scenarios. Our preferred interpretation, constrained by the observed heat flux, implies minimal CO$_2$ degassing from ocean water. Hence, the pH recorded by phosphates should closely approximate that of the ocean; our best estimate is pH ~10.6. Such a high pH seems to reflect a major role of silicates enriched in Na, Mg, or Fe(II) interacting extensively with ocean water. Silica nanoparticles would not form or would subsequently dissolve if the pH is too high (>10.5). The outgassing model presented here provides a new path to quantify the dissolved concentrations of volatile species. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 01:58:51 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Glein",
"Christopher R.",
""
],
[
"Truong",
"Ngoc",
""
]
] |
2506.20938 | Joshua Hoke Davis | Joshua H. Davis, Daniel Nichols, Ishan Khillan, Abhinav Bhatele | ParEval-Repo: A Benchmark Suite for Evaluating LLMs with Repository-level HPC Translation Tasks | 11 pages, 5 figures | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | GPGPU architectures have become significantly diverse in recent years, which has led to an emergence of a variety of specialized programming models and software stacks to support them. While portable execution models exist, they still require significant developer effort to port to and optimize for different hardware architectures. Recent advances in large language models (LLMs) can help us reduce some of this programmer burden. In this paper, we present a novel benchmark and testing framework, ParEval-Repo, which can be used to evaluate the efficacy of LLM-based approaches in automatically translating entire codebases across GPGPU execution models. ParEval-Repo includes several scientific computing and AI mini-applications in a range of programming models, and levels of repository complexity. We use ParEval-Repo to evaluate a range of state-of-the-art open-source and commercial LLMs, with both a non-agentic and a top-down agentic approach. We assess code generated by the LLMs and approaches in terms of compilability, functional correctness, categories of build errors, and the cost of translation in terms of the number of inference tokens. Our results demonstrate that LLM translation of scientific applications is feasible for small programs but difficulty with generating functional build systems and cross-file dependencies pose challenges in scaling to larger codebases. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:01:11 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Davis",
"Joshua H.",
""
],
[
"Nichols",
"Daniel",
""
],
[
"Khillan",
"Ishan",
""
],
[
"Bhatele",
"Abhinav",
""
]
] |
2506.20939 | Chad Mourning | Chad Mourning, Zhewei Wang, Justin Murray | AIR-VIEW: The Aviation Image Repository for Visibility Estimation of Weather, A Dataset and Benchmark | 5 pages, meant as citation for dataset | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Machine Learning for aviation weather is a growing area of research for providing low-cost alternatives for traditional, expensive weather sensors; however, in the area of atmospheric visibility estimation, publicly available datasets, tagged with visibility estimates, of distances relevant for aviation, of diverse locations, of sufficient size for use in supervised learning, are absent. This paper introduces a new dataset which represents the culmination of a year-long data collection campaign of images from the FAA weather camera network suitable for this purpose. We also present a benchmark when applying three commonly used approaches and a general-purpose baseline when trained and tested on three publicly available datasets, in addition to our own, when compared against a recently ratified ASTM standard. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:04:04 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Mourning",
"Chad",
""
],
[
"Wang",
"Zhewei",
""
],
[
"Murray",
"Justin",
""
]
] |
2506.20940 | Tao Li | Tao Li, Meng-Long Xiao, Xin-Fang Zhang | Two-dimensional greedy randomized Kaczmarz methods for solving large-scale linear systems | arXiv admin note: text overlap with arXiv:2506.16106 | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider a novel two-dimensional randomized Kaczmarz method and its improved version with simple random sampling, which chooses two active rows with probability proportional to the square of their cross-product-like constant, for solving large-scale linear systems. From the greedy selection strategy with grasping two larger entries of the residual vector at each iteration, we then devise a two-dimensional greedy randomized Kaczmarz method. To improve the above methods further, motivated by the semi-randomized Kaczmarz method and Chebyshev's law of large numbers, we propose a two-dimensional semi-randomized Kaczmarz method and its modified version with simple random sampling, which is particularly advantageous for big data problems. Theoretically, we prove that the proposed methods converge to the unique least-norm solution of the consistent linear systems. Numerical results on some practical applications illustrate the superiority of the proposed methods compared with some existing ones in terms of computing time. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:08:27 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Li",
"Tao",
""
],
[
"Xiao",
"Meng-Long",
""
],
[
"Zhang",
"Xin-Fang",
""
]
] |
2506.20941 | Keivan Rezaei | Keivan Rezaei, Mehrdad Saberi, Abhilasha Ravichander, Soheil Feizi | Model State Arithmetic for Machine Unlearning | Preprint. Work in progress | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language models are trained on massive corpora of web data, which may include private data, copyrighted material, factually inaccurate data, or data that degrades model performance. Eliminating the influence of such problematic datapoints through complete retraining -- by repeatedly pretraining the model on datasets that exclude these specific instances -- is computationally prohibitive. For this reason, unlearning algorithms have emerged that aim to eliminate the influence of particular datapoints, while otherwise preserving the model -- at a low computational cost. However, precisely estimating and undoing the influence of individual datapoints has proved to be challenging. In this work, we propose a new algorithm, MSA, for estimating and undoing the influence of datapoints -- by leveraging model checkpoints i.e. artifacts capturing model states at different stages of pretraining. Our experimental results demonstrate that MSA consistently outperforms existing machine unlearning algorithms across multiple benchmarks, models, and evaluation metrics, suggesting that MSA could be an effective approach towards more flexible large language models that are capable of data erasure. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:16:16 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Rezaei",
"Keivan",
""
],
[
"Saberi",
"Mehrdad",
""
],
[
"Ravichander",
"Abhilasha",
""
],
[
"Feizi",
"Soheil",
""
]
] |
2506.20942 | Yubo Jin | Yubo Jin, Jian-Shu Li, Dongwen Liu, Binyong Sun | Period relations for Rankin-Selberg convolutions for $\mathrm{GL}(n)\times\mathrm{GL}(n)$ | null | null | null | null | math.NT math.RT | http://creativecommons.org/licenses/by/4.0/ | In this paper, we study the special values of Rankin-Selberg L-functions as a continuation of [LLS24]. Utilizing the modular symbol approach, we prove the rationality and period relations for critical values of Rankin-Selberg L-functions for $\mathrm{GL}(n)\times\mathrm{GL}(n)$ over any number field that contains a CM field. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:18:34 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Jin",
"Yubo",
""
],
[
"Li",
"Jian-Shu",
""
],
[
"Liu",
"Dongwen",
""
],
[
"Sun",
"Binyong",
""
]
] |
2506.20943 | Shubin Yu | Shubin Yu, Chen Yang, Chun-Lei Tang | Normalized solutions for the NLS equation with mixed fractional Laplacians and combined nonlinearities | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We look for normalized solutions to the nonlinear Schr\"{o}dinger equation with mixed fractional Laplacians and combined nonlinearities $$ \left\{\begin{array}{ll} (-\Delta)^{s_{1}} u+(-\Delta)^{s_{2}} u=\lambda u+\mu |u|^{q-2}u+|u|^{p-2}u \ \text{in}\;{\mathbb{R}^{N}}, \\[0.1cm] \int_{\mathbb{R}^{N}}|u|^2\mathrm dx=a^2, \end{array} \right. $$ where $N\geq 2,\;0<s_2<s_1<1, \mu>0$ and $\lambda\in\mathbb R$ appears as an unknown Lagrange multiplier. We mainly focus on some special cases, including fractional Sobolev subcritical or critical exponent. More precisely, for $2<q<2+\frac{4s_2}{N}<2+\frac{4s_1}{N}<p<2_{s_1}^{\ast}:=\frac{2N}{N-2s_1}$, we prove that the above problem has at least two solutions: a ground state with negative energy and a solution of mountain pass type with positive energy. For $2<q<2+\frac{4s_2}{N}$ and $p=2_{s_1}^{\ast}$, we also obtain the existence of ground states. Our results extend some previous ones of Chergui et al. (Calc. Var. Partial Differ. Equ., 2023) and Luo et al. (Adv. Nonlinear Stud., 2022). | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:19:08 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yu",
"Shubin",
""
],
[
"Yang",
"Chen",
""
],
[
"Tang",
"Chun-Lei",
""
]
] |
2506.20944 | Long-Khanh Pham | Van-Hoang Phan, Long-Khanh Pham, Dang Vu, Anh-Duy Tran, Minh-Son Dao | E-FreeM2: Efficient Training-Free Multi-Scale and Cross-Modal News Verification via MLLMs | Accepted to AsiaCCS 2025 @ SCID | null | null | null | cs.MM cs.CR | http://creativecommons.org/licenses/by/4.0/ | The rapid spread of misinformation in mobile and wireless networks presents critical security challenges. This study introduces a training-free, retrieval-based multimodal fact verification system that leverages pretrained vision-language models and large language models for credibility assessment. By dynamically retrieving and cross-referencing trusted data sources, our approach mitigates vulnerabilities of traditional training-based models, such as adversarial attacks and data poisoning. Additionally, its lightweight design enables seamless edge device integration without extensive on-device processing. Experiments on two fact-checking benchmarks achieve SOTA results, confirming its effectiveness in misinformation detection and its robustness against various attack vectors, highlighting its potential to enhance security in mobile and wireless communication environments. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:20:45 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Phan",
"Van-Hoang",
""
],
[
"Pham",
"Long-Khanh",
""
],
[
"Vu",
"Dang",
""
],
[
"Tran",
"Anh-Duy",
""
],
[
"Dao",
"Minh-Son",
""
]
] |
2506.20945 | Rui Niu | Rui Niu, Weihao Wu, Jie Chen, Long Ma, Zhiyong Wu | A Multi-Stage Framework for Multimodal Controllable Speech Synthesis | Accepted by ICME2025 | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Controllable speech synthesis aims to control the style of generated speech using reference input, which can be of various modalities. Existing face-based methods struggle with robustness and generalization due to data quality constraints, while text prompt methods offer limited diversity and fine-grained control. Although multimodal approaches aim to integrate various modalities, their reliance on fully matched training data significantly constrains their performance and applicability. This paper proposes a 3-stage multimodal controllable speech synthesis framework to address these challenges. For face encoder, we use supervised learning and knowledge distillation to tackle generalization issues. Furthermore, the text encoder is trained on both text-face and text-speech data to enhance the diversity of the generated speech. Experimental results demonstrate that this method outperforms single-modal baseline methods in both face based and text prompt based speech synthesis, highlighting its effectiveness in generating high-quality speech. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:23:42 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Niu",
"Rui",
""
],
[
"Wu",
"Weihao",
""
],
[
"Chen",
"Jie",
""
],
[
"Ma",
"Long",
""
],
[
"Wu",
"Zhiyong",
""
]
] |
2506.20946 | Donggoo Kang | Donggoo Kang, Jangyeong Kim, Dasol Jeong, Junyoung Choi, Jeonga Wi, Hyunmin Lee, Joonho Gwon, Joonki Paik | Consistent Zero-shot 3D Texture Synthesis Using Geometry-aware Diffusion and Temporal Video Models | null | null | null | null | cs.GR cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | Current texture synthesis methods, which generate textures from fixed viewpoints, suffer from inconsistencies due to the lack of global context and geometric understanding. Meanwhile, recent advancements in video generation models have demonstrated remarkable success in achieving temporally consistent videos. In this paper, we introduce VideoTex, a novel framework for seamless texture synthesis that leverages video generation models to address both spatial and temporal inconsistencies in 3D textures. Our approach incorporates geometry-aware conditions, enabling precise utilization of 3D mesh structures. Additionally, we propose a structure-wise UV diffusion strategy, which enhances the generation of occluded areas by preserving semantic information, resulting in smoother and more coherent textures. VideoTex not only achieves smoother transitions across UV boundaries but also ensures high-quality, temporally stable textures across video frames. Extensive experiments demonstrate that VideoTex outperforms existing methods in texture fidelity, seam blending, and stability, paving the way for dynamic real-time applications that demand both visual quality and temporal coherence. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:25:16 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Kang",
"Donggoo",
""
],
[
"Kim",
"Jangyeong",
""
],
[
"Jeong",
"Dasol",
""
],
[
"Choi",
"Junyoung",
""
],
[
"Wi",
"Jeonga",
""
],
[
"Lee",
"Hyunmin",
""
],
[
"Gwon",
"Joonho",
""
],
[
"Paik",
"Joonki",
""
]
] |
2506.20947 | Dejie Yang | Dejie Yang and Zhu Xu and Xinjie Gao and Yang Liu | Hierarchical Sub-action Tree for Continuous Sign Language Recognition | null | null | null | null | cs.CV cs.MM | http://creativecommons.org/licenses/by/4.0/ | Continuous sign language recognition (CSLR) aims to transcribe untrimmed videos into glosses, which are typically textual words. Recent studies indicate that the lack of large datasets and precise annotations has become a bottleneck for CSLR due to insufficient training data. To address this, some works have developed cross-modal solutions to align visual and textual modalities. However, they typically extract textual features from glosses without fully utilizing their knowledge. In this paper, we propose the Hierarchical Sub-action Tree (HST), termed HST-CSLR, to efficiently combine gloss knowledge with visual representation learning. By incorporating gloss-specific knowledge from large language models, our approach leverages textual information more effectively. Specifically, we construct an HST for textual information representation, aligning visual and textual modalities step-by-step and benefiting from the tree structure to reduce computational complexity. Additionally, we impose a contrastive alignment enhancement to bridge the gap between the two modalities. Experiments on four datasets (PHOENIX-2014, PHOENIX-2014T, CSL-Daily, and Sign Language Gesture) demonstrate the effectiveness of our HST-CSLR. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:27:50 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yang",
"Dejie",
""
],
[
"Xu",
"Zhu",
""
],
[
"Gao",
"Xinjie",
""
],
[
"Liu",
"Yang",
""
]
] |
2506.20948 | Sunil Naik | Jean-Marc Deshouillers and Sunil Naik | On coprimality of consecutive elements in certain sequences | null | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The study of finding blocks of primes in certain arithmetic sequences is one of the classical problems in number theory. It is also very interesting to study blocks of consecutive elements in such sequences that are pairwise coprime. In this context, we show that if $f$ is a twice continuously differentiable real-valued function on $[1, \infty)$ such that $f''(x) \to 0$ as $x \to \infty$ and $\limsup_{x \to \infty} f'(x) = \infty$, then there exist arbitrarily long blocks of pairwise coprime consecutive elements in the sequence $(\lfloor f(n) \rfloor)_n$. This result refines the qualitative part of a recent result by the first author, Drmota and M\"{u}llner.
We also prove that there exists a subset $\mathcal{A} \subseteq \mathbb{N}$ having upper Banach density one such that for any two distinct integers $m, n \in \mathcal{A}$, the integers $\lfloor f(m) \rfloor$ and $\lfloor f(n) \rfloor$ are pairwise coprime. Further, we show that there exist arbitrarily long blocks of consecutive elements in the sequence $(\lfloor f(n) \rfloor)_n$ such that no two of them are pairwise coprime. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:28:02 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Deshouillers",
"Jean-Marc",
""
],
[
"Naik",
"Sunil",
""
]
] |
2506.20949 | Chenkai Sun | Chenkai Sun, Denghui Zhang, ChengXiang Zhai, Heng Ji | Beyond Reactive Safety: Risk-Aware LLM Alignment via Long-Horizon Simulation | null | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Given the growing influence of language model-based agents on high-stakes societal decisions, from public policy to healthcare, ensuring their beneficial impact requires understanding the far-reaching implications of their suggestions. We propose a proof-of-concept framework that projects how model-generated advice could propagate through societal systems on a macroscopic scale over time, enabling more robust alignment. To assess the long-term safety awareness of language models, we also introduce a dataset of 100 indirect harm scenarios, testing models' ability to foresee adverse, non-obvious outcomes from seemingly harmless user prompts. Our approach achieves not only over 20% improvement on the new dataset but also an average win rate exceeding 70% against strong baselines on existing safety benchmarks (AdvBench, SafeRLHF, WildGuardMix), suggesting a promising direction for safer agents. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:28:58 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Sun",
"Chenkai",
""
],
[
"Zhang",
"Denghui",
""
],
[
"Zhai",
"ChengXiang",
""
],
[
"Ji",
"Heng",
""
]
] |
2506.20950 | Porter Morgan | R. \.Inan\c{c} Baykur, Porter Morgan | On nonorientable $4$--manifolds | 36 pages, 14 figures | null | null | null | math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present several structural results on closed, nonorientable, smooth $4$--manifolds, extending analogous results and machinery for the orientable case. We prove the existence of simplified broken Lefschetz fibrations and simplified trisections on nonorientable $4$--manifolds, yielding descriptions of them via factorizations in mapping class groups of nonorientable surfaces. With these tools in hand, we classify low genera simplified broken Lefschetz fibrations on nonorientable $4$--manifolds. We also establish that every closed, smooth $4$--manifold is obtained by surgery along a link of tori in a connected sum of copies of $\mathbb{CP}^2$, $S^1 \times S^3$ and $S^1\widetilde{\times} S^3$. Our proofs make use of topological modifications of singularities, handlebody decompositions, and mapping classes of surfaces. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:33:07 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Baykur",
"R. İnanç",
""
],
[
"Morgan",
"Porter",
""
]
] |
2506.20951 | Amanda Broadmeadow | Amanda Broadmeadow (University of Maryland College Park), Marta Sewi{\l}o (NASA Goddard Space Flight Center and University of Maryland College Park), Lee Mundy (University of Maryland College Park), Roya Hamedani Golshan (I. Physikalisches Institut der Universit\"at zu K\"oln), Kazuki Tokuda (Kyushu University and Kagawa University), Thomas M\"oller (I. Physikalisches Institut der Universit\"at zu K\"oln), Remy Indebetouw (University of Virginia and National Radio Astronomy Observatory), Joana M. Oliveira (Keele University), Steven B. Charnley (NASA Goddard Space Flight Center), Jennifer Wiseman (NASA Goddard Space Flight Center), Naoto Harada (The University of Tokyo), Peter Schilke (I. Physikalisches Institut der Universit\"at zu K\"oln) | ALMA Observations of Molecular Complexity in the Large Magellanic Cloud: Probing the Star-forming Region N160 | Comments: 49 pages, 34 figures, 5 tables (including appendices); Accepted for publication in the Astrophysical Journal | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Hot cores are small ($\lesssim$0.1 pc), dense ($\geq$10$^6$ cm$^{-3}$), and hot ($>$100 K) regions around massive protostars and are one of the main production sites of complex organic molecules (COMs, $\geq6$ atoms, including carbon). The Large Magellanic Cloud (LMC) is an ideal place to study hot core and COM formation in an environment that is different from our Galaxy, though prior to this study there have only been nine detections of extragalactic hot cores (seven in the LMC and two in the Small Magellanic Cloud, SMC). Here, we report 1.2 mm continuum and molecular line observations with the Atacama Large Millimeter/submillimeter Array (ALMA) in the star-forming region N160 that we named N160A-mm. We identify six 1.2 mm continuum sources, four of which are associated with methanol (CH$_3$OH) emission. Another COM, methyl cyanide (CH$_3$CN) is associated with the brightest source, N160A-mmA, the most chemically rich source in the field. Using the XCLASS software, we perform spectral modeling to estimate rotational temperatures and total column densities of detected molecular species for four sources. Based on the temperature exceeding 100 K, small size, and high H$_2$ number density, we identify N160A-mmA as a hot core. We compare the molecular abundances of this newly detected hot core with those previously detected in the LMC and SMC, as well as with a sample of Galactic hot cores, and discuss the complex nature of N160A-mmA. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:34:50 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Broadmeadow",
"Amanda",
"",
"University of Maryland College Park"
],
[
"Sewiło",
"Marta",
"",
"NASA Goddard Space Flight Center and University of Maryland College Park"
],
[
"Mundy",
"Lee",
"",
"University of Maryland College Park"
],
[
"Golshan",
"Roya Hamedani",
"",
"I. Physikalisches Institut der Universität zu Köln"
],
[
"Tokuda",
"Kazuki",
"",
"Kyushu University and Kagawa University"
],
[
"Möller",
"Thomas",
"",
"I. Physikalisches Institut der Universität zu Köln"
],
[
"Indebetouw",
"Remy",
"",
"University of Virginia and National Radio Astronomy Observatory"
],
[
"Oliveira",
"Joana M.",
"",
"Keele University"
],
[
"Charnley",
"Steven B.",
"",
"NASA Goddard Space Flight Center"
],
[
"Wiseman",
"Jennifer",
"",
"NASA Goddard Space Flight Center"
],
[
"Harada",
"Naoto",
"",
"The University of Tokyo"
],
[
"Schilke",
"Peter",
"",
"I. Physikalisches Institut der Universität zu Köln"
]
] |
2506.20952 | Kyosuke Ishibashi | Kyosuke Ishibashi, Atsushi Saito, Zin Y. Tun, Lucas Ray, Megan C. Coram, Akihiro Sakurai, Allison M. Okamura, and Ko Yamamoto | Effect of Haptic Feedback on Avoidance Behavior and Visual Exploration in Dynamic VR Pedestrian Environment | null | null | null | null | cs.HC cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Human crowd simulation in virtual reality (VR) is a powerful tool with potential applications including emergency evacuation training and assessment of building layout. While haptic feedback in VR enhances immersive experience, its effect on walking behavior in dense and dynamic pedestrian flows is unknown. Through a user study, we investigated how haptic feedback changes user walking motion in crowded pedestrian flows in VR. The results indicate that haptic feedback changed users' collision avoidance movements, as measured by increased walking trajectory length and change in pelvis angle. The displacements of users' lateral position and pelvis angle were also increased in the instantaneous response to a collision with a non-player character (NPC), even when the NPC was inside the field of view. Haptic feedback also enhanced users' awareness and visual exploration when an NPC approached from the side and back. Furthermore, variation in walking speed was increased by the haptic feedback. These results suggested that the haptic feedback enhanced users' sensitivity to a collision in VR environment. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:37:46 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Ishibashi",
"Kyosuke",
""
],
[
"Saito",
"Atsushi",
""
],
[
"Tun",
"Zin Y.",
""
],
[
"Ray",
"Lucas",
""
],
[
"Coram",
"Megan C.",
""
],
[
"Sakurai",
"Akihiro",
""
],
[
"Okamura",
"Allison M.",
""
],
[
"Yamamoto",
"Ko",
""
]
] |
2506.20953 | Jhih-Hong Lyu | Jhih-Hong Lyu and Tai-Chia Lin | Asymptotic Analysis of Boundary Layer Solutions to Poisson-Boltzmann Type Equations in General Bounded Smooth Domains | 45 pages, 5 figures | null | null | null | math.AP math-ph math.CA math.MP | http://creativecommons.org/licenses/by/4.0/ | We study the boundary layer solution to singular perturbation problems involving Poisson-Boltzmann (PB) type equations with a small parameter $\epsilon$ in general bounded smooth domains (including multiply connected domains) under the Robin boundary condition. The PB type equations include the classical PB, modified PB and charge-conserving PB (CCPB) equations, which are mathematical models for the electric potential and ion distributions. The CCPB equations present particular analytical challenges due to their nonlocal nonlinearity introduced through integral terms enforcing charge conservation. Using the principal coordinate system, exponential-type estimates and the moving plane agruments, we rigorously prove asymptotic expansions of boundary layer solutions throughout the whole domain. The solution domain is partitioned into three characteristic regions based on the distance from the boundary: Region I, where the distance from the boundary is at most $T\sqrt\epsilon$, Region II, where the distance ranges between $T\sqrt\epsilon$ and $\epsilon^\beta$, and Region III, where the distance is at least $\epsilon^\beta$, for given parameters $T>0$ and $0<\beta<1/2$. In Region I, we derive second-order asymptotic formulas explicitly incorporating the effects of boundary mean curvature, while exponential decay estimates are established for Regions II and III. Furthermore, we obtain asymptotic expansions for key physical quantities, including the electric potential, electric field, total ionic charge density and total ionic charge, revealing how domain geometry regulates electrostatic interactions. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:38:18 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Lyu",
"Jhih-Hong",
""
],
[
"Lin",
"Tai-Chia",
""
]
] |
2506.20954 | Qingrui Zhang | Xueming Liu, Lin Li, Xiang Zhou, Qingrui Zhang, and Tianjiang Hu | Cooperative Circumnavigation for Multi-Quadrotor Systems via Onboard Sensing | 8 Pages, 7 figures. Accepted by RA-L | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A cooperative circumnavigation framework is proposed for multi-quadrotor systems to enclose and track a moving target without reliance on external localization systems. The distinct relationships between quadrotor-quadrotor and quadrotor-target interactions are evaluated using a heterogeneous perception strategy and corresponding state estimation algorithms. A modified Kalman filter is developed to fuse visual-inertial odometry with range measurements to enhance the accuracy of inter-quadrotor relative localization. An event-triggered distributed Kalman filter is designed to achieve robust target state estimation under visual occlusion by incorporating neighbor measurements and estimated inter-quadrotor relative positions. Using the estimation results, a cooperative circumnavigation controller is constructed, leveraging an oscillator-based autonomous formation flight strategy. We conduct extensive indoor and outdoor experiments to validate the efficiency of the proposed circumnavigation framework in occluded environments. Furthermore, a quadrotor failure experiment highlights the inherent fault tolerance property of the proposed framework, underscoring its potential for deployment in search-and-rescue operations. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:41:55 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Liu",
"Xueming",
""
],
[
"Li",
"Lin",
""
],
[
"Zhou",
"Xiang",
""
],
[
"Zhang",
"Qingrui",
""
],
[
"Hu",
"Tianjiang",
""
]
] |
2506.20955 | Xiaoding Shi | Yazhou Chen, Yi Peng, Xiaoding Shi, Xiaoping Wang | Phase Transition in Non-isentropic Compressible Immiscible Two-Phase Flow with van der Waals Equation of State | 33 pages, 2 figures | null | null | null | math.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This study establishes the global well-posedness of the compressible non-isentropic Navier-Stokes/Allen-Cahn system governed by the van der Waals equation of state $p(\rho,\theta)=- a\rho^2+\frac{R\theta\rho}{1-b\rho}$ and degenerate thermal conductivity $\kappa(\theta)=\tilde{\kappa}\theta^\beta$, where $p$, $\rho$ and $\theta$ are the pressure, the density and the temperature of the flow respectively, and $a,b,R,\tilde\kappa$ are positive constants related to the physical properties of the flow. Navier-Stokes/Allen-Cahn system models immiscible two-phase flow with diffusive interfaces, where the non-monotonic pressure-density relationship in the van der Waals equation drives gas-liquid phase transitions. By developing a refined $L^2$-energy framework, we prove the existence and uniqueness of global strong solutions to the one-dimensional Cauchy problem for non-vacuum and finite-temperature initial data, without imposing smallness restrictions on the initial conditions. The findings demonstrate that despite non-monotonic pressure inducing substantial density fluctuations and triggering phase transitions, all physical quantities remain bounded over finite time intervals. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:42:12 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Yazhou",
""
],
[
"Peng",
"Yi",
""
],
[
"Shi",
"Xiaoding",
""
],
[
"Wang",
"Xiaoping",
""
]
] |
2506.20956 | Sunil Naik | Jean-Marc Deshouillers and Sunil Naik | Coprimality of elements in regular sequences with polynomial growth | null | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The investigation of primes in certain arithmetic sequences is one of the fundamental problems in number theory and especially, finding blocks of distinct primes has gained a lot of attention in recent years. In this context, we prove the existence of long blocks of $k$-wise coprime elements in certain regular sequences. More precisely, we prove that for any positive integers $H \geq k \geq 2$ and for a real-valued $k$-times continuously differentiable function $f \in \mathcal{C}^k\left( [1, \infty)\right)$ satisfying $\lim_{x \to \infty} f^{(k)}(x) = 0$ and $\limsup_{x \to \infty} f^{(k-1)}(x) = \infty$, there exist infinitely many positive integers $n$ such that $$ \gcd\left( \lfloor f(n+i_1)\rfloor, \lfloor f(n+i_2)\rfloor, \cdots, \lfloor f(n+i_k)\rfloor \right) ~=~ 1 $$ for any integers $1 \leq i_1 < i_2 < \cdots < i_k \leq H$. Further, we show that there exists a subset $\mathcal{A} \subseteq \mathbb{N}$ having upper Banach density one such that $$ \gcd\left(\lfloor f(n_1) \rfloor, \lfloor f(n_2) \rfloor, \cdots, \lfloor f(n_k) \rfloor\right) ~=~ 1 $$ for any distinct integers $n_1, n_2, \cdots, n_k \in \mathcal{A}$. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:42:35 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Deshouillers",
"Jean-Marc",
""
],
[
"Naik",
"Sunil",
""
]
] |
2506.20957 | Patrick Chen | Jiameng Chen, Xiantao Cai, Jia Wu, Wenbin Hu | Antibody Design and Optimization with Multi-scale Equivariant Graph Diffusion Models for Accurate Complex Antigen Binding | 9 pages, 4 figures, accepted at IJCAI 2025 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Antibody design remains a critical challenge in therapeutic and diagnostic development, particularly for complex antigens with diverse binding interfaces. Current computational methods face two main limitations: (1) capturing geometric features while preserving symmetries, and (2) generalizing novel antigen interfaces. Despite recent advancements, these methods often fail to accurately capture molecular interactions and maintain structural integrity. To address these challenges, we propose \textbf{AbMEGD}, an end-to-end framework integrating \textbf{M}ulti-scale \textbf{E}quivariant \textbf{G}raph \textbf{D}iffusion for antibody sequence and structure co-design. Leveraging advanced geometric deep learning, AbMEGD combines atomic-level geometric features with residue-level embeddings, capturing local atomic details and global sequence-structure interactions. Its E(3)-equivariant diffusion method ensures geometric precision, computational efficiency, and robust generalizability for complex antigens. Furthermore, experiments using the SAbDab database demonstrate a 10.13\% increase in amino acid recovery, 3.32\% rise in improvement percentage, and a 0.062~\AA\ reduction in root mean square deviation within the critical CDR-H3 region compared to DiffAb, a leading antibody design model. These results highlight AbMEGD's ability to balance structural integrity with improved functionality, establishing a new benchmark for sequence-structure co-design and affinity optimization. The code is available at: https://github.com/Patrick221215/AbMEGD. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:45:38 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Jiameng",
""
],
[
"Cai",
"Xiantao",
""
],
[
"Wu",
"Jia",
""
],
[
"Hu",
"Wenbin",
""
]
] |
2506.20958 | Norimasa Sasabe | Y. Yamasaki, N. Sasabe, Y. Ishii, Y. Sekiguchi, A. Sumiyoshiya, Y. Tanimoto, Y. Kotani, T. Nakamura, H. Nomura | High-Throughput Mapping of Magnetic Properties via the on-the-fly XMCD spectroscopy in a Combinatorial Fe-Co-Ni Film | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-throughput X-ray magnetic circular dichroism (XMCD) spectroscopy was conducted on Fe-Co-Ni compositionally graded films to systematically analyze the variation of magnetic properties as a function of composition. On-the-fly XMCD measurement enabled rapid spectral acquisition. Measurement time was reduced approximately tenfold compared to conventional stepwise methods, while maintaining high precision. The obtained XMCD spectra were processed using the Savitzky-Golay denoising technique, and element-specific magnetic properties were extracted using XMCD sum rules. By mapping the orbital and spin magnetic moments across the composition gradient, we identified key regions exhibiting enhanced soft magnetic properties. This study demonstrates the effectiveness of high-throughput synchrotron-based spectroscopy for accelerating materials discovery and optimizing functional magnetic materials. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:45:42 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yamasaki",
"Y.",
""
],
[
"Sasabe",
"N.",
""
],
[
"Ishii",
"Y.",
""
],
[
"Sekiguchi",
"Y.",
""
],
[
"Sumiyoshiya",
"A.",
""
],
[
"Tanimoto",
"Y.",
""
],
[
"Kotani",
"Y.",
""
],
[
"Nakamura",
"T.",
""
],
[
"Nomura",
"H.",
""
]
] |
2506.20959 | Zhong-Ying Fan | Hong-Ming Cui and Zhong-Ying Fan | Analytical approach to criticality of AdS black holes | null | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | We establish a hidden symmetry between the specific volumes of the coexistent phases and hence develop an analytical approach to study criticality of AdS black holes. In particular, using the method, we solve the coexistence line exactly for a variety of black holes, including the charged AdS black hole (in the four and the five dimensions), the rotating AdS black hole, the Gauss-Bonnet black hole and the quantum BTZ black hole. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:51:16 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Cui",
"Hong-Ming",
""
],
[
"Fan",
"Zhong-Ying",
""
]
] |
2506.20960 | Yiman Zhang | Yiman Zhang, Ziheng Luo, Qiangyu Yan, Wei He, Borui Jiang, Xinghao Chen, Kai Han | OmniEval: A Benchmark for Evaluating Omni-modal Models with Visual, Auditory, and Textual Inputs | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | In this paper, we introduce OmniEval, a benchmark for evaluating omni-modality models like MiniCPM-O 2.6, which encompasses visual, auditory, and textual inputs. Compared with existing benchmarks, our OmniEval has several distinctive features: (i) Full-modal collaboration: We design evaluation tasks that highlight the strong coupling between audio and video, requiring models to effectively leverage the collaborative perception of all modalities; (ii) Diversity of videos: OmniEval includes 810 audio-visual synchronized videos, 285 Chinese videos and 525 English videos; (iii) Diversity and granularity of tasks: OmniEval contains 2617 question-answer pairs, comprising 1412 open-ended questions and 1205 multiple-choice questions. These questions are divided into 3 major task types and 12 sub-task types to achieve comprehensive evaluation. Among them, we introduce a more granular video localization task named Grounding. Then we conduct experiments on OmniEval with several omni-modality models. We hope that our OmniEval can provide a platform for evaluating the ability to construct and understand coherence from the context of all modalities. Codes and data could be found at https://omnieval.github.io/. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 02:54:24 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhang",
"Yiman",
""
],
[
"Luo",
"Ziheng",
""
],
[
"Yan",
"Qiangyu",
""
],
[
"He",
"Wei",
""
],
[
"Jiang",
"Borui",
""
],
[
"Chen",
"Xinghao",
""
],
[
"Han",
"Kai",
""
]
] |
2506.20961 | Jean Felipe | H. G. Fargnoli, J. C. C. Felipe, G. Gazzola | Ambiguities in the generation of CFJ-terms in a 5-dimensional extension of QED in one loop | 13 pages | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | In this paper, we investigate a Lorentz symmetry breaking extension of Quantum Electrodynamics (QED), incorporating 5-dimensional CPT-odd terms at the one-loop level. Employing an independent regularization method, we systematically separate the divergences into regularization-independent contributions and regularization-dependent (surface terms). We demonstrate that these surface terms, potentially contributing to the radiative generation of Carroll-Field-Jackiw (CFJ) terms, remain undetermined even when imposing the Ward-Takahashi identity. Our results are consistent with existing literature and emphasize the necessity of a careful and consistent treatment of regularization ambiguities in quantum field theories involving symmetry breaking extensions. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:00:18 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Fargnoli",
"H. G.",
""
],
[
"Felipe",
"J. C. C.",
""
],
[
"Gazzola",
"G.",
""
]
] |
2506.20962 | Kaifeng Shen | The STAR Collaboration | Measurement of inclusive $J/\psi$ production in Au+Au collisions at $\sqrt{s_\mathrm{NN}} = 54.4$ GeV at STAR | null | null | null | null | nucl-ex hep-ex | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This article presents measurements of inclusive $J/\psi$ production at midrapidity ($\left|y\right| <$ 1.0) in Au+Au collisions at $\sqrt{s_\mathrm{NN}} = 54.4$ GeV with the STAR detector at the Relativistic Heavy Ion Collider. A suppression of the $J/\psi$ yield, quantified using the nuclear modification factors ($R_{\mathrm{AA}}$, $R_{\mathrm{CP}}$), is observed with respect to the scaled production in $p$+$p$ collisions. The dependence of $R_{\mathrm{AA}}$ on collision centrality and $J/\psi$ transverse momentum is measured with improved precision compared to previous measurements at 39 and 62.4 GeV, while the centrality dependence of $R_{\mathrm{CP}}$ is measured and compared to the same results at 39, 62.4, and 200 GeV. In central collisions, no significant collision energy dependence of $R_{\mathrm{AA}}$ is found within uncertainties for collision energies between 17.3 and 200 GeV. Two transport model calculations that include dissociation and regeneration contributions are consistent with the experimental results within uncertainties. Although no significant collision energy dependence of the $J/\psi$ suppression in high energy heavy-ion collisions up to $\sqrt{s_\mathrm{NN}} = 200$ GeV is observed within uncertainties, the newly measured results at 54.4 GeV Au+Au collisions provide additional constraints on theoretical calculations of the hot medium evolution and cold nuclear matter effects. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:00:37 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"The STAR Collaboration",
"",
""
]
] |
2506.20963 | Zhengjun Huang | Fangyuan Zhang, Zhengjun Huang, Yingli Zhou, Qintian Guo, Zhixun Li, Wensheng Luo, Di Jiang, Yixiang Fang, Xiaofang Zhou | EraRAG: Efficient and Incremental Retrieval Augmented Generation for Growing Corpora | Under review | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph-based Retrieval-Augmented Generation (Graph-RAG) enhances large language models (LLMs) by structuring retrieval over an external corpus. However, existing approaches typically assume a static corpus, requiring expensive full-graph reconstruction whenever new documents arrive, limiting their scalability in dynamic, evolving environments. To address these limitations, we introduce EraRAG, a novel multi-layered Graph-RAG framework that supports efficient and scalable dynamic updates. Our method leverages hyperplane-based Locality-Sensitive Hashing (LSH) to partition and organize the original corpus into hierarchical graph structures, enabling efficient and localized insertions of new data without disrupting the existing topology. The design eliminates the need for retraining or costly recomputation while preserving high retrieval accuracy and low latency. Experiments on large-scale benchmarks demonstrate that EraRag achieves up to an order of magnitude reduction in update time and token consumption compared to existing Graph-RAG systems, while providing superior accuracy performance. This work offers a practical path forward for RAG systems that must operate over continually growing corpora, bridging the gap between retrieval efficiency and adaptability. Our code and data are available at https://github.com/EverM0re/EraRAG-Official. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:01:33 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhang",
"Fangyuan",
""
],
[
"Huang",
"Zhengjun",
""
],
[
"Zhou",
"Yingli",
""
],
[
"Guo",
"Qintian",
""
],
[
"Li",
"Zhixun",
""
],
[
"Luo",
"Wensheng",
""
],
[
"Jiang",
"Di",
""
],
[
"Fang",
"Yixiang",
""
],
[
"Zhou",
"Xiaofang",
""
]
] |
2506.20964 | Luca Weishaupt | Chengkuan Chen, Luca L. Weishaupt, Drew F. K. Williamson, Richard J. Chen, Tong Ding, Bowen Chen, Anurag Vaidya, Long Phi Le, Guillaume Jaume, Ming Y. Lu, and Faisal Mahmood | Evidence-based diagnostic reasoning with multi-agent copilot for human pathology | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Pathology is experiencing rapid digital transformation driven by whole-slide imaging and artificial intelligence (AI). While deep learning-based computational pathology has achieved notable success, traditional models primarily focus on image analysis without integrating natural language instruction or rich, text-based context. Current multimodal large language models (MLLMs) in computational pathology face limitations, including insufficient training data, inadequate support and evaluation for multi-image understanding, and a lack of autonomous, diagnostic reasoning capabilities. To address these limitations, we introduce PathChat+, a new MLLM specifically designed for human pathology, trained on over 1 million diverse, pathology-specific instruction samples and nearly 5.5 million question answer turns. Extensive evaluations across diverse pathology benchmarks demonstrated that PathChat+ substantially outperforms the prior PathChat copilot, as well as both state-of-the-art (SOTA) general-purpose and other pathology-specific models. Furthermore, we present SlideSeek, a reasoning-enabled multi-agent AI system leveraging PathChat+ to autonomously evaluate gigapixel whole-slide images (WSIs) through iterative, hierarchical diagnostic reasoning, reaching high accuracy on DDxBench, a challenging open-ended differential diagnosis benchmark, while also capable of generating visually grounded, humanly-interpretable summary reports. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:02:16 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Chengkuan",
""
],
[
"Weishaupt",
"Luca L.",
""
],
[
"Williamson",
"Drew F. K.",
""
],
[
"Chen",
"Richard J.",
""
],
[
"Ding",
"Tong",
""
],
[
"Chen",
"Bowen",
""
],
[
"Vaidya",
"Anurag",
""
],
[
"Le",
"Long Phi",
""
],
[
"Jaume",
"Guillaume",
""
],
[
"Lu",
"Ming Y.",
""
],
[
"Mahmood",
"Faisal",
""
]
] |
2506.20965 | Craig Wright PhD | Craig Steven Wright | Rational Miner Behaviour, Protocol Stability, and Time Preference: An Austrian and Game-Theoretic Analysis of Bitcoin's Incentive Environment | Approximately 10,770 words, 0 figure, 0 table. Submitted to The Quarterly Journal of Austrian Economics | null | null | null | econ.GN cs.CR cs.GT cs.NI q-fin.EC q-fin.GN | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper integrates Austrian capital theory with repeated game theory to examine strategic miner behaviour under different institutional conditions in blockchain systems. It shows that when protocol rules are mutable, effective time preference rises, undermining rational long-term planning and cooperative equilibria. Using formal game-theoretic analysis and Austrian economic principles, the paper demonstrates how mutable protocols shift miner incentives from productive investment to political rent-seeking and influence games. The original Bitcoin protocol is interpreted as an institutional anchor: a fixed rule-set enabling calculability and low time preference. Drawing on the work of Bohm-Bawerk, Mises, and Hayek, the argument is made that protocol immutability is essential for restoring strategic coherence, entrepreneurial confidence, and sustainable network equilibrium. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:04:21 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Wright",
"Craig Steven",
""
]
] |
2506.20966 | Aoqun Jin | Tian-Yu Xiang, Ao-Qun Jin, Xiao-Hu Zhou, Mei-Jiang Gui, Xiao-Liang Xie, Shi-Qi Liu, Shuang-Yi Wang, Sheng-Bin Duan, Fu-Chao Xie, Wen-Kai Wang, Si-Cheng Wang, Ling-Yun Li, Tian Tu, Zeng-Guang Hou | Parallels Between VLA Model Post-Training and Human Motor Learning: Progress, Challenges, and Trends | null | null | null | null | cs.RO cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision-language-action (VLA) models extend vision-language models (VLM) by integrating action generation modules for robotic manipulation. Leveraging strengths of VLM in vision perception and instruction understanding, VLA models exhibit promising generalization across diverse manipulation tasks. However, applications demanding high precision and accuracy reveal performance gaps without further adaptation. Evidence from multiple domains highlights the critical role of post-training to align foundational models with downstream applications, spurring extensive research on post-training VLA models. VLA model post-training aims to address the challenge of improving an embodiment's ability to interact with the environment for the given tasks, analogous to the process of humans motor skills acquisition. Accordingly, this paper reviews post-training strategies for VLA models through the lens of human motor learning, focusing on three dimensions: environments, embodiments, and tasks. A structured taxonomy is introduced aligned with human learning mechanisms: (1) enhancing environmental perception, (2) improving embodiment awareness, (3) deepening task comprehension, and (4) multi-component integration. Finally, key challenges and trends in post-training VLA models are identified, establishing a conceptual framework to guide future research. This work delivers both a comprehensive overview of current VLA model post-training methods from a human motor learning perspective and practical insights for VLA model development. (Project website: https://github.com/AoqunJin/Awesome-VLA-Post-Training) | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:06:57 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Xiang",
"Tian-Yu",
""
],
[
"Jin",
"Ao-Qun",
""
],
[
"Zhou",
"Xiao-Hu",
""
],
[
"Gui",
"Mei-Jiang",
""
],
[
"Xie",
"Xiao-Liang",
""
],
[
"Liu",
"Shi-Qi",
""
],
[
"Wang",
"Shuang-Yi",
""
],
[
"Duan",
"Sheng-Bin",
""
],
[
"Xie",
"Fu-Chao",
""
],
[
"Wang",
"Wen-Kai",
""
],
[
"Wang",
"Si-Cheng",
""
],
[
"Li",
"Ling-Yun",
""
],
[
"Tu",
"Tian",
""
],
[
"Hou",
"Zeng-Guang",
""
]
] |
2506.20967 | Lingling Cai | Lingling Cai, Kang Zhao, Hangjie Yuan, Xiang Wang, Yingya Zhang, Kejie Huang | DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing | Zero-shot video editing | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The advent of Video Diffusion Transformers (Video DiTs) marks a milestone in video generation. However, directly applying existing video editing methods to Video DiTs often incurs substantial computational overhead, due to resource-intensive attention modification or finetuning. To alleviate this problem, we present DFVEdit, an efficient zero-shot video editing method tailored for Video DiTs. DFVEdit eliminates the need for both attention modification and fine-tuning by directly operating on clean latents via flow transformation. To be more specific, we observe that editing and sampling can be unified under the continuous flow perspective. Building upon this foundation, we propose the Conditional Delta Flow Vector (CDFV) -- a theoretically unbiased estimation of DFV -- and integrate Implicit Cross Attention (ICA) guidance as well as Embedding Reinforcement (ER) to further enhance editing quality. DFVEdit excels in practical efficiency, offering at least 20x inference speed-up and 85\% memory reduction on Video DiTs compared to attention-engineering-based editing methods. Extensive quantitative and qualitative experiments demonstrate that DFVEdit can be seamlessly applied to popular Video DiTs (e.g., CogVideoX and Wan2.1), attaining state-of-the-art performance on structural fidelity, spatial-temporal consistency, and editing quality. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:10:13 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Cai",
"Lingling",
""
],
[
"Zhao",
"Kang",
""
],
[
"Yuan",
"Hangjie",
""
],
[
"Wang",
"Xiang",
""
],
[
"Zhang",
"Yingya",
""
],
[
"Huang",
"Kejie",
""
]
] |
2506.20968 | Yuan Ji Xu Dr. | Yuanji Xu, Huiyuan Zhang, Maoyuan Feng, Fuyang Tian | The electronic structures, magnetic transition and Fermi surface instability of room-temperature altermagnet KV$_{2}$Se$_{2}$O | null | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Altermagnetism has recently emerged as a distinct and fundamental class of magnetic order. Exploring its interplay with quantum phenomena such as unconventional superconductivity, density-wave instabilities, and many-body effects represents a compelling frontier. In this work, we theoretically confirm the presence of high-temperature metallic altermagnetism in KV$_2$Se$_2$O. We demonstrate that the anomalous metal-insulator-metal transition arises from a Lifshitz transition associated with Fermi surface reconstruction. The previously reported spin-density wave gap is found to lie below the Fermi level in our study and is now recognized to be attributed to the V-shaped density of states, originating from orbital-selective and sublattice-resolved half-metal-like behavior on a specific V atom. Furthermore, we identify the instability from the nesting of spin-momentum-locked two-dimensional Fermi surfaces, which induces the SDW state. These findings position KV$_2$Se$_2$O as a promising platform for investigating the interplay among altermagnetism, unconventional superconductivity, and density-wave order. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:13:33 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Xu",
"Yuanji",
""
],
[
"Zhang",
"Huiyuan",
""
],
[
"Feng",
"Maoyuan",
""
],
[
"Tian",
"Fuyang",
""
]
] |
2506.20969 | Parv Maheshwari | Shruti Bansal, Wenshan Wang, Yifei Liu, Parv Maheshwari | ThermalDiffusion: Visual-to-Thermal Image-to-Image Translation for Autonomous Navigation | Accepted at Thermal Infrared in Robotics (TIRO) Workshop, ICRA 2025 | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Autonomous systems rely on sensors to estimate the environment around them. However, cameras, LiDARs, and RADARs have their own limitations. In nighttime or degraded environments such as fog, mist, or dust, thermal cameras can provide valuable information regarding the presence of objects of interest due to their heat signature. They make it easy to identify humans and vehicles that are usually at higher temperatures compared to their surroundings. In this paper, we focus on the adaptation of thermal cameras for robotics and automation, where the biggest hurdle is the lack of data. Several multi-modal datasets are available for driving robotics research in tasks such as scene segmentation, object detection, and depth estimation, which are the cornerstone of autonomous systems. However, they are found to be lacking in thermal imagery. Our paper proposes a solution to augment these datasets with synthetic thermal data to enable widespread and rapid adaptation of thermal cameras. We explore the use of conditional diffusion models to convert existing RGB images to thermal images using self-attention to learn the thermal properties of real-world objects. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:18:22 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Bansal",
"Shruti",
""
],
[
"Wang",
"Wenshan",
""
],
[
"Liu",
"Yifei",
""
],
[
"Maheshwari",
"Parv",
""
]
] |
2506.20970 | Haijia Jin | Haijia Jin, Jun Wu, Weijie Yuan, Fan Liu, and Yuanhao Cui | Co-Design of Sensing, Communications, and Control for Low-Altitude Wireless Networks | null | null | null | null | eess.SP | http://creativecommons.org/licenses/by/4.0/ | The rapid advancement of Internet of Things (IoT) services and the evolution toward the sixth generation (6G) have positioned unmanned aerial vehicles (UAVs) as critical enablers of low-altitude wireless networks (LAWNs). This work investigates the co-design of integrated sensing, communication, and control ($\mathbf{SC^{2}}$) for multi-UAV cooperative systems with finite blocklength (FBL) transmission. In particular, the UAVs continuously monitor the state of the field robots and transmit their observations to the robot controller to ensure stable control while cooperating to localize an unknown sensing target (ST). To this end, a weighted optimization problem is first formulated by jointly considering the control and localization performance in terms of the linear quadratic regulator (LQR) cost and the determinant of the Fisher information matrix (FIM), respectively. The resultant problem, optimizing resource allocations, the UAVs' deployment positions, and multi-user scheduling, is non-convex. To circumvent this challenge, we first derive a closed-form expression of the LQR cost with respect to other variables. Subsequently, the non-convex optimization problem is decomposed into a series of sub-problems by leveraging the alternating optimization (AO) approach, in which the difference of convex functions (DC) programming and projected gradient descent (PGD) method are employed to obtain an efficient near-optimal solution. Furthermore, the convergence and computational complexity of the proposed algorithm are thoroughly analyzed. Extensive simulation results are presented to validate the effectiveness of our proposed approach compared to the benchmark schemes and reveal the trade-off between control and sensing performance. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:19:27 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Jin",
"Haijia",
""
],
[
"Wu",
"Jun",
""
],
[
"Yuan",
"Weijie",
""
],
[
"Liu",
"Fan",
""
],
[
"Cui",
"Yuanhao",
""
]
] |
2506.20971 | Shihui Feng | Shihui Feng, Huilin Zhang, Dragan Ga\v{s}evi\'c | Where is AIED Headed? Key Topics and Emerging Frontiers (2020-2024) | null | null | null | null | cs.SI | http://creativecommons.org/licenses/by/4.0/ | In this study, we analyze 2,398 research articles published between 2020 and 2024 across eight core venues related to the field of Artificial Intelligence in Education (AIED). Using a three-step knowledge co-occurrence network analysis, we analyze the knowledge structure of the field, the evolving knowledge clusters, and the emerging frontiers. Our findings reveal that AIED research remains strongly technically focused, with sustained themes such as intelligent tutoring systems, learning analytics, and natural language processing, alongside rising interest in large language models (LLMs) and generative artificial intelligence (GenAI). By tracking the bridging keywords over the past five years, we identify four emerging frontiers in AIED--LLMs, GenAI, multimodal learning analytics, and human-AI collaboration. The current research interests in GenAI are centered around GAI-driven personalization, self-regulated learning, feedback, assessment, motivation, and ethics.The key research interests and emerging frontiers in AIED reflect a growing emphasis on co-adaptive, human-centered AI for education. This study provides the first large-scale field-level mapping of AIED's transformation in the GenAI era and sheds light on the future research development and educational practices. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:24:30 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Feng",
"Shihui",
""
],
[
"Zhang",
"Huilin",
""
],
[
"Gašević",
"Dragan",
""
]
] |
2506.20972 | Wenze Li | Wenze Li | Wild Bootstrap Inference for Linear Regressions with Many Covariates | null | null | null | null | econ.EM | http://creativecommons.org/licenses/by/4.0/ | We propose a simple modification to the wild bootstrap procedure and establish its asymptotic validity for linear regression models with many covariates and heteroskedastic errors. Monte Carlo simulations show that the modified wild bootstrap has excellent finite sample performance compared with alternative methods that are based on standard normal critical values, especially when the sample size is small and/or the number of controls is of the same order of magnitude as the sample size. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:30:26 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Li",
"Wenze",
""
]
] |
2506.20973 | Ashwin Joy | Ankan Biswas, Amal Manoharan and Ashwin Joy | Topological Entropy of Stationary Turbulent Flows | null | null | null | null | physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Topological entropy serves as a viable candidate for quantifying mixing and complexity of a highly chaotic system. Particularly in turbulence, this is determined as the exponential stretching rate of a fluid material line that typically necessitates a Lagrangian description. We extend the recent work [A. Manoharan, S. Subramanian, and A. Joy, arXiv:2412.08996] to three dimensions, and present an exact Eulerian framework for the topological entropy in stationary turbulent flows that is devoid of free parameters. The only prerequisite is a distribution of eigenvalues of the local strain-rate tensor and their decorrelation times. This can be easily obtained from a single wire probe at a fixed location, thereby eliminating the need for Lagrangian particle tracking, which remains formidable due to chaotic nature of the flow. We believe that our results lend great utility in experiments aimed at understanding transport and mixing in many industrial and natural flows. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:30:46 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Biswas",
"Ankan",
""
],
[
"Manoharan",
"Amal",
""
],
[
"Joy",
"Ashwin",
""
]
] |
2506.20974 | Zhejia Dong | Zhejia Dong, Corwin Zigler and Youjin Lee | Disentangling network dependence among multiple variables | null | null | null | null | stat.ME | http://creativecommons.org/licenses/by/4.0/ | When two variables depend on the same or similar underlying network, their shared network dependence structure can lead to spurious associations. While statistical associations between two variables sampled from interconnected subjects are a common inferential goal across various fields, little research has focused on how to disentangle shared dependence for valid statistical inference. We revisit two different approaches from distinct fields that may address shared network dependence: the pre-whitening approach, commonly used in time series analysis to remove the shared temporal dependence, and the network autocorrelation model, widely used in network analysis often to examine or account for autocorrelation of the outcome variable. We demonstrate how each approach implicitly entails assumptions about how a variable of interest propagates among nodes via network ties given the network structure. We further propose adaptations of existing pre-whitening methods to the network setting by explicitly reflecting underlying assumptions about "level of interaction" that induce network dependence, while accounting for its unique complexities. Our simulation studies demonstrate the effectiveness of the two approaches in reducing spurious associations due to shared network dependence when their respective assumptions hold. However, the results also show the sensitivity to assumption violations, underscoring the importance of correctly specifying the shared dependence structure based on available network information and prior knowledge about the interactions driving dependence. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:34:08 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Dong",
"Zhejia",
""
],
[
"Zigler",
"Corwin",
""
],
[
"Lee",
"Youjin",
""
]
] |
2506.20975 | Stephen L. Olsen | Stephen Lars Olsen | You never have enough J/$\psi$ events: the case for a J/$\psi$ factory | A write up of a talk presented at the 50 Years Discovery of the J Particle Symposium in Beijing in October 2024 | null | null | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | In a talk at an IHEP-Beijing symposium celebrating the 50th anniversary of the discovery of the J/$\psi$, I reminisced about some of the interesting phenomena that were observed in J/$\psi$ decays in the BES, BESII, and BESIII detectors during the 35 year long BES experimental program. The three order of magnitude increase in the J/$\psi$ event samples and the improved detector capabilities that occurred during this time led to a persistent series of interesting discoveries and new insights. As more J/$\psi$ events were collected, the scientific interest in them increased and the breadth of physics topics that are addressed by them expanded. In addition, I speculated on what might happen if, in the next few decades, the magnitude of J/$\psi$ data samples and the detector capabilities continued to increase at similar paces. In particular, I emphasize the possibilities of searches for CP violation in strange hyperon decays and testing the CPT~theorem with strangeness-tagged neutral kaon decays. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:34:20 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Olsen",
"Stephen Lars",
""
]
] |
2506.20976 | Aida Abiad | Aida Abiad, Luuk Reijnders | Eigenvalue bounds for distance-edge colorings | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | For a fixed positive integer $t$, we consider the graph colouring problem in which edges at distance at most $t$ are given distinct colours. We obtain sharp lower bounds for the distance-$t$ chromatic index, the least number of colours necessary for such a colouring. Our bounds are of algebraic nature; they depend on the eigenvalues of the line graph and on a polynomial which can be found using integer linear programming methods. We show several graph classes that attain equality for our bounds, and also present some computational results which illustrate the bound's performance. Lastly, we investigate the implications the spectral approach has to the Erd\H{o}s-Ne\v{s}et\v{r}il conjecture, and derive some conditions which a graph must satisfy if we could use it to obtain a counter example through the proposed spectral methods. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:38:41 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Abiad",
"Aida",
""
],
[
"Reijnders",
"Luuk",
""
]
] |
2506.20977 | Tao Liu | Tao Liu, Dafeng Zhang, Gengchen Li, Shizhuo Liu, Yongqi Song, Senmao Li, Shiqi Yang, Boqian Li, Kai Wang, Yaxing Wang | From Cradle to Cane: A Two-Pass Framework for High-Fidelity Lifespan Face Aging | 30 pages, 12 figures | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Face aging has become a crucial task in computer vision, with applications ranging from entertainment to healthcare. However, existing methods struggle with achieving a realistic and seamless transformation across the entire lifespan, especially when handling large age gaps or extreme head poses. The core challenge lies in balancing age accuracy and identity preservation--what we refer to as the Age-ID trade-off. Most prior methods either prioritize age transformation at the expense of identity consistency or vice versa. In this work, we address this issue by proposing a two-pass face aging framework, named Cradle2Cane, based on few-step text-to-image (T2I) diffusion models. The first pass focuses on solving age accuracy by introducing an adaptive noise injection (AdaNI) mechanism. This mechanism is guided by including prompt descriptions of age and gender for the given person as the textual condition. Also, by adjusting the noise level, we can control the strength of aging while allowing more flexibility in transforming the face. However, identity preservation is weakly ensured here to facilitate stronger age transformations. In the second pass, we enhance identity preservation while maintaining age-specific features by conditioning the model on two identity-aware embeddings (IDEmb): SVR-ArcFace and Rotate-CLIP. This pass allows for denoising the transformed image from the first pass, ensuring stronger identity preservation without compromising the aging accuracy. Both passes are jointly trained in an end-to-end way. Extensive experiments on the CelebA-HQ test dataset, evaluated through Face++ and Qwen-VL protocols, show that our Cradle2Cane outperforms existing face aging methods in age accuracy and identity consistency. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:48:28 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Liu",
"Tao",
""
],
[
"Zhang",
"Dafeng",
""
],
[
"Li",
"Gengchen",
""
],
[
"Liu",
"Shizhuo",
""
],
[
"Song",
"Yongqi",
""
],
[
"Li",
"Senmao",
""
],
[
"Yang",
"Shiqi",
""
],
[
"Li",
"Boqian",
""
],
[
"Wang",
"Kai",
""
],
[
"Wang",
"Yaxing",
""
]
] |
2506.20978 | Naihe Feng | Naihe Feng, Yi Sui, Shiyi Hou, Jesse C. Cresswell, Ga Wu | Response Quality Assessment for Retrieval-Augmented Generation via Conditional Conformal Factuality | Accepted by SIGIR 2025 short paper, 5 pages, Code is available at https://github.com/n4feng/ResponseQualityAssessment | null | 10.1145/3726302.3730244 | null | cs.IR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Existing research on Retrieval-Augmented Generation (RAG) primarily focuses on improving overall question-answering accuracy, often overlooking the quality of sub-claims within generated responses. Recent methods that attempt to improve RAG trustworthiness, such as through auto-evaluation metrics, lack probabilistic guarantees or require ground truth answers. To address these limitations, we propose Conformal-RAG, a novel framework inspired by recent applications of conformal prediction (CP) on large language models (LLMs). Conformal-RAG leverages CP and internal information from the RAG mechanism to offer statistical guarantees on response quality. It ensures group-conditional coverage spanning multiple sub-domains without requiring manual labelling of conformal sets, making it suitable for complex RAG applications. Compared to existing RAG auto-evaluation methods, Conformal-RAG offers statistical guarantees on the quality of refined sub-claims, ensuring response reliability without the need for ground truth answers. Additionally, our experiments demonstrate that by leveraging information from the RAG system, Conformal-RAG retains up to 60\% more high-quality sub-claims from the response compared to direct applications of CP to LLMs, while maintaining the same reliability guarantee. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:52:56 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Feng",
"Naihe",
""
],
[
"Sui",
"Yi",
""
],
[
"Hou",
"Shiyi",
""
],
[
"Cresswell",
"Jesse C.",
""
],
[
"Wu",
"Ga",
""
]
] |
2506.20979 | Weichen Dai | Weichen Dai, Kangcheng Ma, Jiaxin Wang, Kecen Pan, Yuhang Ming, Hua Zhang, Wanzeng Kong | 3D Scene-Camera Representation with Joint Camera Photometric Optimization | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Representing scenes from multi-view images is a crucial task in computer vision with extensive applications. However, inherent photometric distortions in the camera imaging can significantly degrade image quality. Without accounting for these distortions, the 3D scene representation may inadvertently incorporate erroneous information unrelated to the scene, diminishing the quality of the representation. In this paper, we propose a novel 3D scene-camera representation with joint camera photometric optimization. By introducing internal and external photometric model, we propose a full photometric model and corresponding camera representation. Based on simultaneously optimizing the parameters of the camera representation, the proposed method effectively separates scene-unrelated information from the 3D scene representation. Additionally, during the optimization of the photometric parameters, we introduce a depth regularization to prevent the 3D scene representation from fitting scene-unrelated information. By incorporating the camera model as part of the mapping process, the proposed method constructs a complete map that includes both the scene radiance field and the camera photometric model. Experimental results demonstrate that the proposed method can achieve high-quality 3D scene representations, even under conditions of imaging degradation, such as vignetting and dirt. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:53:38 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Dai",
"Weichen",
""
],
[
"Ma",
"Kangcheng",
""
],
[
"Wang",
"Jiaxin",
""
],
[
"Pan",
"Kecen",
""
],
[
"Ming",
"Yuhang",
""
],
[
"Zhang",
"Hua",
""
],
[
"Kong",
"Wanzeng",
""
]
] |
2506.20980 | Ziyu Zheng | Ziyu Zheng, Yaming Yang, Ziyu Guan, Wei Zhao, Weigang Lu | Enhancing Homophily-Heterophily Separation: Relation-Aware Learning in Heterogeneous Graphs | accepted by KDD 2025 | null | null | null | cs.SI cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-world networks usually have a property of node heterophily, that is, the connected nodes usually have different features or different labels. This heterophily issue has been extensively studied in homogeneous graphs but remains under-explored in heterogeneous graphs, where there are multiple types of nodes and edges. Capturing node heterophily in heterogeneous graphs is very challenging since both node/edge heterogeneity and node heterophily should be carefully taken into consideration. Existing methods typically convert heterogeneous graphs into homogeneous ones to learn node heterophily, which will inevitably lose the potential heterophily conveyed by heterogeneous relations. To bridge this gap, we propose Relation-Aware Separation of Homophily and Heterophily (RASH), a novel contrastive learning framework that explicitly models high-order semantics of heterogeneous interactions and adaptively separates homophilic and heterophilic patterns. Particularly, RASH introduces dual heterogeneous hypergraphs to encode multi-relational bipartite subgraphs and dynamically constructs homophilic graphs and heterophilic graphs based on relation importance. A multi-relation contrastive loss is designed to align heterogeneous and homophilic/heterophilic views by maximizing mutual information. In this way, RASH simultaneously resolves the challenges of heterogeneity and heterophily in heterogeneous graphs. Extensive experiments on benchmark datasets demonstrate the effectiveness of RASH across various downstream tasks. The code is available at: https://github.com/zhengziyu77/RASH. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:54:06 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zheng",
"Ziyu",
""
],
[
"Yang",
"Yaming",
""
],
[
"Guan",
"Ziyu",
""
],
[
"Zhao",
"Wei",
""
],
[
"Lu",
"Weigang",
""
]
] |
2506.20981 | Wen-Jie Lu | Jian Du, Haohao Qian, Shikun Zhang, Wen-jie Lu, Donghang Lu, Yongchuan Niu, Bo Jiang, Yongjun Zhao, and Qiang Yan | PrivacyGo: Privacy-Preserving Ad Measurement with Multidimensional Intersection | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | This paper tackles the challenging and practical problem of multi-identifier private user profile matching for privacy-preserving ad measurement, a cornerstone of modern advertising analytics. We introduce a comprehensive cryptographic framework leveraging reversed Oblivious Pseudorandom Functions (OPRF) and novel blind key rotation techniques to support secure matching across multiple identifiers. Our design prevents cross-identifier linkages and includes a differentially private mechanism to obfuscate intersection sizes, mitigating risks such as membership inference attacks.
We present a concrete construction of our protocol that achieves both strong privacy guarantees and high efficiency. It scales to large datasets, offering a practical and scalable solution for privacy-centric applications like secure ad conversion tracking. By combining rigorous cryptographic principles with differential privacy, our work addresses a critical need in the advertising industry, setting a new standard for privacy-preserving ad measurement frameworks. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:54:19 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Du",
"Jian",
""
],
[
"Qian",
"Haohao",
""
],
[
"Zhang",
"Shikun",
""
],
[
"Lu",
"Wen-jie",
""
],
[
"Lu",
"Donghang",
""
],
[
"Niu",
"Yongchuan",
""
],
[
"Jiang",
"Bo",
""
],
[
"Zhao",
"Yongjun",
""
],
[
"Yan",
"Qiang",
""
]
] |
2506.20982 | Martin Ruskov | Martin Ruskov | Our Coding Adventure: Using LLMs to Personalise the Narrative of a Tangible Programming Robot for Preschoolers | accepted at D-SAIL Workshop - Transformative Curriculum Design: Digitalization, Sustainability, and AI Literacy for 21st Century Learning | null | null | null | cs.CY cs.RO | http://creativecommons.org/licenses/by/4.0/ | Finding balanced ways to employ Large Language Models (LLMs) in education is a challenge due to inherent risks of poor understanding of the technology and of a susceptible audience. This is particularly so with younger children, who are known to have difficulties with pervasive screen time. Working with a tangible programming robot called Cubetto, we propose an approach to benefit from the capabilities of LLMs by employing such models in the preparation of personalised storytelling, necessary for preschool children to get accustomed to the practice of commanding the robot. We engage in action research to develop an early version of a formalised process to rapidly prototype game stories for Cubetto. Our approach has both reproducible results, because it employs open weight models, and is model-agnostic, because we test it with 5 different LLMs. We document on one hand the process, the used materials and prompts, and on the other the learning experience and outcomes. We deem the generation successful for the intended purposes of using the results as a teacher aid. Testing the models on 4 different task scenarios, we encounter issues of consistency and hallucinations and document the corresponding evaluation process and attempts (some successful and some not) to overcome these issues. Importantly, the process does not expose children to LLMs directly. Rather, the technology is used to help teachers easily develop personalised narratives on children's preferred topics. We believe our method is adequate for preschool classes and we are planning to further experiment in real-world educational settings. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:54:25 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Ruskov",
"Martin",
""
]
] |
2506.20983 | Wenjie Xuan | Wenjie Xuan, Jing Zhang, Juhua Liu, Bo Du, Dacheng Tao | Rethink Sparse Signals for Pose-guided Text-to-image Generation | accepted by ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent works favored dense signals (e.g., depth, DensePose), as an alternative to sparse signals (e.g., OpenPose), to provide detailed spatial guidance for pose-guided text-to-image generation. However, dense representations raised new challenges, including editing difficulties and potential inconsistencies with textual prompts. This fact motivates us to revisit sparse signals for pose guidance, owing to their simplicity and shape-agnostic nature, which remains underexplored. This paper proposes a novel Spatial-Pose ControlNet(SP-Ctrl), equipping sparse signals with robust controllability for pose-guided image generation. Specifically, we extend OpenPose to a learnable spatial representation, making keypoint embeddings discriminative and expressive. Additionally, we introduce keypoint concept learning, which encourages keypoint tokens to attend to the spatial positions of each keypoint, thus improving pose alignment. Experiments on animal- and human-centric image generation tasks demonstrate that our method outperforms recent spatially controllable T2I generation approaches under sparse-pose guidance and even matches the performance of dense signal-based methods. Moreover, SP-Ctrl shows promising capabilities in diverse and cross-species generation through sparse signals. Codes will be available at https://github.com/DREAMXFAR/SP-Ctrl. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:57:44 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Xuan",
"Wenjie",
""
],
[
"Zhang",
"Jing",
""
],
[
"Liu",
"Juhua",
""
],
[
"Du",
"Bo",
""
],
[
"Tao",
"Dacheng",
""
]
] |
2506.20984 | Lan Bo | Lan Bo, Songli Dai, Xichao Zhang, Masahito Mochizuki, Xiaohong Xu, Zean Tian, and Yan Zhou | All-electric control of skyrmion-bimeron transition in van der Waals heterostructures | 11 pages, 5 figures | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Two-dimensional van der Waals materials offer a versatile platform for manipulating atomic-scale topological spin textures. In this study, using first-principles and micromagnetic calculations, we demonstrate a reversible transition between magnetic skyrmions and bimerons in a MoTeI/In_2Se_3 multiferroic heterostructure. The physical origin lies in the reorientation of the easy axis of magnetic anisotropy, triggered by the reversal of ferroelectric polarization. We show that the transition operates effectively under both static and dynamic conditions, exhibiting remarkable stability and flexibility. Notably, this transition can be achieved entirely through electric control, without requiring any external magnetic field. Furthermore, we propose a binary encoding scheme based on the skyrmion-bimeron transition, presenting a promising path toward energy-efficient spintronic applications. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:58:36 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Bo",
"Lan",
""
],
[
"Dai",
"Songli",
""
],
[
"Zhang",
"Xichao",
""
],
[
"Mochizuki",
"Masahito",
""
],
[
"Xu",
"Xiaohong",
""
],
[
"Tian",
"Zean",
""
],
[
"Zhou",
"Yan",
""
]
] |
2506.20985 | Scott Wesley | Neil J. Ross and Scott Wesley | Cutoff Theorems for the Equivalence of Parameterized Quantum Circuits (Extended) | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Many promising quantum algorithms in economics, medical science, and material science rely on circuits that are parameterized by a large number of angles. To ensure that these algorithms are efficient, these parameterized circuits must be heavily optimized. However, most quantum circuit optimizers are not verified, so this procedure is known to be error-prone. For this reason, there is growing interest in the design of equivalence checking algorithms for parameterized quantum circuits. In this paper, we define a generalized class of parameterized circuits with arbitrary rotations and show that this problem is decidable for cyclotomic gate sets. We propose a cutoff-based procedure which reduces the problem of verifying the equivalence of parameterized quantum circuits to the problem of verifying the equivalence of finitely many parameter-free quantum circuits. Because the number of parameter-free circuits grows exponentially with the number of parameters, we also propose a probabilistic variant of the algorithm for cases when the number of parameters is intractably large. We show that our techniques extend to equivalence modulo global phase, and describe an efficient angle sampling procedure for cyclotomic gate sets. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 03:58:51 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Ross",
"Neil J.",
""
],
[
"Wesley",
"Scott",
""
]
] |
2506.20986 | Xiao Zhang | Xiao Zhang, Yongqiang Ma, Haodong Jing, Nanning Zheng | EVA: Mixture-of-Experts Semantic Variant Alignment for Compositional Zero-Shot Learning | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Compositional Zero-Shot Learning (CZSL) investigates compositional generalization capacity to recognize unknown state-object pairs based on learned primitive concepts. Existing CZSL methods typically derive primitives features through a simple composition-prototype mapping, which is suboptimal for a set of individuals that can be divided into distinct semantic subsets. Moreover, the all-to-one cross-modal primitives matching neglects compositional divergence within identical states or objects, limiting fine-grained image-composition alignment. In this study, we propose EVA, a Mixture-of-Experts Semantic Variant Alignment framework for CZSL. Specifically, we introduce domain-expert adaption, leveraging multiple experts to achieve token-aware learning and model high-quality primitive representations. To enable accurate compositional generalization, we further present semantic variant alignment to select semantically relevant representation for image-primitives matching. Our method significantly outperforms other state-of-the-art CZSL methods on three popular benchmarks in both closed- and open-world settings, demonstrating the efficacy of the proposed insight. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:00:55 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhang",
"Xiao",
""
],
[
"Ma",
"Yongqiang",
""
],
[
"Jing",
"Haodong",
""
],
[
"Zheng",
"Nanning",
""
]
] |
2506.20987 | Van-Hai Bui | Akash Mahajan, Shivam Chaturvedi, Srijita Das, Wencong Su, Van-Hai Bui | Optimal Parameter Design for Power Electronic Converters Using a Probabilistic Learning-Based Stochastic Surrogate Model | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | The selection of optimal design for power electronic converter parameters involves balancing efficiency and thermal constraints to ensure high performance without compromising safety. This paper introduces a probabilistic-learning-based stochastic surrogate modeling framework to address this challenge and significantly reduce the time required during the design phase. The approach begins with a neural network classifier that evaluates the feasibility of parameter configurations, effectively filtering out unsafe and/or impractical inputs. Subsequently, a probabilistic prediction model estimates the converter's efficiency and temperature while quantifying prediction uncertainty, providing both performance insights and reliability metrics. Finally, a heuristic optimization-based model is employed to optimize a multi-objective function that maximizes efficiency while adhering to thermal constraints. The optimization process incorporates penalty terms to discourage solutions that violate practical thresholds, ensuring actionable and realistic recommendations. An advanced heuristic optimization method is used to find the optimal solution and is compared with several well-known search algorithms, including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Simulated Annealing (SA), Tabu-Search (TS), and Stochastic Hill Climbing (SHC). The results demonstrate significant improvements in predictive accuracy and optimization outcomes, offering a robust solution for advancing power electronics design. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:00:58 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Mahajan",
"Akash",
""
],
[
"Chaturvedi",
"Shivam",
""
],
[
"Das",
"Srijita",
""
],
[
"Su",
"Wencong",
""
],
[
"Bui",
"Van-Hai",
""
]
] |
2506.20988 | Zhixuan Chen | Zhixuan Chen, Junlin Hou, Liqi Lin, Yihui Wang, Yequan Bie, Xi Wang, Yanning Zhou, Ronald Cheong Kin Chan and Hao Chen | Segment Anything in Pathology Images with Natural Language | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pathology image segmentation is crucial in computational pathology for analyzing histological features relevant to cancer diagnosis and prognosis. However, current methods face major challenges in clinical applications due to limited annotated data and restricted category definitions. To address these limitations, we propose PathSegmentor, the first text-prompted segmentation foundation model designed specifically for pathology images. We also introduce PathSeg , the largest and most comprehensive dataset for pathology segmentation, built from 17 public sources and containing 275k image-mask-label triples across 160 diverse categories. With PathSegmentor, users can perform semantic segmentation using natural language prompts, eliminating the need for laborious spatial inputs such as points or boxes. Extensive experiments demonstrate that PathSegmentor outperforms specialized models with higher accuracy and broader applicability, while maintaining a compact architecture. It significantly surpasses existing spatial- and text-prompted models by 0.145 and 0.429 in overall Dice scores, respectively, showing strong robustness in segmenting complex structures and generalizing to external datasets. Moreover, PathSegmentor's outputs enhance the interpretability of diagnostic models through feature importance estimation and imaging biomarker discovery, offering pathologists evidence-based support for clinical decision-making. This work advances the development of explainable AI in precision oncology. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:01:40 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Zhixuan",
""
],
[
"Hou",
"Junlin",
""
],
[
"Lin",
"Liqi",
""
],
[
"Wang",
"Yihui",
""
],
[
"Bie",
"Yequan",
""
],
[
"Wang",
"Xi",
""
],
[
"Zhou",
"Yanning",
""
],
[
"Chan",
"Ronald Cheong Kin",
""
],
[
"Chen",
"Hao",
""
]
] |
2506.20989 | Eric Zhang | Eric Zhang, Leshem Choshen, Jacob Andreas | Can Gradient Descent Simulate Prompting? | 14 pages, 2 figures | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | There are two primary ways of incorporating new information into a language model (LM): changing its prompt or changing its parameters, e.g. via fine-tuning. Parameter updates incur no long-term storage cost for model changes. However, for many model updates, prompting is significantly more effective: prompted models can generalize robustly from single examples and draw logical inferences that do not occur under standard fine-tuning. Can models be modified so that fine-tuning does emulate prompting? This paper describes a method for meta-training LMs such that gradient updates emulate the effects of conditioning on new information. Our approach uses tools from gradient-based meta-learning but uses an LM's own prompted predictions as targets, eliminating the need for ground-truth labels. Subsequent gradient descent training recovers some (and occasionally all) of prompted model performance -- showing improvement on the ``reversal curse'' tasks, and answering questions about text passages after a single gradient update. These results suggest that, with appropriate initialization, gradient descent can be surprisingly expressive. Our results suggest new avenues for long-context modeling and offer insight into the generalization capabilities of gradient-based learning. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:06:20 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Zhang",
"Eric",
""
],
[
"Choshen",
"Leshem",
""
],
[
"Andreas",
"Jacob",
""
]
] |
2506.20990 | Yifan Yang | Yifan Yang, Zhen Zhang, Rupak Vignesh Swaminathan, Jing Liu, Nathan Susanj, Zheng Zhang | SharpZO: Hybrid Sharpness-Aware Vision Language Model Prompt Tuning via Forward-Only Passes | null | null | null | null | cs.LG cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | Fine-tuning vision language models (VLMs) has achieved remarkable performance across various downstream tasks; yet, it requires access to model gradients through backpropagation (BP), making them unsuitable for memory-constrained, inference-only edge devices. To address this limitation, previous work has explored various BP-free fine-tuning methods. However, these approaches often rely on high-variance evolutionary strategies (ES) or zeroth-order (ZO) optimization, and often fail to achieve satisfactory performance. In this paper, we propose a hybrid Sharpness-aware Zeroth-order optimization (SharpZO) approach, specifically designed to enhance the performance of ZO VLM fine-tuning via a sharpness-aware warm-up training. SharpZO features a two-stage optimization process: a sharpness-aware ES stage that globally explores and smooths the loss landscape to construct a strong initialization, followed by a fine-grained local search via sparse ZO optimization. The entire optimization relies solely on forward passes. Detailed theoretical analysis and extensive experiments on CLIP models demonstrate that SharpZO significantly improves accuracy and convergence speed, achieving up to 7% average gain over state-of-the-art forward-only methods. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:07:14 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yang",
"Yifan",
""
],
[
"Zhang",
"Zhen",
""
],
[
"Swaminathan",
"Rupak Vignesh",
""
],
[
"Liu",
"Jing",
""
],
[
"Susanj",
"Nathan",
""
],
[
"Zhang",
"Zheng",
""
]
] |
2506.20991 | Chade Li | Chade Li, Pengju Zhang, Yihong Wu | TSDASeg: A Two-Stage Model with Direct Alignment for Interactive Point Cloud Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid advancement of 3D vision-language models (VLMs) has spurred significant interest in interactive point cloud processing tasks, particularly for real-world applications. However, existing methods often underperform in point-level tasks, such as segmentation, due to missing direct 3D-text alignment, limiting their ability to link local 3D features with textual context. To solve this problem, we propose TSDASeg, a Two-Stage model coupled with a Direct cross-modal Alignment module and memory module for interactive point cloud Segmentation. We introduce the direct cross-modal alignment module to establish explicit alignment between 3D point clouds and textual/2D image data. Within the memory module, we employ multiple dedicated memory banks to separately store text features, visual features, and their cross-modal correspondence mappings. These memory banks are dynamically leveraged through self-attention and cross-attention mechanisms to update scene-specific features based on prior stored data, effectively addressing inconsistencies in interactive segmentation results across diverse scenarios. Experiments conducted on multiple 3D instruction, reference, and semantic segmentation datasets demonstrate that the proposed method achieves state-of-the-art performance. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:10:33 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Li",
"Chade",
""
],
[
"Zhang",
"Pengju",
""
],
[
"Wu",
"Yihong",
""
]
] |
2506.20992 | Craig Wright PhD | Craig Steven Wright | Institutional Noise, Strategic Deviation, and Intertemporal Collapse: A Formal Model of Miner Behaviour under Protocol Uncertainty | 40 pages, submitted to QJAE | null | null | null | econ.GN cs.CE cs.CY cs.GT cs.SI q-fin.EC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper develops a formal game-theoretic model to examine how protocol mutability disrupts cooperative mining behaviour in blockchain systems. Using a repeated game framework with stochastic rule shocks, we show that even minor uncertainty in institutional rules increases time preference and induces strategic deviation. Fixed-rule environments support long-term investment and stable equilibrium strategies; in contrast, mutable protocols lead to short-termism, higher discounting, and collapse of coordinated engagement. Simulation results identify instability zones in the parameter space where rational mining gives way to extractive or arbitrage conduct. These findings support an Austrian economic interpretation: calculability requires rule stability. Institutional noise undermines the informational basis for productive action. We conclude that protocol design must be treated as a constitutional economic constraint, not a discretionary variable, if sustainable cooperation is to emerge in decentralised systems. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:12:06 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Wright",
"Craig Steven",
""
]
] |
2506.20993 | Dhruv Kumar | Adithya Chittem, Aishna Shrivastava, Sai Tarun Pendela, Jagat Sesh Challa, Dhruv Kumar | SAC: A Framework for Measuring and Inducing Personality Traits in LLMs with Dynamic Intensity Control | Under review | null | null | null | cs.CL cs.AI cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) have gained significant traction across a wide range of fields in recent years. There is also a growing expectation for them to display human-like personalities during interactions. To meet this expectation, numerous studies have proposed methods for modelling LLM personalities through psychometric evaluations. However, most existing models face two major limitations: they rely on the Big Five (OCEAN) framework, which only provides coarse personality dimensions, and they lack mechanisms for controlling trait intensity. In this paper, we address this gap by extending the Machine Personality Inventory (MPI), which originally used the Big Five model, to incorporate the 16 Personality Factor (16PF) model, allowing expressive control over sixteen distinct traits. We also developed a structured framework known as Specific Attribute Control (SAC) for evaluating and dynamically inducing trait intensity in LLMs. Our method introduces adjective-based semantic anchoring to guide trait intensity expression and leverages behavioural questions across five intensity factors: \textit{Frequency}, \textit{Depth}, \textit{Threshold}, \textit{Effort}, and \textit{Willingness}. Through experimentation, we find that modelling intensity as a continuous spectrum yields substantially more consistent and controllable personality expression compared to binary trait toggling. Moreover, we observe that changes in target trait intensity systematically influence closely related traits in psychologically coherent directions, suggesting that LLMs internalize multi-dimensional personality structures rather than treating traits in isolation. Our work opens new pathways for controlled and nuanced human-machine interactions in domains such as healthcare, education, and interviewing processes, bringing us one step closer to truly human-like social machines. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:12:15 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chittem",
"Adithya",
""
],
[
"Shrivastava",
"Aishna",
""
],
[
"Pendela",
"Sai Tarun",
""
],
[
"Challa",
"Jagat Sesh",
""
],
[
"Kumar",
"Dhruv",
""
]
] |
2506.20994 | M{\aa}ns I. Andersson | M{\aa}ns I. Andersson, Martin Karp, Niclas Jansson and Stefano Markidis | Portable High-Performance Kernel Generation for a Computational Fluid Dynamics Code with DaCe | null | null | null | null | cs.DC cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the emergence of new high-performance computing (HPC) accelerators, such as Nvidia and AMD GPUs, efficiently targeting diverse hardware architectures has become a major challenge for HPC application developers. The increasing hardware diversity in HPC systems often necessitates the development of architecture-specific code, hindering the sustainability of large-scale scientific applications. In this work, we leverage DaCe, a data-centric parallel programming framework, to automate the generation of high-performance kernels. DaCe enables automatic code generation for multicore processors and various accelerators, reducing the burden on developers who would otherwise need to rewrite code for each new architecture. Our study demonstrates DaCe's capabilities by applying its automatic code generation to a critical computational kernel used in Computational Fluid Dynamics (CFD). Specifically, we focus on Neko, a Fortran-based solver that employs the spectral-element method, which relies on small tensor operations. We detail the formulation of this computational kernel using DaCe's Stateful Dataflow Multigraph (SDFG) representation and discuss how this approach facilitates high-performance code generation. Additionally, we outline the workflow for seamlessly integrating DaCe's generated code into the Neko solver. Our results highlight the portability and performance of the generated code across multiple platforms, including Nvidia GH200, Nvidia A100, and AMD MI250X GPUs, with competitive performance results. By demonstrating the potential of automatic code generation, we emphasise the feasibility of using portable solutions to ensure the long-term sustainability of large-scale scientific applications. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:17:38 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Andersson",
"Måns I.",
""
],
[
"Karp",
"Martin",
""
],
[
"Jansson",
"Niclas",
""
],
[
"Markidis",
"Stefano",
""
]
] |
2506.20995 | Akio Hayakawa | Akio Hayakawa, Masato Ishii, Takashi Shibuya, Yuki Mitsufuji | Step-by-Step Video-to-Audio Synthesis via Negative Audio Guidance | null | null | null | null | cs.CV cs.LG cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | We propose a novel step-by-step video-to-audio generation method that sequentially produces individual audio tracks, each corresponding to a specific sound event in the video. Our approach mirrors traditional Foley workflows, aiming to capture all sound events induced by a given video comprehensively. Each generation step is formulated as a guided video-to-audio synthesis task, conditioned on a target text prompt and previously generated audio tracks. This design is inspired by the idea of concept negation from prior compositional generation frameworks. To enable this guided generation, we introduce a training framework that leverages pre-trained video-to-audio models and eliminates the need for specialized paired datasets, allowing training on more accessible data. Experimental results demonstrate that our method generates multiple semantically distinct audio tracks for a single input video, leading to higher-quality composite audio synthesis than existing baselines. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:20:08 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Hayakawa",
"Akio",
""
],
[
"Ishii",
"Masato",
""
],
[
"Shibuya",
"Takashi",
""
],
[
"Mitsufuji",
"Yuki",
""
]
] |
2506.20996 | Kevin Mitchell | Kevin A. Mitchell, Md Mainul Hasan Sabbir, Sean Ricarte, Brandon Klein, and Daniel A. Beller | Modeling active nematics via the nematic locking principle | null | null | null | null | cond-mat.soft nlin.CD | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Active nematic systems consist of rod-like internally driven subunits that interact with one another to form large-scale coherent flows. They are important examples of far-from-equilibrium fluids, which exhibit a wealth of nonlinear behavior. This includes active turbulence, in which topological defects braid around one another in a chaotic fashion. One of the most studied examples of active nematics is a dense two-dimensional layer of microtubules, crosslinked by kinesin molecular motors that inject extensile deformations into the fluid. Though numerous studies have modeled microtubule-based active nematics, no consensus has emerged on how to fully capture the features of the experimental system. To better understand the foundations for modeling this system, we propose a fundamental principle we call the nematic locking principle: individual microtubules cannot rotate without all neighboring microtubules also rotating. Physically, this is justified by the high density of the microtubules, their elongated nature, and their corresponding steric interactions. We assert that nematic locking holds throughout the majority of the material but breaks down in the neighborhood of topological defects and other regions of low density. We derive the most general nematic transport equation consistent with this principle and also derive the most general term that violates it. We examine the standard Beris-Edwards approach used to model this system and show that it violates nematic locking throughout the majority of the material. We then propose a modification to the Beris-Edwards model that enforces nematic locking nearly everywhere. This modification shuts off fracturing except in regions where the order parameter is reduced. The resulting simulations show strong nematic locking throughout the bulk of the material, consistent with experimental observation. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:23:59 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Mitchell",
"Kevin A.",
""
],
[
"Sabbir",
"Md Mainul Hasan",
""
],
[
"Ricarte",
"Sean",
""
],
[
"Klein",
"Brandon",
""
],
[
"Beller",
"Daniel A.",
""
]
] |
2506.20997 | Shuang Liu | Chao Yang, Zhizheng Pan, Min Fang, Xian Zhong Zheng, Binyang Liu, Guoliang Li, Tian-Rui Sun, Ji-An Jiang, Miaomiao Zhang, Zhen Wan, Shuang Liu, Han Qu, Ji Yang, Xu Kong, Wenhao Liu, Yiping Shu, Jiang Chang, Tinggui Wang, Lulu Fan, Yongquan Xue, Wentao Luo, Hongxin Zhang, Zheng Lou, Haibin Zhao, Bin Li, Hairen Wang, Dazhi Yao, Jian Wang, Hongfei Zhang, Feng Li, Hao Liu, Ming Liang, Jinlong Tang, Yuheng Zhang, Man Qiao, Run Wen, Zongfei Lyu | A Glimpse of Satellite Galaxies in the Milky Way with the 2.5-meter Wide Field Survey Telescope (WFST): Bootes III and Draco | 17 pages, 12 figures, 3 tables. Accepted for publication in ApJ | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We carry out deep imaging of the Milky Way satellite galaxies, Bootes III and Draco, with WFST as one pilot observing program to demonstrate the capability of WFST. Combining catalogs with PS1 DR2 and Gaia DR3, we derive proper motions for candidate member stars in these two satellite galaxies over a 12-year time baseline, yielding uncertainties of ~1.8 mas/yr at 21 mag and ~3.0 mas/yr at 22 mag in the r band. The proper motions derived from bright and faint stars are consistent, indicating no significant variation in proper motion across stellar luminosity as these galaxies undergo tidal interactions with the MW. Meanwhile, we suggest that Bootes III represents the bound remnant of the progenitor galaxy that gave rise to the Styx stream, as evidenced by its elongated density profile and overdensity in both spatial and kinematic space. This is the first paper to use WFST to measure the proper motions of faint stars in Milky Way satellite galaxies. More detailed analyses will be presented in forthcoming papers from the wide field survey (WFS) program. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:26:57 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yang",
"Chao",
""
],
[
"Pan",
"Zhizheng",
""
],
[
"Fang",
"Min",
""
],
[
"Zheng",
"Xian Zhong",
""
],
[
"Liu",
"Binyang",
""
],
[
"Li",
"Guoliang",
""
],
[
"Sun",
"Tian-Rui",
""
],
[
"Jiang",
"Ji-An",
""
],
[
"Zhang",
"Miaomiao",
""
],
[
"Wan",
"Zhen",
""
],
[
"Liu",
"Shuang",
""
],
[
"Qu",
"Han",
""
],
[
"Yang",
"Ji",
""
],
[
"Kong",
"Xu",
""
],
[
"Liu",
"Wenhao",
""
],
[
"Shu",
"Yiping",
""
],
[
"Chang",
"Jiang",
""
],
[
"Wang",
"Tinggui",
""
],
[
"Fan",
"Lulu",
""
],
[
"Xue",
"Yongquan",
""
],
[
"Luo",
"Wentao",
""
],
[
"Zhang",
"Hongxin",
""
],
[
"Lou",
"Zheng",
""
],
[
"Zhao",
"Haibin",
""
],
[
"Li",
"Bin",
""
],
[
"Wang",
"Hairen",
""
],
[
"Yao",
"Dazhi",
""
],
[
"Wang",
"Jian",
""
],
[
"Zhang",
"Hongfei",
""
],
[
"Li",
"Feng",
""
],
[
"Liu",
"Hao",
""
],
[
"Liang",
"Ming",
""
],
[
"Tang",
"Jinlong",
""
],
[
"Zhang",
"Yuheng",
""
],
[
"Qiao",
"Man",
""
],
[
"Wen",
"Run",
""
],
[
"Lyu",
"Zongfei",
""
]
] |
2506.20998 | Yeon Ji Song | Yeon-Ji Song, Jaein Kim, Byung-Ju Kim, Byoung-Tak Zhang | DBMovi-GS: Dynamic View Synthesis from Blurry Monocular Video via Sparse-Controlled Gaussian Splatting | CVPRW 2025, Neural Fields Beyond Conventional Cameras | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Novel view synthesis is a task of generating scenes from unseen perspectives; however, synthesizing dynamic scenes from blurry monocular videos remains an unresolved challenge that has yet to be effectively addressed. Existing novel view synthesis methods are often constrained by their reliance on high-resolution images or strong assumptions about static geometry and rigid scene priors. Consequently, their approaches lack robustness in real-world environments with dynamic object and camera motion, leading to instability and degraded visual fidelity. To address this, we propose Motion-aware Dynamic View Synthesis from Blurry Monocular Video via Sparse-Controlled Gaussian Splatting (DBMovi-GS), a method designed for dynamic view synthesis from blurry monocular videos. Our model generates dense 3D Gaussians, restoring sharpness from blurry videos and reconstructing detailed 3D geometry of the scene affected by dynamic motion variations. Our model achieves robust performance in novel view synthesis under dynamic blurry scenes and sets a new benchmark in realistic novel view synthesis for blurry monocular video inputs. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:28:48 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Song",
"Yeon-Ji",
""
],
[
"Kim",
"Jaein",
""
],
[
"Kim",
"Byung-Ju",
""
],
[
"Zhang",
"Byoung-Tak",
""
]
] |
2506.20999 | Soujun Kitagawa | Soujun Kitagawa | Signed Minkowski decompositions of convex polygons into minimum simplices and factorization of max-plus functions | 24 pages, 3 figures | null | null | null | math.CO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Signed Minkowski decomposition is an expression of a polytope as a Minkowski sum and difference of smaller polytopes. Signed Minkowski decompositions of a polytope can be interpreted as factorizations of a max-plus (tropical) function. We review two relations about Minkowski decompositions, and we prove that any 2-dimensional integral polytopes (polygons) have a signed Minkowski decomposition which consists of integral points, integral line segments of length 1, and integral triangles of area 1/2. From this result, we also obtain that any max-plus functions with two variables and integer coefficients can be expressed as a set of specified form of simpler max-plus functions. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:28:56 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Kitagawa",
"Soujun",
""
]
] |
2506.21000 | Wei Li | Zhiyi Chen, Zhilin Ren, Wei Li, ZhenZhen Huo, ZhuangZheng Wang, Ye Liu, Bowen Hu, Wanting Chen, Ting Xu, Artemiy Leonov, Chenyan Zhang, Bernhard Hommel, Tingyong Feng | Modulating task outcome value to mitigate real-world procrastination via noninvasive brain stimulation | null | null | null | null | q-bio.NC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Procrastination represents one of the most prevalent behavioral problems affecting individual health and societal productivity. Although it is often conceptualized as a form of self-control failure, its underlying neurocognitive mechanisms are poorly understood. A leading model posits that procrastination arises from imbalanced competing motivations: the avoidance of negative task aversiveness and the pursuit of positive task outcomes, yet this theoretical framework has not fully validated in real-world settings and not applied effectively to guide interventions. Here, we addressed this gap with a preregistered, double-blind, randomized controlled trial. We applied seven sessions of high-definition transcranial direct current stimulation (HD-tDCS) to the left dorsolateral prefrontal cortex (DLPFC), a key region for self-control, in chronic procrastinators. Using the intensive experience sampling method (iESM), we assessed the effect of anodal HD-tDCS on real-world procrastination behavior at offline after-effect (2-day interval) and long-term retention (6-month follow-up). We found that this neuromodulation produced a lasting reduction in real-world procrastination, with effects sustained at a 6-month follow-up. While the intervention both decreased task aversiveness and increased perceived task outcome value, causal mediation analysis revealed a striking mechanism: the increase in task outcome value uniquely and sufficiently mediated the entire behavioral improvement. In conclusion, these findings provide causal evidence that enhancing DLPFC function mitigates procrastination by selectively amplifying the valuation of future rewards, not by simply reducing negative feelings about the task. This establishes a precise, value-driven neurocognitive pathway for self-control and offers a validated, theory-driven strategy for intervention. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:30:51 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chen",
"Zhiyi",
""
],
[
"Ren",
"Zhilin",
""
],
[
"Li",
"Wei",
""
],
[
"Huo",
"ZhenZhen",
""
],
[
"Wang",
"ZhuangZheng",
""
],
[
"Liu",
"Ye",
""
],
[
"Hu",
"Bowen",
""
],
[
"Chen",
"Wanting",
""
],
[
"Xu",
"Ting",
""
],
[
"Leonov",
"Artemiy",
""
],
[
"Zhang",
"Chenyan",
""
],
[
"Hommel",
"Bernhard",
""
],
[
"Feng",
"Tingyong",
""
]
] |
2506.21001 | Qiuyi Qi | Qiuyi Qi, Xin Li, Ming Kong, Zikang Xu, Bingdi Chen, Qiang Zhu, S Kevin Zhou | Style-Aligned Image Composition for Robust Detection of Abnormal Cells in Cytopathology | MIDL 2025 Oral | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Challenges such as the lack of high-quality annotations, long-tailed data distributions, and inconsistent staining styles pose significant obstacles to training neural networks to detect abnormal cells in cytopathology robustly. This paper proposes a style-aligned image composition (SAIC) method that composes high-fidelity and style-preserved pathological images to enhance the effectiveness and robustness of detection models. Without additional training, SAIC first selects an appropriate candidate from the abnormal cell bank based on attribute guidance. Then, it employs a high-frequency feature reconstruction to achieve a style-aligned and high-fidelity composition of abnormal cells and pathological backgrounds. Finally, it introduces a large vision-language model to filter high-quality synthesis images. Experimental results demonstrate that incorporating SAIC-synthesized images effectively enhances the performance and robustness of abnormal cell detection for tail categories and styles, thereby improving overall detection performance. The comprehensive quality evaluation further confirms the generalizability and practicality of SAIC in clinical application scenarios. Our code will be released at https://github.com/Joey-Qi/SAIC. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:32:21 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Qi",
"Qiuyi",
""
],
[
"Li",
"Xin",
""
],
[
"Kong",
"Ming",
""
],
[
"Xu",
"Zikang",
""
],
[
"Chen",
"Bingdi",
""
],
[
"Zhu",
"Qiang",
""
],
[
"Zhou",
"S Kevin",
""
]
] |
2506.21002 | Takumi Yoshimatsu | Takumi Yoshimatsu, Shumpei Takezaki and Seiichi Uchida | Inverse Scene Text Removal | 17 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scene text removal (STR) aims to erase textual elements from images. It was originally intended for removing privacy-sensitiveor undesired texts from natural scene images, but is now also appliedto typographic images. STR typically detects text regions and theninpaints them. Although STR has advanced through neural networksand synthetic data, misuse risks have increased. This paper investi-gates Inverse STR (ISTR), which analyzes STR-processed images andfocuses on binary classification (detecting whether an image has un-dergone STR) and localizing removed text regions. We demonstrate inexperiments that these tasks are achievable with high accuracies, en-abling detection of potential misuse and improving STR. We also at-tempt to recover the removed text content by training a text recognizerto understand its difficulty. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:32:35 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yoshimatsu",
"Takumi",
""
],
[
"Takezaki",
"Shumpei",
""
],
[
"Uchida",
"Seiichi",
""
]
] |
2506.21003 | Steven Walton | Steven Walton, Valeriy Klyukin, Maksim Artemev, Denis Derkach, Nikita Orlov, Humphrey Shi | Distilling Normalizing Flows | Published in eLVM @ CVPR (https://openaccess.thecvf.com/content/CVPR2025W/eLVM/html/Walton_Distilling_Normalizing_Flows_CVPRW_2025_paper) | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Explicit density learners are becoming an increasingly popular technique for generative models because of their ability to better model probability distributions. They have advantages over Generative Adversarial Networks due to their ability to perform density estimation and having exact latent-variable inference. This has many advantages, including: being able to simply interpolate, calculate sample likelihood, and analyze the probability distribution. The downside of these models is that they are often more difficult to train and have lower sampling quality.
Normalizing flows are explicit density models, that use composable bijective functions to turn an intractable probability function into a tractable one. In this work, we present novel knowledge distillation techniques to increase sampling quality and density estimation of smaller student normalizing flows. We seek to study the capacity of knowledge distillation in Compositional Normalizing Flows to understand the benefits and weaknesses provided by these architectures. Normalizing flows have unique properties that allow for a non-traditional forms of knowledge transfer, where we can transfer that knowledge within intermediate layers. We find that through this distillation, we can make students significantly smaller while making substantial performance gains over a non-distilled student. With smaller models there is a proportionally increased throughput as this is dependent upon the number of bijectors, and thus parameters, in the network. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:34:28 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Walton",
"Steven",
""
],
[
"Klyukin",
"Valeriy",
""
],
[
"Artemev",
"Maksim",
""
],
[
"Derkach",
"Denis",
""
],
[
"Orlov",
"Nikita",
""
],
[
"Shi",
"Humphrey",
""
]
] |
2506.21004 | Emmanuel Terzoudis-Lumsden | Emmanuel W.C. Terzoudis-Lumsden, Alireza Sadri, Matthew Weyland, Laure Bourgeois, Stephanie M. Ribet, Georgios Varnavides, Colin Ophus, Timothy C. Petersen, Scott D. Findlay | Quantitative structure determination from experimental four-dimensional scanning transmission electron microscopy via the scattering matrix | 24 pages, 9 figures | null | null | null | cond-mat.mtrl-sci physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Considerable inroads have recently been made on algorithms to determine the sample potential from four-dimensional scanning transmission electron microscopy data from thick samples where multiple scattering cannot be neglected. This paper further develops the scattering matrix approach to such structure determination. Through simulation, we demonstrate how this approach can be modified to better handle partial spatial coherence, unknown probe defocus, and information from the dark field region. By combining these developments we reconstruct the electrostatic potential of a monolithic SrTiO$_3$ crystal showing good quantitative agreement with the expected structure. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:37:39 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Terzoudis-Lumsden",
"Emmanuel W. C.",
""
],
[
"Sadri",
"Alireza",
""
],
[
"Weyland",
"Matthew",
""
],
[
"Bourgeois",
"Laure",
""
],
[
"Ribet",
"Stephanie M.",
""
],
[
"Varnavides",
"Georgios",
""
],
[
"Ophus",
"Colin",
""
],
[
"Petersen",
"Timothy C.",
""
],
[
"Findlay",
"Scott D.",
""
]
] |
2506.21005 | Trung Nghia Le | Lam-Huy Nguyen and Thinh-Phuc Nguyen and Thanh-Hai Nguyen and Gia-Huy Dinh and Minh-Triet Tran and Trung-Nghia Le | VisionGuard: Synergistic Framework for Helmet Violation Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Enforcing helmet regulations among motorcyclists is essential for enhancing road safety and ensuring the effectiveness of traffic management systems. However, automatic detection of helmet violations faces significant challenges due to environmental variability, camera angles, and inconsistencies in the data. These factors hinder reliable detection of motorcycles and riders and disrupt consistent object classification. To address these challenges, we propose VisionGuard, a synergistic multi-stage framework designed to overcome the limitations of frame-wise detectors, especially in scenarios with class imbalance and inconsistent annotations. VisionGuard integrates two key components: Adaptive Labeling and Contextual Expander modules. The Adaptive Labeling module is a tracking-based refinement technique that enhances classification consistency by leveraging a tracking algorithm to assign persistent labels across frames and correct misclassifications. The Contextual Expander module improves recall for underrepresented classes by generating virtual bounding boxes with appropriate confidence scores, effectively addressing the impact of data imbalance. Experimental results show that VisionGuard improves overall mAP by 3.1% compared to baseline detectors, demonstrating its effectiveness and potential for real-world deployment in traffic surveillance systems, ultimately promoting safety and regulatory compliance. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:45:18 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Nguyen",
"Lam-Huy",
""
],
[
"Nguyen",
"Thinh-Phuc",
""
],
[
"Nguyen",
"Thanh-Hai",
""
],
[
"Dinh",
"Gia-Huy",
""
],
[
"Tran",
"Minh-Triet",
""
],
[
"Le",
"Trung-Nghia",
""
]
] |
2506.21006 | Tyler Ward | Tyler Ward, Xiaoqin Wang, Braxton McFarland, Md Atik Ahamed, Sahar Nozad, Talal Arshad, Hafsa Nebbache, Jin Chen, Abdullah Imran | Detection of Breast Cancer Lumpectomy Margin with SAM-incorporated Forward-Forward Contrastive Learning | 19 pages, 7 figures, 3 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Complete removal of cancer tumors with a negative specimen margin during lumpectomy is essential in reducing breast cancer recurrence. However, 2D specimen radiography (SR), the current method used to assess intraoperative specimen margin status, has limited accuracy, resulting in nearly a quarter of patients requiring additional surgery. To address this, we propose a novel deep learning framework combining the Segment Anything Model (SAM) with Forward-Forward Contrastive Learning (FFCL), a pre-training strategy leveraging both local and global contrastive learning for patch-level classification of SR images. After annotating SR images with regions of known maligancy, non-malignant tissue, and pathology-confirmed margins, we pre-train a ResNet-18 backbone with FFCL to classify margin status, then reconstruct coarse binary masks to prompt SAM for refined tumor margin segmentation. Our approach achieved an AUC of 0.8455 for margin classification and segmented margins with a 27.4% improvement in Dice similarity over baseline models, while reducing inference time to 47 milliseconds per image. These results demonstrate that FFCL-SAM significantly enhances both the speed and accuracy of intraoperative margin assessment, with strong potential to reduce re-excision rates and improve surgical outcomes in breast cancer treatment. Our code is available at https://github.com/tbwa233/FFCL-SAM/. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:46:28 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Ward",
"Tyler",
""
],
[
"Wang",
"Xiaoqin",
""
],
[
"McFarland",
"Braxton",
""
],
[
"Ahamed",
"Md Atik",
""
],
[
"Nozad",
"Sahar",
""
],
[
"Arshad",
"Talal",
""
],
[
"Nebbache",
"Hafsa",
""
],
[
"Chen",
"Jin",
""
],
[
"Imran",
"Abdullah",
""
]
] |
2506.21007 | Benjamin Bursik | Benjamin Bursik, Nikolaos Karadimitriou, Holger Steeb, Joachim Gross | Static Contact Angles of Mixtures: Classical Density Functional Theory and Experimental Investigation | 10 pages, 7 figures | null | null | null | physics.flu-dyn cond-mat.stat-mech physics.chem-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This work assesses a classical density functional theory (DFT) model for predicting macroscopic static contact angles of pure substances and mixtures by comparison to own experimental data. We employ a DFT with a Helmholtz energy functional based on the perturbed-chain statistical associating fluid theory (PC-SAFT) for the fluid-fluid interactions and an effective external potential for the solid-fluid interactions. The solid substrate is characterized by adjusting a single solid-solid interaction energy parameter to a single contact angle value of $n$-octane, while all other results are predictions based on Berthelot-Lorentz combining rules. The surface tensions between solid, liquid and vapor phases are determined from one-dimensional DFT calculations, and Young's equation is used to calculate the contact angle. A non-polar polytetrafluoroethylene (PTFE, Teflon) substrate is used, and experiments are carried out using the sessile droplet method with graphical evaluation of the contact angle. Accurate results are obtained for pure substances except for monohydric alcohols. Contact angles for monohydric alcohols are systematically overestimated, and we show that this is partially due to neglecting orientational effects. The approach provides accurate results for mixtures whenever the respective pure substance contact angle is described well. This comprises mixtures of non-polar, polar, and hydrogen bonding substances, including mixtures with water. Our results suggest that DFT based on PC-SAFT provides a fast and accurate method for prediction of contact angles for a wide range of pure substances and mixtures. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:57:04 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Bursik",
"Benjamin",
""
],
[
"Karadimitriou",
"Nikolaos",
""
],
[
"Steeb",
"Holger",
""
],
[
"Gross",
"Joachim",
""
]
] |
2506.21008 | Bang Gong | Bang Gong, Luchao Qi, Jiaye Wu, Zhicheng Fu, Chunbo Song, David W. Jacobs, John Nicholson, Roni Sengupta | The Aging Multiverse: Generating Condition-Aware Facial Aging Tree via Training-Free Diffusion | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce the Aging Multiverse, a framework for generating multiple plausible facial aging trajectories from a single image, each conditioned on external factors such as environment, health, and lifestyle. Unlike prior methods that model aging as a single deterministic path, our approach creates an aging tree that visualizes diverse futures. To enable this, we propose a training-free diffusion-based method that balances identity preservation, age accuracy, and condition control. Our key contributions include attention mixing to modulate editing strength and a Simulated Aging Regularization strategy to stabilize edits. Extensive experiments and user studies demonstrate state-of-the-art performance across identity preservation, aging realism, and conditional alignment, outperforming existing editing and age-progression models, which often fail to account for one or more of the editing criteria. By transforming aging into a multi-dimensional, controllable, and interpretable process, our approach opens up new creative and practical avenues in digital storytelling, health education, and personalized visualization. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 04:57:47 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Gong",
"Bang",
""
],
[
"Qi",
"Luchao",
""
],
[
"Wu",
"Jiaye",
""
],
[
"Fu",
"Zhicheng",
""
],
[
"Song",
"Chunbo",
""
],
[
"Jacobs",
"David W.",
""
],
[
"Nicholson",
"John",
""
],
[
"Sengupta",
"Roni",
""
]
] |
2506.21009 | Ayaka Yasunaga | Ayaka Yasunaga, Hideo Saito, Shohei Mori | User-in-the-Loop View Sampling with Error Peaking Visualization | Accepted at IEEE ICIP 2025, Project Page: https://mediated-reality.github.io/projects/yasunaga_icip25/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Augmented reality (AR) provides ways to visualize missing view samples for novel view synthesis. Existing approaches present 3D annotations for new view samples and task users with taking images by aligning the AR display. This data collection task is known to be mentally demanding and limits capture areas to pre-defined small areas due to the ideal but restrictive underlying sampling theory. To free users from 3D annotations and limited scene exploration, we propose using locally reconstructed light fields and visualizing errors to be removed by inserting new views. Our results show that the error-peaking visualization is less invasive, reduces disappointment in final results, and is satisfactory with fewer view samples in our mobile view synthesis system. We also show that our approach can contribute to recent radiance field reconstruction for larger scenes, such as 3D Gaussian splatting. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:01:41 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yasunaga",
"Ayaka",
""
],
[
"Saito",
"Hideo",
""
],
[
"Mori",
"Shohei",
""
]
] |
2506.21010 | Shin'ichi Nojiri | Shin'ichi Nojiri, S.D. Odintsov, V.K. Oikonomou | Phantom Crossing and Oscillating Dark Energy with $F(R)$ Gravity | LaTeX 27 pages | null | null | KEK-TH-2734, KEK-Cosmo-0383 | gr-qc astro-ph.CO hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we shall consider how a dynamical oscillating and phantom crossing dark energy era can be realised in the context of $F(R)$ gravity. We approach the topic from a theoretical standpoint, considering all the conditions that may lead to a consistent phantom crossing behaviour and separately how the $F(R)$ gravity context may realise an oscillating dark energy era. Apart from our qualitative considerations, we study quantitatively two $F(R)$ gravity dark energy models, which are viable cosmologically and also exhibit simultaneously phantom crossing behaviour and oscillating dark energy. We consider these models by numerically solving the field equations using appropriate statefinder parameters engineered for dark energy studies. As we show, $F(R)$ provides a natural extension of Einstein's general relativity, which can naturally realise a transition from a phantom era to a quintessential era, a feature supported by recent observational data, without resorting to phantom scalar fields to realise the phantom evolution. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:02:19 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Nojiri",
"Shin'ichi",
""
],
[
"Odintsov",
"S. D.",
""
],
[
"Oikonomou",
"V. K.",
""
]
] |
2506.21011 | Qizhi Xie | Qizhi Xie, Kun Yuan, Yunpeng Qu, Jiachao Gong, Mingda Wu, Ming Sun, Chao Zhou, Jihong Zhu | Bridging Video Quality Scoring and Justification via Large Multimodal Models | 15 pages, 4 figures, 8 tables | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Classical video quality assessment (VQA) methods generate a numerical score to judge a video's perceived visual fidelity and clarity. Yet, a score fails to describe the video's complex quality dimensions, restricting its applicability. Benefiting from the linguistic output, adapting video large multimodal models (LMMs) to VQA via instruction tuning has the potential to address this issue. The core of the approach lies in the video quality-centric instruction data. Previous explorations mainly focus on the image domain, and their data generation processes heavily rely on human quality annotations and proprietary systems, limiting data scalability and effectiveness. To address these challenges, we propose the Score-based Instruction Generation (SIG) pipeline. Specifically, SIG first scores multiple quality dimensions of an unlabeled video and maps scores to text-defined levels. It then explicitly incorporates a hierarchical Chain-of-Thought (CoT) to model the correlation between specific dimensions and overall quality, mimicking the human visual system's reasoning process. The automated pipeline eliminates the reliance on expert-written quality descriptions and proprietary systems, ensuring data scalability and generation efficiency. To this end, the resulting Score2Instruct (S2I) dataset contains over 320K diverse instruction-response pairs, laying the basis for instruction tuning. Moreover, to advance video LMMs' quality scoring and justification abilities simultaneously, we devise a progressive tuning strategy to fully unleash the power of S2I. Built upon SIG, we further curate a benchmark termed S2I-Bench with 400 open-ended questions to better evaluate the quality justification capacity of video LMMs. Experimental results on the S2I-Bench and existing benchmarks indicate that our method consistently improves quality scoring and justification capabilities across multiple video LMMs. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:02:25 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Xie",
"Qizhi",
""
],
[
"Yuan",
"Kun",
""
],
[
"Qu",
"Yunpeng",
""
],
[
"Gong",
"Jiachao",
""
],
[
"Wu",
"Mingda",
""
],
[
"Sun",
"Ming",
""
],
[
"Zhou",
"Chao",
""
],
[
"Zhu",
"Jihong",
""
]
] |
2506.21012 | Huan Wang | Huan Wang, Haoran Li, Huaming Chen, Jun Yan, Jiahua Shi, Jun Shen | FedSC: Federated Learning with Semantic-Aware Collaboration | 12 pages, KDD 2025 | null | 10.1145/3711896.3736957 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated learning (FL) aims to train models collaboratively across clients without sharing data for privacy-preserving. However, one major challenge is the data heterogeneity issue, which refers to the biased labeling preferences at multiple clients. A number of existing FL methods attempt to tackle data heterogeneity locally (e.g., regularizing local models) or globally (e.g., fine-tuning global model), often neglecting inherent semantic information contained in each client. To explore the possibility of using intra-client semantically meaningful knowledge in handling data heterogeneity, in this paper, we propose Federated Learning with Semantic-Aware Collaboration (FedSC) to capture client-specific and class-relevant knowledge across heterogeneous clients. The core idea of FedSC is to construct relational prototypes and consistent prototypes at semantic-level, aiming to provide fruitful class underlying knowledge and stable convergence signals in a prototype-wise collaborative way. On the one hand, FedSC introduces an inter-contrastive learning strategy to bring instance-level embeddings closer to relational prototypes with the same semantics and away from distinct classes. On the other hand, FedSC devises consistent prototypes via a discrepancy aggregation manner, as a regularization penalty to constrain the optimization region of the local model. Moreover, a theoretical analysis for FedSC is provided to ensure a convergence guarantee. Experimental results on various challenging scenarios demonstrate the effectiveness of FedSC and the efficiency of crucial components. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:04:55 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Wang",
"Huan",
""
],
[
"Li",
"Haoran",
""
],
[
"Chen",
"Huaming",
""
],
[
"Yan",
"Jun",
""
],
[
"Shi",
"Jiahua",
""
],
[
"Shen",
"Jun",
""
]
] |
2506.21013 | June-Young Kim | June-Young Kim | Chiral-odd generalized parton distributions in the large-$N_{c}$ limit of QCD: Next-to-leading-order contributions | 25 pages, 6 figures | null | null | JLAB-THY-25-4386 | hep-ph hep-lat | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the nucleon's chiral-odd generalized parton distribution functions (GPDs) in the large-$N_c$ limit of QCD. Extending previous work on the leading-order contribution in the $1/N_c$ expansion, we focus on the next-to-leading-order contributions and provide a complete set of flavor-singlet and flavor-non-singlet chiral-odd GPDs. This study includes the derivation of the spin-flavor structure of the baryon matrix element of the chiral-odd operator, the proof of the polynomiality property and associated sum rules, and numerical estimates based on the gradient expansion. The spin-flavor structure of the nucleon matrix element is interpreted through a multipole expansion in the transverse momentum transfer, leading to the definition of multipole mean-field GPDs. Using these GPDs as the basis of our analysis, we take their $m$-th moments and demonstrate the polynomiality property within the mean-field framework, making use of discrete symmetries in the proof. In particular, the first moments ($m = 1$) of the GPDs are related to the nucleon tensor form factors, as generally required. By performing a gradient expansion, we compute the chiral-odd GPDs and present numerical estimates. Our results show good agreement with lattice QCD predictions. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:10:00 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Kim",
"June-Young",
""
]
] |
2506.21014 | Shaojian Qiu Dr. | Shaojian Qiu, Mengyang Huang, Jiahao Cheng | Boosting Vulnerability Detection with Inter-function Multilateral Association Insights | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vulnerability detection is a crucial yet challenging technique for ensuring the security of software systems. Currently, most deep learning-based vulnerability detection methods focus on stand-alone functions, neglecting the complex inter-function interrelations, particularly the multilateral associations. This oversight can fail to detect vulnerabilities in these interrelations. To address this gap, we present an Inter-Function Multilateral Association analysis framework for Vulnerability Detection (IFMA-VD). The cornerstone of the IFMA-VD lies in constructing a code behavior hypergraph and utilizing hyperedge convolution to extract multilateral association features. Specifically, we first parse functions into a code property graph to generate intra-function features. Following this, we construct a code behavior hypergraph by segmenting the program dependency graph to isolate and encode behavioral features into hyperedges. Finally, we utilize a hypergraph network to capture the multilateral association knowledge for augmenting vulnerability detection. We evaluate IFMA-VD on three widely used vulnerability datasets and demonstrate improvements in F-measure and Recall compared to baseline methods. Additionally, we illustrate that multilateral association features can boost code feature representation and validate the effectiveness of IFMA-VD on real-world datasets. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:11:21 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Qiu",
"Shaojian",
""
],
[
"Huang",
"Mengyang",
""
],
[
"Cheng",
"Jiahao",
""
]
] |
2506.21015 | Qingyue Jiao | Qingyue Jiao, Kangyu Zheng, Yiyu Shi, Zhiding Liang | HybridQ: Hybrid Classical-Quantum Generative Adversarial Network for Skin Disease Image Generation | null | null | null | null | cs.CV cs.LG quant-ph | http://creativecommons.org/licenses/by/4.0/ | Machine learning-assisted diagnosis is gaining traction in skin disease detection, but training effective models requires large amounts of high-quality data. Skin disease datasets often suffer from class imbalance, privacy concerns, and object bias, making data augmentation essential. While classical generative models are widely used, they demand extensive computational resources and lengthy training time. Quantum computing offers a promising alternative, but existing quantum-based image generation methods can only yield grayscale low-quality images. Through a novel classical-quantum latent space fusion technique, our work overcomes this limitation and introduces the first classical-quantum generative adversarial network (GAN) capable of generating color medical images. Our model outperforms classical deep convolutional GANs and existing hybrid classical-quantum GANs in both image generation quality and classification performance boost when used as data augmentation. Moreover, the performance boost is comparable with that achieved using state-of-the-art classical generative models, yet with over 25 times fewer parameters and 10 times fewer training epochs. Such results suggest a promising future for quantum image generation as quantum hardware advances. Finally, we demonstrate the robust performance of our model on real IBM quantum machine with hardware noise. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:14:45 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Jiao",
"Qingyue",
""
],
[
"Zheng",
"Kangyu",
""
],
[
"Shi",
"Yiyu",
""
],
[
"Liang",
"Zhiding",
""
]
] |
2506.21016 | Bala Chidambaram | B. Chidambaram, A. Hilbert, M. Silva | Fault-Tolerant Spacecraft Attitude Determination using State Estimation Techniques | 8 pages, 19 figures | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | The extended and unscented Kalman filter, and the particle filter provide a robust framework for fault-tolerant attitude estimation on spacecraft. This paper explores how each filter performs for a large satellite in a low earth orbit. Additionally, various techniques, built on these filters, for fault detection, isolation and recovery from erroneous sensor measurements, are analyzed. Key results from this analysis include filter performance for various fault modes. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:19:38 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Chidambaram",
"B.",
""
],
[
"Hilbert",
"A.",
""
],
[
"Silva",
"M.",
""
]
] |
2506.21017 | Fuyan Ma | Fuyan Ma, Yiran He, Bin Sun, Shutao Li | Multimodal Prompt Alignment for Facial Expression Recognition | To appear in ICCV2025 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Prompt learning has been widely adopted to efficiently adapt vision-language models (VLMs) like CLIP for various downstream tasks. Despite their success, current VLM-based facial expression recognition (FER) methods struggle to capture fine-grained textual-visual relationships, which are essential for distinguishing subtle differences between facial expressions. To address this challenge, we propose a multimodal prompt alignment framework for FER, called MPA-FER, that provides fine-grained semantic guidance to the learning process of prompted visual features, resulting in more precise and interpretable representations. Specifically, we introduce a multi-granularity hard prompt generation strategy that utilizes a large language model (LLM) like ChatGPT to generate detailed descriptions for each facial expression. The LLM-based external knowledge is injected into the soft prompts by minimizing the feature discrepancy between the soft prompts and the hard prompts. To preserve the generalization abilities of the pretrained CLIP model, our approach incorporates prototype-guided visual feature alignment, ensuring that the prompted visual features from the frozen image encoder align closely with class-specific prototypes. Additionally, we propose a cross-modal global-local alignment module that focuses on expression-relevant facial features, further improving the alignment between textual and visual features. Extensive experiments demonstrate our framework outperforms state-of-the-art methods on three FER benchmark datasets, while retaining the benefits of the pretrained model and minimizing computational costs. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:28:57 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Ma",
"Fuyan",
""
],
[
"He",
"Yiran",
""
],
[
"Sun",
"Bin",
""
],
[
"Li",
"Shutao",
""
]
] |
2506.21018 | Lei Hao | Lei Hao, Lina Xu, Chang Liu, and Yanni Dong | LASFNet: A Lightweight Attention-Guided Self-Modulation Feature Fusion Network for Multimodal Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Effective deep feature extraction via feature-level fusion is crucial for multimodal object detection. However, previous studies often involve complex training processes that integrate modality-specific features by stacking multiple feature-level fusion units, leading to significant computational overhead. To address this issue, we propose a new fusion detection baseline that uses a single feature-level fusion unit to enable high-performance detection, thereby simplifying the training process. Based on this approach, we propose a lightweight attention-guided self-modulation feature fusion network (LASFNet), which introduces a novel attention-guided self-modulation feature fusion (ASFF) module that adaptively adjusts the responses of fusion features at both global and local levels based on attention information from different modalities, thereby promoting comprehensive and enriched feature generation. Additionally, a lightweight feature attention transformation module (FATM) is designed at the neck of LASFNet to enhance the focus on fused features and minimize information loss. Extensive experiments on three representative datasets demonstrate that, compared to state-of-the-art methods, our approach achieves a favorable efficiency-accuracy trade-off, reducing the number of parameters and computational cost by as much as 90% and 85%, respectively, while improving detection accuracy (mAP) by 1%-3%. The code will be open-sourced at https://github.com/leileilei2000/LASFNet. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:32:33 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Hao",
"Lei",
""
],
[
"Xu",
"Lina",
""
],
[
"Liu",
"Chang",
""
],
[
"Dong",
"Yanni",
""
]
] |
2506.21019 | Li-Xue Yue | Li-Xue Yue, Da-Ming Chen | Statistical Strong Lensing as a Test of Conformal Gravity | 19 pages, 3 figures, 2 tables, Published in Universe Journal | Universe 2025, 11(6), 178 | 10.3390/universe11060178 | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | As an alternative gravitational theory to General Relativity (GR), Conformal Gravity (CG) can be verified through astronomical observations. Currently, Mannheim and Kazanas have provided vacuum solutions for cosmological and local gravitational systems, and these solutions may resolve the dark matter and dark energy issues encountered in GR, making them particularly valuable. For static, spherically symmetric systems, CG predicts an additional linear potential generated by luminous matter in addition to the conventional Newtonian potential. This extra potential is expected to account for the observations of galaxies and galaxy clusters without the need of dark matter. It is characterized by the parameter $\gamma^*$, which corresponds to the linear potential generated by the unit of the solar mass, and it is thus a universal constant. The value of $\gamma^\ast$ was determined by fitting the rotation curve data of spiral galaxies. These predictions of CG should also be verified by the observations of strong gravitational lensing. In this study, building upon the previous research, we tested CG via strong lensing statistics. We used a well-defined sample that consisted of both galaxies and galaxy clusters. This allowed us to test CG through statistical strong lensing in a way similar to the conventional approach in GR. As anticipated, our results were consistent with previous studies, namely that the fitted $\gamma^*$ is much larger than that from rotation curves. Intriguingly, we further discovered that, in order to fit the strong lensing data of another sample, the value of $\gamma^*$ cannot be a constant, as is required in CG. Instead, we derived a formula for $\gamma^*$ as a function of the stellar mass $M_*$ of the galaxies or galaxy clusters. It was found that $\gamma^*$ decreases as $M_*$ increases. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:40:31 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Yue",
"Li-Xue",
""
],
[
"Chen",
"Da-Ming",
""
]
] |
2506.21020 | Mallory Flynn PhD | Mallory J Flynn and Paul Gustafson | Leveraging Relational Evidence: Population Size Estimation on Tree-Structured Data with the Weighted Multiplier Method | 26 pages, 15 figures (7 in main text, 8 in supplementary), 4 tables | null | null | null | stat.ME | http://creativecommons.org/licenses/by/4.0/ | Populations of interest are often hidden from data for a variety of reasons, though their magnitude remains important in determining resource allocation and appropriate policy. One popular approach to population size estimation, the multiplier method, is a back-calculation tool requiring only a marginal subpopulation size and an estimate of the proportion belonging to this subgroup. Another approach is to use Bayesian methods, which are inherently well-suited to incorporating multiple data sources. However, both methods have their drawbacks. A framework for applying the multiplier method which combines information from several known subpopulations has not yet been established; Bayesian models, though able to incorporate complex dependencies and various data sources, can be difficult for researchers in less technical fields to design and implement. Increasing data collection and linkage across diverse fields suggests accessible methods of estimating population size with synthesized data are needed. We propose an extension to the well-known multiplier method which is applicable to tree-structured data, where multiple subpopulations and corresponding proportions combine to generate a population size estimate via the minimum variance estimator. The methodology and resulting estimates are compared with those from a Bayesian hierarchical model, for both simulated and real world data. Subsequent analysis elucidates which data are key to estimation in each method, and examines robustness and feasibility of this new methodology. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:46:50 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Flynn",
"Mallory J",
""
],
[
"Gustafson",
"Paul",
""
]
] |
2506.21021 | Kihyun Jung | T. Akutsu, M. Ando, M. Aoumi, A. Araya, Y. Aso, L. Baiotti, R. Bajpai, K. Cannon, A. H.-Y. Chen, D. Chen, H. Chen, A. Chiba, C. Chou, M. Eisenmann, K. Endo, T. Fujimori, S. Garg, D. Haba, S. Haino, R. Harada, H. Hayakawa, K. Hayama, S. Fujii, Y. Himemoto, N. Hirata, C. Hirose, H.-F. Hsieh, H.-Y. Hsieh, C. Hsiung, S.-H. Hsu, K. Ide, R. Iden, S. Ikeda, H. Imafuku, R. Ishikawa, Y. Itoh, M. Iwaya, H-B. Jin, K. Jung, T. Kajita, I. Kaku, M. Kamiizumi, N. Kanda, H. Kato, T. Kato, R. Kawamoto, S. Kim, K. Kobayashi, K. Kohri, K. Kokeyama, K. Komori, A. K. H. Kong, T. Koyama, J. Kume, S. Kuroyanagi, S. Kuwahara, K. Kwak, S. Kwon, H. W. Lee, R. Lee, S. Lee, K. L. Li, L. C.-C. Lin, E. T. Lin, Y.-C. Lin, G. C. Liu, K. Maeda, M. Meyer-Conde, Y. Michimura, K. Mitsuhashi, O. Miyakawa, S. Miyoki, S. Morisaki, Y. Moriwaki, M. Murakoshi, K. Nakagaki, K. Nakamura, H. Nakano, T. Narikawa, L. Naticchioni, L. Nguyen Quynh, Y. Nishino, A. Nishizawa, K. Obayashi, M. Ohashi, M. Onishi, K. Oohara, S. Oshino, R. Ozaki, M. A. Page, K.-C. Pan, B.-J. Park, J. Park, F. E. Pena Arellano, N. Ruhama, S. Saha, K. Sakai, Y. Sakai, R. Sato, S. Sato, Y. Sato, Y. Sato, T. Sawada, Y. Sekiguchi, N. Sembo, L. Shao, Z.-H. Shi, R. Shimomura, H. Shinkai, S. Singh, K. Somiya, I. Song, H. Sotani, Y. Sudo, K. Suzuki, M. Suzuki, H. Tagoshi, K. Takada, H. Takahashi, R. Takahashi, A. Takamori, S. Takano, H. Takeda, K. Takeshita, M. Tamaki, K. Tanaka, S. J. Tanaka, A. Taruya, T. Tomaru, T. Tomura, S. Tsuchida, N. Uchikata, T. Uchiyama, T. Uehara, K. Ueno, T. Ushiba, H. Wang, T. Washimi, C. Wu, H. Wu, K. Yamamoto, T. Yamamoto, T. S. Yamamoto, R. Yamazaki, Y. Yang, S.-W. Yeh, J. Yokoyama, T. Yokozawa, H. Yuzurihara, Z.-C. Zhao, Z.-H. Zhu (KAGRA Collaboration) and Y.-M Kim | Identification of Noise-Associated Glitches in KAGRA O3GK with Hveto | To appear in Progress of Theoretical and Experimental Physics (PTEP), accepted June 2025 | null | null | null | gr-qc astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transient noise ("glitches") in gravitational wave detectors can mimic or obscure true signals, significantly reducing detection sensitivity. Identifying and excluding glitch-contaminated data segments is therefore crucial for enhancing the performance of gravitational-wave searches. We perform a noise analysis of the KAGRA data obtained during the O3GK observation. Our analysis is performed with hierarchical veto (Hveto) which identifies noises based on the statistical time correlation between the main channel and the auxiliary channels. A total of 2,531 noises were vetoed by 28 auxiliary channels with the configuration (i.e., signal-to-noise threshold set to 8) that we chose for Hveto. We identify vetoed events as glitches on the spectrogram via visual examination after plotting them with Q-transformation. By referring to the Gravity Spy project, we categorize 2,354 glitches into six types: blip, helix, scratchy, and scattered light, which correspond to those listed in Gravity Spy, and dot and line, which are not found in the Gravity Spy classification and are thus named based on their spectrogram morphology in KAGRA data. The remaining 177 glitches are determined not to belong to any of these six types. We show how the KAGRA glitch types are related to each subsystem of KAGRA. To investigate the possible correlation between the main channel and the round winner - an auxiliary channel statistically associated with the main channel for vetoing purposes - we visually examine the similarity or difference in the glitch pattern on the spectrogram. We compare the qualitative correlation found through visual examination with coherence, which is known to provide quantitative measurement for the correlation between the main channel and each auxiliary channel. Our comprehensive noise analysis will help improve the data quality of KAGRA by applying it to future KAGRA observation data. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:47:44 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Akutsu",
"T.",
"",
"KAGRA Collaboration"
],
[
"Ando",
"M.",
"",
"KAGRA Collaboration"
],
[
"Aoumi",
"M.",
"",
"KAGRA Collaboration"
],
[
"Araya",
"A.",
"",
"KAGRA Collaboration"
],
[
"Aso",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Baiotti",
"L.",
"",
"KAGRA Collaboration"
],
[
"Bajpai",
"R.",
"",
"KAGRA Collaboration"
],
[
"Cannon",
"K.",
"",
"KAGRA Collaboration"
],
[
"Chen",
"A. H. -Y.",
"",
"KAGRA Collaboration"
],
[
"Chen",
"D.",
"",
"KAGRA Collaboration"
],
[
"Chen",
"H.",
"",
"KAGRA Collaboration"
],
[
"Chiba",
"A.",
"",
"KAGRA Collaboration"
],
[
"Chou",
"C.",
"",
"KAGRA Collaboration"
],
[
"Eisenmann",
"M.",
"",
"KAGRA Collaboration"
],
[
"Endo",
"K.",
"",
"KAGRA Collaboration"
],
[
"Fujimori",
"T.",
"",
"KAGRA Collaboration"
],
[
"Garg",
"S.",
"",
"KAGRA Collaboration"
],
[
"Haba",
"D.",
"",
"KAGRA Collaboration"
],
[
"Haino",
"S.",
"",
"KAGRA Collaboration"
],
[
"Harada",
"R.",
"",
"KAGRA Collaboration"
],
[
"Hayakawa",
"H.",
"",
"KAGRA Collaboration"
],
[
"Hayama",
"K.",
"",
"KAGRA Collaboration"
],
[
"Fujii",
"S.",
"",
"KAGRA Collaboration"
],
[
"Himemoto",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Hirata",
"N.",
"",
"KAGRA Collaboration"
],
[
"Hirose",
"C.",
"",
"KAGRA Collaboration"
],
[
"Hsieh",
"H. -F.",
"",
"KAGRA Collaboration"
],
[
"Hsieh",
"H. -Y.",
"",
"KAGRA Collaboration"
],
[
"Hsiung",
"C.",
"",
"KAGRA Collaboration"
],
[
"Hsu",
"S. -H.",
"",
"KAGRA Collaboration"
],
[
"Ide",
"K.",
"",
"KAGRA Collaboration"
],
[
"Iden",
"R.",
"",
"KAGRA Collaboration"
],
[
"Ikeda",
"S.",
"",
"KAGRA Collaboration"
],
[
"Imafuku",
"H.",
"",
"KAGRA Collaboration"
],
[
"Ishikawa",
"R.",
"",
"KAGRA Collaboration"
],
[
"Itoh",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Iwaya",
"M.",
"",
"KAGRA Collaboration"
],
[
"Jin",
"H-B.",
"",
"KAGRA Collaboration"
],
[
"Jung",
"K.",
"",
"KAGRA Collaboration"
],
[
"Kajita",
"T.",
"",
"KAGRA Collaboration"
],
[
"Kaku",
"I.",
"",
"KAGRA Collaboration"
],
[
"Kamiizumi",
"M.",
"",
"KAGRA Collaboration"
],
[
"Kanda",
"N.",
"",
"KAGRA Collaboration"
],
[
"Kato",
"H.",
"",
"KAGRA Collaboration"
],
[
"Kato",
"T.",
"",
"KAGRA Collaboration"
],
[
"Kawamoto",
"R.",
"",
"KAGRA Collaboration"
],
[
"Kim",
"S.",
"",
"KAGRA Collaboration"
],
[
"Kobayashi",
"K.",
"",
"KAGRA Collaboration"
],
[
"Kohri",
"K.",
"",
"KAGRA Collaboration"
],
[
"Kokeyama",
"K.",
"",
"KAGRA Collaboration"
],
[
"Komori",
"K.",
"",
"KAGRA Collaboration"
],
[
"Kong",
"A. K. H.",
"",
"KAGRA Collaboration"
],
[
"Koyama",
"T.",
"",
"KAGRA Collaboration"
],
[
"Kume",
"J.",
"",
"KAGRA Collaboration"
],
[
"Kuroyanagi",
"S.",
"",
"KAGRA Collaboration"
],
[
"Kuwahara",
"S.",
"",
"KAGRA Collaboration"
],
[
"Kwak",
"K.",
"",
"KAGRA Collaboration"
],
[
"Kwon",
"S.",
"",
"KAGRA Collaboration"
],
[
"Lee",
"H. W.",
"",
"KAGRA Collaboration"
],
[
"Lee",
"R.",
"",
"KAGRA Collaboration"
],
[
"Lee",
"S.",
"",
"KAGRA Collaboration"
],
[
"Li",
"K. L.",
"",
"KAGRA Collaboration"
],
[
"Lin",
"L. C. -C.",
"",
"KAGRA Collaboration"
],
[
"Lin",
"E. T.",
"",
"KAGRA Collaboration"
],
[
"Lin",
"Y. -C.",
"",
"KAGRA Collaboration"
],
[
"Liu",
"G. C.",
"",
"KAGRA Collaboration"
],
[
"Maeda",
"K.",
"",
"KAGRA Collaboration"
],
[
"Meyer-Conde",
"M.",
"",
"KAGRA Collaboration"
],
[
"Michimura",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Mitsuhashi",
"K.",
"",
"KAGRA Collaboration"
],
[
"Miyakawa",
"O.",
"",
"KAGRA Collaboration"
],
[
"Miyoki",
"S.",
"",
"KAGRA Collaboration"
],
[
"Morisaki",
"S.",
"",
"KAGRA Collaboration"
],
[
"Moriwaki",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Murakoshi",
"M.",
"",
"KAGRA Collaboration"
],
[
"Nakagaki",
"K.",
"",
"KAGRA Collaboration"
],
[
"Nakamura",
"K.",
"",
"KAGRA Collaboration"
],
[
"Nakano",
"H.",
"",
"KAGRA Collaboration"
],
[
"Narikawa",
"T.",
"",
"KAGRA Collaboration"
],
[
"Naticchioni",
"L.",
"",
"KAGRA Collaboration"
],
[
"Quynh",
"L. Nguyen",
"",
"KAGRA Collaboration"
],
[
"Nishino",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Nishizawa",
"A.",
"",
"KAGRA Collaboration"
],
[
"Obayashi",
"K.",
"",
"KAGRA Collaboration"
],
[
"Ohashi",
"M.",
"",
"KAGRA Collaboration"
],
[
"Onishi",
"M.",
"",
"KAGRA Collaboration"
],
[
"Oohara",
"K.",
"",
"KAGRA Collaboration"
],
[
"Oshino",
"S.",
"",
"KAGRA Collaboration"
],
[
"Ozaki",
"R.",
"",
"KAGRA Collaboration"
],
[
"Page",
"M. A.",
"",
"KAGRA Collaboration"
],
[
"Pan",
"K. -C.",
"",
"KAGRA Collaboration"
],
[
"Park",
"B. -J.",
"",
"KAGRA Collaboration"
],
[
"Park",
"J.",
"",
"KAGRA Collaboration"
],
[
"Arellano",
"F. E. Pena",
"",
"KAGRA Collaboration"
],
[
"Ruhama",
"N.",
"",
"KAGRA Collaboration"
],
[
"Saha",
"S.",
"",
"KAGRA Collaboration"
],
[
"Sakai",
"K.",
"",
"KAGRA Collaboration"
],
[
"Sakai",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Sato",
"R.",
"",
"KAGRA Collaboration"
],
[
"Sato",
"S.",
"",
"KAGRA Collaboration"
],
[
"Sato",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Sato",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Sawada",
"T.",
"",
"KAGRA Collaboration"
],
[
"Sekiguchi",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Sembo",
"N.",
"",
"KAGRA Collaboration"
],
[
"Shao",
"L.",
"",
"KAGRA Collaboration"
],
[
"Shi",
"Z. -H.",
"",
"KAGRA Collaboration"
],
[
"Shimomura",
"R.",
"",
"KAGRA Collaboration"
],
[
"Shinkai",
"H.",
"",
"KAGRA Collaboration"
],
[
"Singh",
"S.",
"",
"KAGRA Collaboration"
],
[
"Somiya",
"K.",
"",
"KAGRA Collaboration"
],
[
"Song",
"I.",
"",
"KAGRA Collaboration"
],
[
"Sotani",
"H.",
"",
"KAGRA Collaboration"
],
[
"Sudo",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Suzuki",
"K.",
"",
"KAGRA Collaboration"
],
[
"Suzuki",
"M.",
"",
"KAGRA Collaboration"
],
[
"Tagoshi",
"H.",
"",
"KAGRA Collaboration"
],
[
"Takada",
"K.",
"",
"KAGRA Collaboration"
],
[
"Takahashi",
"H.",
"",
"KAGRA Collaboration"
],
[
"Takahashi",
"R.",
"",
"KAGRA Collaboration"
],
[
"Takamori",
"A.",
"",
"KAGRA Collaboration"
],
[
"Takano",
"S.",
"",
"KAGRA Collaboration"
],
[
"Takeda",
"H.",
"",
"KAGRA Collaboration"
],
[
"Takeshita",
"K.",
"",
"KAGRA Collaboration"
],
[
"Tamaki",
"M.",
"",
"KAGRA Collaboration"
],
[
"Tanaka",
"K.",
"",
"KAGRA Collaboration"
],
[
"Tanaka",
"S. J.",
"",
"KAGRA Collaboration"
],
[
"Taruya",
"A.",
"",
"KAGRA Collaboration"
],
[
"Tomaru",
"T.",
"",
"KAGRA Collaboration"
],
[
"Tomura",
"T.",
"",
"KAGRA Collaboration"
],
[
"Tsuchida",
"S.",
"",
"KAGRA Collaboration"
],
[
"Uchikata",
"N.",
"",
"KAGRA Collaboration"
],
[
"Uchiyama",
"T.",
"",
"KAGRA Collaboration"
],
[
"Uehara",
"T.",
"",
"KAGRA Collaboration"
],
[
"Ueno",
"K.",
"",
"KAGRA Collaboration"
],
[
"Ushiba",
"T.",
"",
"KAGRA Collaboration"
],
[
"Wang",
"H.",
"",
"KAGRA Collaboration"
],
[
"Washimi",
"T.",
"",
"KAGRA Collaboration"
],
[
"Wu",
"C.",
"",
"KAGRA Collaboration"
],
[
"Wu",
"H.",
"",
"KAGRA Collaboration"
],
[
"Yamamoto",
"K.",
"",
"KAGRA Collaboration"
],
[
"Yamamoto",
"T.",
"",
"KAGRA Collaboration"
],
[
"Yamamoto",
"T. S.",
"",
"KAGRA Collaboration"
],
[
"Yamazaki",
"R.",
"",
"KAGRA Collaboration"
],
[
"Yang",
"Y.",
"",
"KAGRA Collaboration"
],
[
"Yeh",
"S. -W.",
"",
"KAGRA Collaboration"
],
[
"Yokoyama",
"J.",
"",
"KAGRA Collaboration"
],
[
"Yokozawa",
"T.",
"",
"KAGRA Collaboration"
],
[
"Yuzurihara",
"H.",
"",
"KAGRA Collaboration"
],
[
"Zhao",
"Z. -C.",
"",
"KAGRA Collaboration"
],
[
"Zhu",
"Z. -H.",
"",
"KAGRA Collaboration"
],
[
"Kim",
"Y. -M",
""
]
] |
2506.21022 | Ze Wang | Ze Wang, Hao Chen, Benran Hu, Jiang Liu, Ximeng Sun, Jialian Wu, Yusheng Su, Xiaodong Yu, Emad Barsoum, Zicheng Liu | Instella-T2I: Pushing the Limits of 1D Discrete Latent Space Image Generation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image tokenization plays a critical role in reducing the computational demands of modeling high-resolution images, significantly improving the efficiency of image and multimodal understanding and generation. Recent advances in 1D latent spaces have reduced the number of tokens required by eliminating the need for a 2D grid structure. In this paper, we further advance compact discrete image representation by introducing 1D binary image latents. By representing each image as a sequence of binary vectors, rather than using traditional one-hot codebook tokens, our approach preserves high-resolution details while maintaining the compactness of 1D latents. To the best of our knowledge, our text-to-image models are the first to achieve competitive performance in both diffusion and auto-regressive generation using just 128 discrete tokens for images up to 1024x1024, demonstrating up to a 32-fold reduction in token numbers compared to standard VQ-VAEs. The proposed 1D binary latent space, coupled with simple model architectures, achieves marked improvements in speed training and inference speed. Our text-to-image models allow for a global batch size of 4096 on a single GPU node with 8 AMD MI300X GPUs, and the training can be completed within 200 GPU days. Our models achieve competitive performance compared to modern image generation models without any in-house private training data or post-training refinements, offering a scalable and efficient alternative to conventional tokenization methods. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:48:36 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Wang",
"Ze",
""
],
[
"Chen",
"Hao",
""
],
[
"Hu",
"Benran",
""
],
[
"Liu",
"Jiang",
""
],
[
"Sun",
"Ximeng",
""
],
[
"Wu",
"Jialian",
""
],
[
"Su",
"Yusheng",
""
],
[
"Yu",
"Xiaodong",
""
],
[
"Barsoum",
"Emad",
""
],
[
"Liu",
"Zicheng",
""
]
] |
2506.21023 | Mallory Flynn PhD | Mallory J Flynn and Paul Gustafson | AutoWMM and JAGStree -- R packages for Population Size Estimation on Relational Tree-Structured Data | 17 pages, 6 figures, 2 tables | null | null | null | stat.CO | http://creativecommons.org/licenses/by/4.0/ | The weighted multiplier method (WMM) is an extension of the traditional method of back-calculation method to estimate the size of a target population, which synthesizes available evidence from multiple subgroups of the target population with known counts and estimated proportions by leveraging the tree-structure inherent to the data. Hierarchical Bayesian models offer an alternative to modeling population size estimation on such a structure, but require non-trivial theoretical and practical knowledge to implement. While the theory underlying the WMM methodology may be more accessible to researchers in diverse fields, a barrier still exists in execution of this method, which requires significant computation. We develop two \texttt{R} packages to help facilitate population size estimation on trees using both the WMM and hierarchical Bayesian modeling; \textit{AutoWMM} simplifies WMM estimation for any general tree topology, and \textit{JAGStree} automates the creation of suitable JAGS MCMC modeling code for these same networks. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:56:07 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Flynn",
"Mallory J",
""
],
[
"Gustafson",
"Paul",
""
]
] |
2506.21024 | Mallory Flynn PhD | Mallory J Flynn and Paul Gustafson and Michael A. Irvine | Estimating the Number of Opioid Overdoses in British Columbia Using Relational Evidence with Tree Structure | 21 pages, 14 figures (8 in main text, 6 in supplementary), 6 tables | null | null | null | stat.AP | http://creativecommons.org/licenses/by/4.0/ | In many fields, populations of interest are hidden from data for a variety of reasons, though their magnitude remains important in determining resource allocation and appropriate policy. In public health and epidemiology, linkages or relationships between sources of data may exist due to intake structure of care providers, referrals, or other related health programming. These relationships often admit a tree structure, with the target population represented by the root, and paths from root-to-leaf representing pathways of care after a health event. In the Canadian province of British Columbia (BC), significant efforts have been made in creating an opioid overdose cohort, a tree-like linked data structure which tracks the movement of individuals along pathways of care after an overdose. In this application, the root node represents the target population, the total number of overdose events occurring in BC during the specified time period. We compare and contrast two methods of estimating the target population size - a weighted multiplier method based on back-calculating estimates from a number of paths and combining these estimates via a variance-minimizing weighted mean, and a fully Bayesian hierarchical model. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 05:56:36 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Flynn",
"Mallory J",
""
],
[
"Gustafson",
"Paul",
""
],
[
"Irvine",
"Michael A.",
""
]
] |
2506.21025 | Yifei Li | Weizhu Bao, Yifei Li, and Dongmin Wang | An energy-stable parametric finite element method for the Willmore flow in three dimensions | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work develops novel energy-stable parametric finite element methods (ES-PFEM) for the Willmore flow and curvature-dependent geometric gradient flows of surfaces in three dimensions. The key to achieving the energy stability lies in the use of two novel geometric identities: (i) a reformulated variational form of the normal velocity field, and (ii) incorporation of the temporal evolution of the mean curvature into the governing equations. These identities enable the derivation of a new variational formulation. By using the parametric finite element method, an implicit fully discrete scheme is subsequently developed, which maintains the energy dissipative property at the fully discrete level. Based on the ES-PFEM, comprehensive insights into the design of ES-PFEM for general curvature-dependent geometric gradient flows and a new understanding of mesh quality improvement in PFEM are provided. In particular, we develop the first PFEM for the Gauss curvature flow of surfaces. Furthermore, a tangential velocity control methodology is applied to improve the mesh quality and enhance the robustness of the proposed numerical method. Extensive numerical experiments confirm that the proposed method preserves energy dissipation properties and maintain good mesh quality in the surface evolution under the Willmore flow. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 06:06:34 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Bao",
"Weizhu",
""
],
[
"Li",
"Yifei",
""
],
[
"Wang",
"Dongmin",
""
]
] |
2506.21026 | Man Qiao | Man Qiao, Mingfeng Liu, Zongfei Lyu, Shuang Liu, Chao Yang, Dong Dong Shi, Fangxia An, Zhizheng Pan, Wenhao Liu, Binyang Liu, Run Wen, Yu Heng Zhang, Xian Zhong Zheng | Balmer Decrement and IRX Break in Tracing Dust Attenuation at Scales of Individual Star-forming Regions in NGC 628 | 20 pages, 20 figures, 1 table; accepted for publication in RAA | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the relationships between infrared excess (IRX=$L_{\rm IR}/L_{\rm UV}$) and Balmer decrement (${\rm H}\alpha/{\rm H}\beta$) as indicators of dust attenuation for 609 ${\rm {H\,{\small II}}}$ regions at scales of $\sim 50-200$ pc in NGC 628, utilizing data from AstroSat, James Webb Space Telescope (JWST) and Multi Unit Spectroscopic Explorer (MUSE). Our findings indicate that about three fifths of the sample ${\rm {H\,{\small II}}}$ regions reside within the regime occupied by local star-forming galaxies (SFGs) along the dust attenuation correlation described by their corresponding color excess parameters $E(B-V)_{\rm IRX} = 0.51\,E(B-V)_{{\rm H}\alpha/{\rm H}\beta}$. Nearly 27$\%$ of the sample exhibits $E(B-V)_{\rm IRX}> E(B-V)_{{\rm H}\alpha/{\rm H}\beta}$, while a small fraction ($\sim 13\%$) displays significantly lower $E(B-V)_{\rm IRX}$ compared to $E(B-V)_{{\rm H}\alpha/{\rm H}\beta}$. These results suggest that the correlation between the two dust attenuation indicators no longer holds for spatially resolved ${\rm {H\,{\small II}}}$ regions. Furthermore, the ratio of $E(B-V)_{\rm IRX}$ to $E(B-V)_{{\rm H}\alpha/{\rm H}\beta}$ remains unaffected by various physical parameters of the ${\rm {H\,{\small II}}}$ regions, including star formation rate (SFR), SFR surface density, infrared luminosity ($L_{\rm IR}$), $L_{\rm IR}$ surface density, stellar mass, gas-phase metallicity, circularized radius, and the distance to galactic center. We argue that the ratio is primarily influenced by the evolution of surrounding interstellar medium (ISM) of the star-forming regions, transitioning from an early dense and thick phase to the late blown-away stage. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 06:07:30 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Qiao",
"Man",
""
],
[
"Liu",
"Mingfeng",
""
],
[
"Lyu",
"Zongfei",
""
],
[
"Liu",
"Shuang",
""
],
[
"Yang",
"Chao",
""
],
[
"Shi",
"Dong Dong",
""
],
[
"An",
"Fangxia",
""
],
[
"Pan",
"Zhizheng",
""
],
[
"Liu",
"Wenhao",
""
],
[
"Liu",
"Binyang",
""
],
[
"Wen",
"Run",
""
],
[
"Zhang",
"Yu Heng",
""
],
[
"Zheng",
"Xian Zhong",
""
]
] |
2506.21027 | Fabio Sigrist | Hans R. K\"unsch and Fabio Sigrist | Simultaneous estimation of the effective reproduction number and the time series of daily infections: Application to Covid-19 | null | null | null | null | stat.ME stat.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The time varying effective reproduction number is an important parameter for communication and policy decisions during an epidemic. It is difficult to estimate because it depends on latent variables such as new infections and other characteristics of an epidemic which have to be inferred from available data. In this paper, we present new statistical methods for a popular model which defines the effective reproduction number based on self-exciting dynamics of new infections. Such a model is conceptually simple and less susceptible to misspecifications than more complicated multi-compartment models. In contrast to the state-of-the-art three-step estimation procedure of \citet{huisman2022estimation}, we present a coherent Bayesian method that approximates the joint posterior of daily new infections and reproduction numbers given the data using a novel Markov chain Monte Carlo (MCMC) algorithm. Comparing our method with that of \citet{huisman2022estimation}, both with daily confirmed cases from Switzerland in the Covid-19 epidemic and with simulated data, we find that our method is more accurate, especially near the beginning and end of the observation period. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 06:09:40 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Künsch",
"Hans R.",
""
],
[
"Sigrist",
"Fabio",
""
]
] |
2506.21028 | Feng Jiang | Feng Jiang, Mangal Prakash, Hehuan Ma, Jianyuan Deng, Yuzhi Guo, Amina Mollaysa, Tommaso Mansi, Rui Liao, Junzhou Huang | TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Molecular property prediction aims to learn representations that map chemical structures to functional properties. While multimodal learning has emerged as a powerful paradigm to learn molecular representations, prior works have largely overlooked textual and taxonomic information of molecules for representation learning. We introduce TRIDENT, a novel framework that integrates molecular SMILES, textual descriptions, and taxonomic functional annotations to learn rich molecular representations. To achieve this, we curate a comprehensive dataset of molecule-text pairs with structured, multi-level functional annotations. Instead of relying on conventional contrastive loss, TRIDENT employs a volume-based alignment objective to jointly align tri-modal features at the global level, enabling soft, geometry-aware alignment across modalities. Additionally, TRIDENT introduces a novel local alignment objective that captures detailed relationships between molecular substructures and their corresponding sub-textual descriptions. A momentum-based mechanism dynamically balances global and local alignment, enabling the model to learn both broad functional semantics and fine-grained structure-function mappings. TRIDENT achieves state-of-the-art performance on 11 downstream tasks, demonstrating the value of combining SMILES, textual, and taxonomic functional annotations for molecular property prediction. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 06:09:47 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Jiang",
"Feng",
""
],
[
"Prakash",
"Mangal",
""
],
[
"Ma",
"Hehuan",
""
],
[
"Deng",
"Jianyuan",
""
],
[
"Guo",
"Yuzhi",
""
],
[
"Mollaysa",
"Amina",
""
],
[
"Mansi",
"Tommaso",
""
],
[
"Liao",
"Rui",
""
],
[
"Huang",
"Junzhou",
""
]
] |
2506.21029 | Andrea Colombo | Andrea Colombo, Dario Carbonaro, Mingzi Zhang, Chi Shen, Ramtin Gharleghi, Ankush Kapoor, Claudio Chiastra, Nigel Jepson, Mark Webster, Susann Beier | Are Ultrathin Stents Optimal for Bifurcation Lesions? Insights from Computational Modelling of Provisional and DK-Crush Techniques | null | null | null | null | physics.med-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Complex coronary bifurcation lesions remain challenging in percutaneous coronary intervention, with stent design and deployment strategy influencing clinical outcomes. This study compares the mechanical and hemodynamic performance of the ultrathin-strut Orsiro and thin-strut Xience Sierra stent in Provisional Side Branch (PSB) and Double Kissing Crush (DKC) techniques. We used finite element analyses of bifurcation stent deployment to assess malapposition, ostium clearance, and arterial wall stress for both techniques. Computational fluid dynamics simulations quantified the luminal exposure to low Time-Averaged Endothelial Shear Stress (TAESS below 0.4 Pa) and high shear rates (above 1000 1/s). In PSB, Orsiro showed higher malapposition (13.0% vs 9.6%) but improved SB ostium clearance (77% vs 64%) and lower low-TAESS exposure (30.3% vs 33.6%) compared to Xience. Orsiro also produced higher arterial wall stresses, particularly during kissing balloon inflation. In DKC, differences in malapposition and ostium clearance diminished between stents, though Orsiro retained a hemodynamic advantage with lower low-TAESS (28.2% vs 36.3%).Stent design influenced outcomes more strongly in PSB, where anatomical interaction and platform-specific behavior impacted both structural and hemodynamic results. In DKC, procedural complexity minimized those differences, making the stenting technique the primary performance driver. Nonetheless, Orsiro consistently preserved more favorable flow conditions. These findings highlight the need to match device selection with lesion characteristics in PSB, while in DKC, optimizing procedural steps may have a greater impact than the choice of stent platform. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 06:09:57 GMT"
}
] | 2025-06-27T00:00:00 | [
[
"Colombo",
"Andrea",
""
],
[
"Carbonaro",
"Dario",
""
],
[
"Zhang",
"Mingzi",
""
],
[
"Shen",
"Chi",
""
],
[
"Gharleghi",
"Ramtin",
""
],
[
"Kapoor",
"Ankush",
""
],
[
"Chiastra",
"Claudio",
""
],
[
"Jepson",
"Nigel",
""
],
[
"Webster",
"Mark",
""
],
[
"Beier",
"Susann",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.