id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2505.22293 | Samuel Frontull | Samuel Frontull and Thomas Str\"ohle | Compensating for Data with Reasoning: Low-Resource Machine Translation with LLMs | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Large Language Models (LLMs) have demonstrated strong capabilities in multilingual machine translation, sometimes even outperforming traditional neural systems. However, previous research has highlighted the challenges of using LLMs, particularly with prompt engineering, for low-resource languages. In this work, we introduce Fragment-Shot Prompting, a novel in-context learning method that segments input and retrieves translation examples based on syntactic coverage, along with Pivoted Fragment-Shot, an extension that enables translation without direct parallel data. We evaluate these methods using GPT-3.5, GPT-4o, o1-mini, LLaMA-3.3, and DeepSeek-R1 for translation between Italian and two Ladin variants, revealing three key findings: (1) Fragment-Shot Prompting is effective for translating into and between the studied low-resource languages, with syntactic coverage positively correlating with translation quality; (2) Models with stronger reasoning abilities make more effective use of retrieved knowledge, generally produce better translations, and enable Pivoted Fragment-Shot to significantly improve translation quality between the Ladin variants; and (3) prompt engineering offers limited, if any, improvements when translating from a low-resource to a high-resource language, where zero-shot prompting already yields satisfactory results. We publicly release our code and the retrieval corpora. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:29:05 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Frontull",
"Samuel",
""
],
[
"Ströhle",
"Thomas",
""
]
] |
2505.22294 | Tymon Frelik | Tomasz Sobczak, Tymon Frelik | Novel pathways in $k$-contact geometry | null | null | null | null | math.DG nlin.SI | http://creativecommons.org/licenses/by/4.0/ | Our study of Goursat distributions originates new types of $k$-contact distributions and Lie systems with applications. In particular, families of generators for Goursat distributions on $\mathbb{R}^4, \mathbb{R}^5$ and $\mathbb{R}^6$ give rise to Lie systems and we characterise Goursat structures that are $k$-contact distributions. Our results are used to study the zero-trailer and other systems via Lie systems and $k$-contact manifolds. New ideas for the development of superposition rules via geometric structures and the characterisation of $k$-contact distributions are given and applied. Some relations of $k$-contact geometry with parabolic Cartan geometries are inspected. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:32:29 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Sobczak",
"Tomasz",
""
],
[
"Frelik",
"Tymon",
""
]
] |
2505.22295 | Giordano Scappucci | Lucas E. A. Stehouwer, Merrit P. Losert, Maia Rigot, Davide Degli Esposti, Sara Mart\'i-S\'anchez, Maximillian Rimbach-Russ, Jordi Arbiol, Mark Friesen, Giordano Scappucci | Engineering Ge profiles in Si/SiGe heterostructures for increased valley splitting | null | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Electron spin qubits in Si/SiGe quantum wells are limited by the small and variable energy separation of the conduction band valleys. While sharp quantum well interfaces are pursued to increase the valley splitting energy deterministically, here we explore an alternative approach to enhance the valley splitting on average. We grow increasingly thinner quantum wells with broad interfaces to controllably increase the overlap of the electron wave function with Ge atoms. In these quantum wells, comprehensive quantum Hall measurements of two-dimensional electron gases reveal a linear correlation between valley splitting and disorder. Benchmarked against quantum wells with sharp interfaces, we demonstrate enhanced valley splitting while maintaining a low-disorder potential environment. Simulations using the experimental Ge concentration profiles predict an average valley splitting in quantum dots that matches the enhancement observed in two-dimensional systems. Our results motivate the experimental realization of quantum dot spin qubits in these heterostructures. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:33:20 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Stehouwer",
"Lucas E. A.",
""
],
[
"Losert",
"Merrit P.",
""
],
[
"Rigot",
"Maia",
""
],
[
"Esposti",
"Davide Degli",
""
],
[
"Martí-Sánchez",
"Sara",
""
],
[
"Rimbach-Russ",
"Maximillian",
""
],
[
"Arbiol",
"Jordi",
""
],
[
"Friesen",
"Mark",
""
],
[
"Scappucci",
"Giordano",
""
]
] |
2505.22296 | Haosheng Zou | Haosheng Zou, Xiaowei Lv, Shousheng Jia, Xiangzheng Zhang | 360-LLaMA-Factory: Plug & Play Sequence Parallelism for Long Post-Training | code at https://github.com/Qihoo360/360-LLaMA-Factory | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Adding sequence parallelism into LLaMA-Factory, we open-sourced 360-LLaMA-Factory at https://github.com/Qihoo360/360-LLaMA-Factory. 360-LLaMA-Factory has received wide recognition and used in models such as Light-R1 arXiv:2503.10460, TinyR1 arXiv:2503.04872, Kaggle AIMO math models and also in large companies' training frameworks. This technical report delves deeper into the different sequence parallel modes behind 360-LLaMA-Factory and discusses our implementation insights. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:33:46 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zou",
"Haosheng",
""
],
[
"Lv",
"Xiaowei",
""
],
[
"Jia",
"Shousheng",
""
],
[
"Zhang",
"Xiangzheng",
""
]
] |
2505.22297 | Jiayu Zhao | Jiajun Yang, Xiaofeng Li, Linlin Yuan, Li Lao, Jiayu Zhao | Revealing the terahertz-laser velocity effect during air filamentation via travelling-wave-antenna model | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | During femtosecond laser filamentation in air, the velocity ratio (K) between the terahertz (THz) phase velocity and the laser group velocity plays a crucial role in THz waves generation. However, K is typically assumed to be unity and its impact has been long overlooked due to the more attention paid to the more easily controlled filament length. Here, we investigate the obscured contribution of K to the THz radiation characteristics by using the improved travelling-wave-antenna (TWA) model. It has been found that, under both single- and two-color laser pumping schemes, K significantly determines the far-field spatial distribution of forward or backward THz radiation, as well as a transition from Bessel- to Cherenkov-type THz emission patterns. These results establish the TWA model as a reliable theoretical tool for studying the mechanisms of THz beam shaping via the designed K. Moreover, for cases of K not being controlled, its value can also be inferred by the proposed TWA model, which could be an effective method to confirm whether the laser ionization front is superluminal or subluminal compared with the generated THz waves. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:36:05 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Yang",
"Jiajun",
""
],
[
"Li",
"Xiaofeng",
""
],
[
"Yuan",
"Linlin",
""
],
[
"Lao",
"Li",
""
],
[
"Zhao",
"Jiayu",
""
]
] |
2505.22298 | Yifan Lu | Yifan Lu, Jing Li, Yigeng Zhou, Yihui Zhang, Wenya Wang, Xiucheng Li, Meishan Zhang, Fangming Liu, Jun Yu, Min Zhang | Adaptive Detoxification: Safeguarding General Capabilities of LLMs through Toxicity-Aware Knowledge Editing | ACL 2025 Findings | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) exhibit impressive language capabilities but remain vulnerable to malicious prompts and jailbreaking attacks. Existing knowledge editing methods for LLM detoxification face two major challenges. First, they often rely on entity-specific localization, making them ineffective against adversarial inputs without explicit entities. Second, these methods suffer from over-editing, where detoxified models reject legitimate queries, compromising overall performance. In this paper, we propose ToxEdit, a toxicity-aware knowledge editing approach that dynamically detects toxic activation patterns during forward propagation. It then routes computations through adaptive inter-layer pathways to mitigate toxicity effectively. This design ensures precise toxicity mitigation while preserving LLMs' general capabilities. To more accurately assess over-editing, we also enhance the SafeEdit benchmark by incorporating instruction-following evaluation tasks. Experimental results on multiple LLMs demonstrate that our ToxEdit outperforms previous state-of-the-art methods in both detoxification performance and safeguarding general capabilities of LLMs. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:37:06 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Lu",
"Yifan",
""
],
[
"Li",
"Jing",
""
],
[
"Zhou",
"Yigeng",
""
],
[
"Zhang",
"Yihui",
""
],
[
"Wang",
"Wenya",
""
],
[
"Li",
"Xiucheng",
""
],
[
"Zhang",
"Meishan",
""
],
[
"Liu",
"Fangming",
""
],
[
"Yu",
"Jun",
""
],
[
"Zhang",
"Min",
""
]
] |
2505.22300 | Simon D\"oring | Radu Curticapean, Simon D\"oring, Daniel Neuen | Counting Small Induced Subgraphs: Scorpions Are Easy but Not Trivial | null | null | null | null | cs.CC cs.DS | http://creativecommons.org/licenses/by/4.0/ | We consider the parameterized problem $\#$IndSub$(\Phi)$ for fixed graph properties $\Phi$: Given a graph $G$ and an integer $k$, this problem asks to count the number of induced $k$-vertex subgraphs satisfying $\Phi$. D\"orfler et al. [Algorithmica 2022] and Roth et al. [SICOMP 2024] conjectured that $\#$IndSub$(\Phi)$ is $\#$W[1]-hard for all non-meager properties $\Phi$, i.e., properties that are nontrivial for infinitely many $k$. This conjecture has been confirmed for several restricted types of properties, including all hereditary properties [STOC 2022] and all edge-monotone properties [STOC 2024].
In this work, we refute this conjecture by showing that scorpion graphs, certain $k$-vertex graphs which were introduced more than 50 years ago in the context of the evasiveness conjecture, can be counted in time $O(n^4)$ for all $k$. A simple variant of this construction results in graph properties that achieve arbitrary intermediate complexity assuming ETH.
We formulate an updated conjecture on the complexity of $\#$IndSub$(\Phi)$ that correctly captures the complexity status of scorpions and related constructions. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:37:30 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Curticapean",
"Radu",
""
],
[
"Döring",
"Simon",
""
],
[
"Neuen",
"Daniel",
""
]
] |
2505.22301 | Malte Henkel | Malte Henkel, Stoimen Stoimenov | Ageing correlators from local scale-invariance | Latex2e, 1+13 pages, 2 figures, 1 table | null | null | null | cond-mat.stat-mech hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | For ageing systems with dynamical exponent $\mathpzc{z}=2$ and with the dominant noise coming from the thermal bath, the form of the two-time autocorrelator as well as the time-space form of the single-time correlator are derived from Schr\"odinger-invariance, generalised to non-equilibrium ageing. These findings reproduce the exact results in the $1D$ Glauber-Ising model at $T=0$ and the critical spherical model in $d>2$ dimensions. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:38:01 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Henkel",
"Malte",
""
],
[
"Stoimenov",
"Stoimen",
""
]
] |
2505.22302 | Triantafyllos Kormpakis | Triantafyllos Kormpakis, Manuel Linares, Jordi Jos\'e | Connecting the m-dots: accretion rates and thermonuclear burst recurrence times on neutron stars and white dwarfs | 15 pages, 4 figures, Accepted in MNRAS | null | 10.1093/mnras/staf853 | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | We present a compilation of observed recurrence times ($t_{\rm rec}$) and infer the corresponding local mass-accretion rates ($\dot m$) for type I X-ray bursts, milliHertz quasi-periodic oscillating sources and recurrent novae eruptions. We construct models of the $t_{\rm rec}-\dot m$ relation for accreting white dwarfs and neutron stars and find that both are roughly consistent with a global inverse linear relation, connecting for the first time thermonuclear runaways on neutron stars and white dwarfs. We find that theoretical models of pure He bursts are in agreement with the best $t_{\rm rec}$ measurements in ultra-compact X-ray binaries at low $\dot m$ (4U~$0614+09$ and 2S~0918-549).
We suggest that the transient Z source XTE~J1701-462 is a slow rotator, based on its mHz QPO properties. Finally, we discuss the implications for thermonuclear ignition and point out that the difference in eruption/burst energy ($E_{b_{WD}}/E_{b_{NS}}=2\times 10^4$) is consistent with the difference in area between neutron stars and white dwarfs $\left((R_{WD}/R_{NS})^2=4\times 10^4\right)$. We conclude that ignitions of thermonuclear shell flashes on neutron stars and white dwarfs depend primarily on the specific mass accretion rate and do not depend on the nature of the underlying compact object. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:38:43 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Kormpakis",
"Triantafyllos",
""
],
[
"Linares",
"Manuel",
""
],
[
"José",
"Jordi",
""
]
] |
2505.22303 | Micha{\l} Szczerbak | Grzegorz Wolny and Micha{\l} Szczerbak | Voice CMS: updating the knowledge base of a digital assistant through conversation | null | null | null | null | cs.HC cs.AI cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, we propose a solution based on a multi-agent LLM architecture and a voice user interface (VUI) designed to update the knowledge base of a digital assistant. Its usability is evaluated in comparison to a more traditional graphical content management system (CMS), with a focus on understanding the relationship between user preferences and the complexity of the information being provided. The findings demonstrate that, while the overall usability of the VUI is rated lower than the graphical interface, it is already preferred by users for less complex tasks. Furthermore, the quality of content entered through the VUI is comparable to that achieved with the graphical interface, even for highly complex tasks. Obtained qualitative results suggest that a hybrid interface combining the strengths of both approaches could address the key challenges identified during the experiment, such as reducing cognitive load through graphical feedback while maintaining the intuitive nature of voice-based interactions. This work highlights the potential of conversational interfaces as a viable and effective method for knowledge management in specific business contexts. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:40:37 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wolny",
"Grzegorz",
""
],
[
"Szczerbak",
"Michał",
""
]
] |
2505.22304 | Jiali Chen | Jiali Chen, Xusen Hei, HongFei Liu, Yuancheng Wei, Zikun Deng, Jiayuan Xie, Yi Cai, Li Qing | CADReview: Automatically Reviewing CAD Programs with Error Detection and Correction | ACL 2025 main conference | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computer-aided design (CAD) is crucial in prototyping 3D objects through geometric instructions (i.e., CAD programs). In practical design workflows, designers often engage in time-consuming reviews and refinements of these prototypes by comparing them with reference images. To bridge this gap, we introduce the CAD review task to automatically detect and correct potential errors, ensuring consistency between the constructed 3D objects and reference images. However, recent advanced multimodal large language models (MLLMs) struggle to recognize multiple geometric components and perform spatial geometric operations within the CAD program, leading to inaccurate reviews. In this paper, we propose the CAD program repairer (ReCAD) framework to effectively detect program errors and provide helpful feedback on error correction. Additionally, we create a dataset, CADReview, consisting of over 20K program-image pairs, with diverse errors for the CAD review task. Extensive experiments demonstrate that our ReCAD significantly outperforms existing MLLMs, which shows great potential in design applications. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:41:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Chen",
"Jiali",
""
],
[
"Hei",
"Xusen",
""
],
[
"Liu",
"HongFei",
""
],
[
"Wei",
"Yuancheng",
""
],
[
"Deng",
"Zikun",
""
],
[
"Xie",
"Jiayuan",
""
],
[
"Cai",
"Yi",
""
],
[
"Qing",
"Li",
""
]
] |
2505.22305 | Md Touhidul Islam | Md Touhidul Islam and Imran Kabir and Md Alimoor Reza and Syed Masum Billah | IKIWISI: An Interactive Visual Pattern Generator for Evaluating the Reliability of Vision-Language Models Without Ground Truth | Accepted at DIS'25 (Funchal, Portugal) | null | 10.1145/3715336.3735754 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present IKIWISI ("I Know It When I See It"), an interactive visual pattern generator for assessing vision-language models in video object recognition when ground truth is unavailable. IKIWISI transforms model outputs into a binary heatmap where green cells indicate object presence and red cells indicate object absence. This visualization leverages humans' innate pattern recognition abilities to evaluate model reliability. IKIWISI introduces "spy objects": adversarial instances users know are absent, to discern models hallucinating on nonexistent items. The tool functions as a cognitive audit mechanism, surfacing mismatches between human and machine perception by visualizing where models diverge from human understanding.
Our study with 15 participants found that users considered IKIWISI easy to use, made assessments that correlated with objective metrics when available, and reached informed conclusions by examining only a small fraction of heatmap cells. This approach not only complements traditional evaluation methods through visual assessment of model behavior with custom object sets, but also reveals opportunities for improving alignment between human perception and machine understanding in vision-language systems. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:41:08 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Islam",
"Md Touhidul",
""
],
[
"Kabir",
"Imran",
""
],
[
"Reza",
"Md Alimoor",
""
],
[
"Billah",
"Syed Masum",
""
]
] |
2505.22306 | Liyuan Wang | Zehua Chen, Yuyang Miao, Liyuan Wang, Luyun Fan, Danilo P. Mandic, Jun Zhu | Versatile Cardiovascular Signal Generation with a Unified Diffusion Transformer | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cardiovascular signals such as photoplethysmography (PPG), electrocardiography (ECG), and blood pressure (BP) are inherently correlated and complementary, together reflecting the health of cardiovascular system. However, their joint utilization in real-time monitoring is severely limited by diverse acquisition challenges from noisy wearable recordings to burdened invasive procedures. Here we propose UniCardio, a multi-modal diffusion transformer that reconstructs low-quality signals and synthesizes unrecorded signals in a unified generative framework. Its key innovations include a specialized model architecture to manage the signal modalities involved in generation tasks and a continual learning paradigm to incorporate varying modality combinations. By exploiting the complementary nature of cardiovascular signals, UniCardio clearly outperforms recent task-specific baselines in signal denoising, imputation, and translation. The generated signals match the performance of ground-truth signals in detecting abnormal health conditions and estimating vital signs, even in unseen domains, while ensuring interpretability for human experts. These advantages position UniCardio as a promising avenue for advancing AI-assisted healthcare. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:45:39 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Chen",
"Zehua",
""
],
[
"Miao",
"Yuyang",
""
],
[
"Wang",
"Liyuan",
""
],
[
"Fan",
"Luyun",
""
],
[
"Mandic",
"Danilo P.",
""
],
[
"Zhu",
"Jun",
""
]
] |
2505.22307 | Manuel Kl\"adtke | Manuel Kl\"adtke and Moritz Schulze Darup | On data usage and predictive behavior of data-driven predictive control with 1-norm regularization | This paper is a preprint of a contribution to the IEEE Control Systems Letters. 6 pages, 3 figures | null | null | null | eess.SY cs.SY math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the data usage and predictive behavior of data-driven predictive control (DPC) with 1-norm regularization. Our analysis enables the offline removal of unused data and facilitates a comparison between the identified symmetric structure and data usage against prior knowledge of the true system. This comparison helps assess the suitability of the DPC scheme for effective control. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:46:22 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Klädtke",
"Manuel",
""
],
[
"Darup",
"Moritz Schulze",
""
]
] |
2505.22308 | Zachary Shinnick | Zachary Shinnick, Liangze Jiang, Hemanth Saratchandran, Anton van den Hengel, Damien Teney | Transformers Pretrained on Procedural Data Contain Modular Structures for Algorithmic Reasoning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pretraining on large, semantically rich datasets is key for developing language models. Surprisingly, recent studies have shown that even synthetic data, generated procedurally through simple semantic-free algorithms, can yield some of the same benefits as natural language pretraining. It is unclear what specific capabilities such simple synthetic data instils in a model, where these capabilities reside in the architecture, and how they manifest within its weights. In this short paper, we identify several beneficial forms of procedural data, together with specific algorithmic reasoning skills that improve in small transformers. Our core finding is that different procedural rules instil distinct but complementary inductive structures in the model. With extensive ablations and partial-transfer experiments, we discover that these structures reside in different parts of the model. Attention layers often carry the most transferable information, but some pretraining rules impart useful structure to MLP blocks instead. Most interestingly, the structures induced by multiple rules can be composed to jointly reinforce multiple capabilities. These results suggest an exciting possibility of disentangling the acquisition of knowledge from reasoning in language models, with the goal of improving their robustness and data efficiency. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:50:09 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Shinnick",
"Zachary",
""
],
[
"Jiang",
"Liangze",
""
],
[
"Saratchandran",
"Hemanth",
""
],
[
"Hengel",
"Anton van den",
""
],
[
"Teney",
"Damien",
""
]
] |
2505.22309 | Xiangling Xu | Xiangling Xu, Marc-Olivier Renou, Igor Klep | Quantitative Tsirelson's Theorems via Approximate Schur's Lemma and Probabilistic Stampfli's Theorems | 23 pages, comments welcome | null | null | null | quant-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tsirelson showed that, in finite dimensions, quantum correlations generated by commuting observables--measurements associated with distinct parties whose operators mutually commute--are equivalent to those obtainable from measurements on separate tensor product factors. We generalize this foundational result to the setting of $\epsilon$-almost commuting observables, establishing two distinct quantitative approximate Tsirelson's theorems. Both theorems show that if a $d$-dimensional bipartite quantum strategy's observables $\epsilon$-almost commute, then they are within $O(\mathrm{poly}(d) \epsilon)$ (in operator norm) of observables from a genuine tensor product strategy. This provides a quantitative counterpart to the asymptotic result of [N. Ozawa, J. Math. Phys. 54, 032202 (2013)] and justifies the tensor product model as an effective model even when subsystem independence is only approximately satisfied.
Our theorems arise from two different but complementary formulations of almost commutation: (i) The first approach utilizes deterministic operator norm bounds relative to specific matrix generators (such as clock and shift matrices), leading to an approximate Schur's Lemma from which the first theorem directly follows. (ii) The second approach employs probabilistic bounds, requiring small commutators only on average against Haar-random single-qubit unitaries. This method yields two novel probabilistic Stampfli's theorems, quantifying distance to scalars based on probabilistic commutation, a result which may be of independent interest. These theorems set the basis for the second approximate Tsirelson's theorem. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:51:27 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Xu",
"Xiangling",
""
],
[
"Renou",
"Marc-Olivier",
""
],
[
"Klep",
"Igor",
""
]
] |
2505.22310 | Shoaib Ahmed Siddiqui | Shoaib Ahmed Siddiqui, Adrian Weller, David Krueger, Gintare Karolina Dziugaite, Michael Curtis Mozer, Eleni Triantafillou | From Dormant to Deleted: Tamper-Resistant Unlearning Through Weight-Space Regularization | null | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent unlearning methods for LLMs are vulnerable to relearning attacks: knowledge believed-to-be-unlearned re-emerges by fine-tuning on a small set of (even seemingly-unrelated) examples. We study this phenomenon in a controlled setting for example-level unlearning in vision classifiers. We make the surprising discovery that forget-set accuracy can recover from around 50% post-unlearning to nearly 100% with fine-tuning on just the retain set -- i.e., zero examples of the forget set. We observe this effect across a wide variety of unlearning methods, whereas for a model retrained from scratch excluding the forget set (gold standard), the accuracy remains at 50%. We observe that resistance to relearning attacks can be predicted by weight-space properties, specifically, $L_2$-distance and linear mode connectivity between the original and the unlearned model. Leveraging this insight, we propose a new class of methods that achieve state-of-the-art resistance to relearning attacks. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:53:08 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Siddiqui",
"Shoaib Ahmed",
""
],
[
"Weller",
"Adrian",
""
],
[
"Krueger",
"David",
""
],
[
"Dziugaite",
"Gintare Karolina",
""
],
[
"Mozer",
"Michael Curtis",
""
],
[
"Triantafillou",
"Eleni",
""
]
] |
2505.22311 | Feibo Jiang | Feibo Jiang, Cunhua Pan, Li Dong, Kezhi Wang, Octavia A. Dobre, and Merouane Debbah | From Large AI Models to Agentic AI: A Tutorial on Future Intelligent Communications | null | null | null | null | cs.AI cs.CY cs.NI eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the advent of 6G communications, intelligent communication systems face multiple challenges, including constrained perception and response capabilities, limited scalability, and low adaptability in dynamic environments. This tutorial provides a systematic introduction to the principles, design, and applications of Large Artificial Intelligence Models (LAMs) and Agentic AI technologies in intelligent communication systems, aiming to offer researchers a comprehensive overview of cutting-edge technologies and practical guidance. First, we outline the background of 6G communications, review the technological evolution from LAMs to Agentic AI, and clarify the tutorial's motivation and main contributions. Subsequently, we present a comprehensive review of the key components required for constructing LAMs. We further categorize LAMs and analyze their applicability, covering Large Language Models (LLMs), Large Vision Models (LVMs), Large Multimodal Models (LMMs), Large Reasoning Models (LRMs), and lightweight LAMs. Next, we propose a LAM-centric design paradigm tailored for communications, encompassing dataset construction and both internal and external learning approaches. Building upon this, we develop an LAM-based Agentic AI system for intelligent communications, clarifying its core components such as planners, knowledge bases, tools, and memory modules, as well as its interaction mechanisms. We also introduce a multi-agent framework with data retrieval, collaborative planning, and reflective evaluation for 6G. Subsequently, we provide a detailed overview of the applications of LAMs and Agentic AI in communication scenarios. Finally, we summarize the research challenges and future directions in current studies, aiming to support the development of efficient, secure, and sustainable next-generation intelligent communication systems. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:54:07 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Jiang",
"Feibo",
""
],
[
"Pan",
"Cunhua",
""
],
[
"Dong",
"Li",
""
],
[
"Wang",
"Kezhi",
""
],
[
"Dobre",
"Octavia A.",
""
],
[
"Debbah",
"Merouane",
""
]
] |
2505.22313 | Kaixuan Wei | Kaixuan Wei, Hector A. Jimenez-Romero, Hadi Amata, Jipeng Sun, Qiang Fu, Felix Heide, Wolfgang Heidrich | Large-Area Fabrication-aware Computational Diffractive Optics | null | null | null | null | physics.optics cs.CV cs.ET cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Differentiable optics, as an emerging paradigm that jointly optimizes optics and (optional) image processing algorithms, has made innovative optical designs possible across a broad range of applications. Many of these systems utilize diffractive optical components (DOEs) for holography, PSF engineering, or wavefront shaping. Existing approaches have, however, mostly remained limited to laboratory prototypes, owing to a large quality gap between simulation and manufactured devices. We aim at lifting the fundamental technical barriers to the practical use of learned diffractive optical systems. To this end, we propose a fabrication-aware design pipeline for diffractive optics fabricated by direct-write grayscale lithography followed by nano-imprinting replication, which is directly suited for inexpensive mass production of large area designs. We propose a super-resolved neural lithography model that can accurately predict the 3D geometry generated by the fabrication process. This model can be seamlessly integrated into existing differentiable optics frameworks, enabling fabrication-aware, end-to-end optimization of computational optical systems. To tackle the computational challenges, we also devise tensor-parallel compute framework centered on distributing large-scale FFT computation across many GPUs. As such, we demonstrate large scale diffractive optics designs up to 32.16 mm $\times$ 21.44 mm, simulated on grids of up to 128,640 by 85,760 feature points. We find adequate agreement between simulation and fabricated prototypes for applications such as holography and PSF engineering. We also achieve high image quality from an imaging system comprised only of a single DOE, with images processed only by a Wiener filter utilizing the simulation PSF. We believe our findings lift the fabrication limitations for real-world applications of diffractive optics and differentiable optical design. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:56:46 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wei",
"Kaixuan",
""
],
[
"Jimenez-Romero",
"Hector A.",
""
],
[
"Amata",
"Hadi",
""
],
[
"Sun",
"Jipeng",
""
],
[
"Fu",
"Qiang",
""
],
[
"Heide",
"Felix",
""
],
[
"Heidrich",
"Wolfgang",
""
]
] |
2505.22314 | Ilane Schroetter | Ilane Schroetter, Olivier Bern\'e, Emeric Bron, Felipe Alarcon, Paul Amiot, Edwin A. Bergin, Christiaan Boersma, Jan Cami, Gavin A. L. Coleman, Emmanuel Dartois, Asuncion Fuente, Javier R. Goicoechea, Emilie Habart, Thomas J. Haworth, Christine Joblin, Franck Le Petit, Takashi Onaka, Els Peeters, Markus R\"olling, Alexander G. G. M. Tielens, Marion Zannese | Solar C/O ratio in planet-forming gas at 1 au in a highly irradiated disk | 20 pages, 14 figures, submitted to Nature Astronomy | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The chemical composition of exoplanets is thought to be influenced by the composition of the disks in which they form. JWST observations have unveiled a variety of species in numerous nearby disks, showing significant variations in the C/O abundance ratio. However, little is known about the composition and C/O ratio of disks around young stars in clusters exposed to strong ultraviolet (UV) radiation from nearby massive stars, which are representative of the environments where most planetary systems form, including ours. We present JWST spectroscopy of d203-504, a young 0.7 $\rm M_{\odot}$ star in the Orion Nebula with a 30 au disk irradiated by nearby massive stars. These observations reveal spectroscopic signatures of CO, H$_2$O, CH$_3^+$, and PAHs. Water and CO are detected in absorption in the inner disk ($r\lesssim 1$ au), where the estimated gas-phase C/O ratio is 0.48, consistent with the Solar value and that of the Orion Nebula. In contrast, \ch{CH3+} and PAHs are found in the extended surface layers of the disk. These results suggest that gas in the inner disk is chemically shielded from UV radiation while the surface layers of the disk experience UV-induced chemistry, potentially depleting their carbon content. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:58:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Schroetter",
"Ilane",
""
],
[
"Berné",
"Olivier",
""
],
[
"Bron",
"Emeric",
""
],
[
"Alarcon",
"Felipe",
""
],
[
"Amiot",
"Paul",
""
],
[
"Bergin",
"Edwin A.",
""
],
[
"Boersma",
"Christiaan",
""
],
[
"Cami",
"Jan",
""
],
[
"Coleman",
"Gavin A. L.",
""
],
[
"Dartois",
"Emmanuel",
""
],
[
"Fuente",
"Asuncion",
""
],
[
"Goicoechea",
"Javier R.",
""
],
[
"Habart",
"Emilie",
""
],
[
"Haworth",
"Thomas J.",
""
],
[
"Joblin",
"Christine",
""
],
[
"Petit",
"Franck Le",
""
],
[
"Onaka",
"Takashi",
""
],
[
"Peeters",
"Els",
""
],
[
"Rölling",
"Markus",
""
],
[
"Tielens",
"Alexander G. G. M.",
""
],
[
"Zannese",
"Marion",
""
]
] |
2505.22315 | Adam B. Cahaya | Adam B. Cahaya | Anti-bosonic current in thermal spin pumping | null | null | null | null | cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate thermal spin pumping in gadolinium iron garnet, focusing on dynamics of antiferromagnetic magnon and its impact on spin and heat transport. Antiferromagnet has right-handed and left-handed magnon modes. Using a two-magnetic-lattice model, we derive temperature-dependent spin-wave spectra and thermal spin pumping constants, revealing distinct transport regimes governed by competing contribution of magnon modes. We show that competition of both modes induces a compensation temperature in thermally generated spin current. This theoretical framework deepens the understanding of spin-caloritronic effects in ferrimagnets, offering new perspectives for spintronic device optimization. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 12:59:37 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Cahaya",
"Adam B.",
""
]
] |
2505.22316 | Konrad \"Ozdemir | Konrad \"Ozdemir, Lukas Kirchdorfer, Keyvan Amiri Elyasi, Han van der Aa, Heiner Stuckenschmidt | Rethinking BPS: A Utility-Based Evaluation Framework | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Business process simulation (BPS) is a key tool for analyzing and optimizing organizational workflows, supporting decision-making by estimating the impact of process changes. The reliability of such estimates depends on the ability of a BPS model to accurately mimic the process under analysis, making rigorous accuracy evaluation essential. However, the state-of-the-art approach to evaluating BPS models has two key limitations. First, it treats simulation as a forecasting problem, testing whether models can predict unseen future events. This fails to assess how well a model captures the as-is process, particularly when process behavior changes from train to test period. Thus, it becomes difficult to determine whether poor results stem from an inaccurate model or the inherent complexity of the data, such as unpredictable drift. Second, the evaluation approach strongly relies on Earth Mover's Distance-based metrics, which can obscure temporal patterns and thus yield misleading conclusions about simulation quality. To address these issues, we propose a novel framework that evaluates simulation quality based on its ability to generate representative process behavior. Instead of comparing simulated logs to future real-world executions, we evaluate whether predictive process monitoring models trained on simulated data perform comparably to those trained on real data for downstream analysis tasks. Empirical results show that our framework not only helps identify sources of discrepancies but also distinguishes between model accuracy and data complexity, offering a more meaningful way to assess BPS quality. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:00:52 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Özdemir",
"Konrad",
""
],
[
"Kirchdorfer",
"Lukas",
""
],
[
"Elyasi",
"Keyvan Amiri",
""
],
[
"van der Aa",
"Han",
""
],
[
"Stuckenschmidt",
"Heiner",
""
]
] |
2505.22317 | Vladimir Zyuzin | Vladimir A. Zyuzin | Magnon thermal Hall effect in insulating antiferromagnets | null | null | null | null | cond-mat.mes-hall cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | In this paper we theoretically discuss thermal Hall effect of magnons in insulating N\'{e}el ordered antiferromagnets at zero external magnetic field. We show that non-zero thermal Hall effect requires absence of any symmetry between the two sublattices. The thermal Hall effect of magnons will be non-zero by a virtue of the spin-momentum splitting of the magnon spectrum due to the Dzyaloshinskii-Moriya interaction as well as anisotropic second-nearest exchange interaction different in the two sublattices, both corresponding to the broken symmetry. We construct a theoretical model in which an external electric field may change the symmetry of the antiferromagnetic system thus altering the thermal Hall effect of magnons. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:01:18 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zyuzin",
"Vladimir A.",
""
]
] |
2505.22318 | Anish Joishy | Ishwar B Balappanawar, Vamshi Krishna Bonagiri, Anish R Joishy, Manas Gaur, Krishnaprasad Thirunarayan, Ponnurangam Kumaraguru | If Pigs Could Fly... Can LLMs Logically Reason Through Counterfactuals? | 16 pages, 5 figures | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) demonstrate impressive reasoning capabilities in familiar contexts, but struggle when the context conflicts with their parametric knowledge. To investigate this phenomenon, we introduce CounterLogic, a dataset containing 1,800 examples across 9 logical schemas, explicitly designed to evaluate logical reasoning through counterfactual (hypothetical knowledge-conflicting) scenarios. Our systematic evaluation of 11 LLMs across 6 different datasets reveals a consistent performance degradation, with accuracies dropping by 27% on average when reasoning through counterfactual information. We propose Self-Segregate, a prompting method enabling metacognitive awareness (explicitly identifying knowledge conflicts) before reasoning. Our method dramatically narrows the average performance gaps from 27% to just 11%, while significantly increasing the overall accuracy (+7.5%). We discuss the implications of these findings and draw parallels to human cognitive processes, particularly on how humans disambiguate conflicting information during reasoning tasks. Our findings offer practical insights for understanding and enhancing LLMs reasoning capabilities in real-world applications, especially where models must logically reason independently of their factual knowledge. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:03:18 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Balappanawar",
"Ishwar B",
""
],
[
"Bonagiri",
"Vamshi Krishna",
""
],
[
"Joishy",
"Anish R",
""
],
[
"Gaur",
"Manas",
""
],
[
"Thirunarayan",
"Krishnaprasad",
""
],
[
"Kumaraguru",
"Ponnurangam",
""
]
] |
2505.22319 | Entong Zhao | Entong Zhao, Yu-Jun Liu, Ka Kwan Pak, Peng Ren, Mengbo Guo, Chengdong He, and Gyu-Boong Jo | Localization behavior in a Hermitian and non-Hermitian Raman lattice | 8 pages, 6 figures | null | null | null | cond-mat.quant-gas quant-ph | http://creativecommons.org/licenses/by/4.0/ | We introduce a flexible Raman lattice system for alakaline-earth like atoms to investigate localization behaviors in a quasi-periodic lattice with controllable non-Hermiticity. We demonstrate that critical phases and mobility edges can arise by adjusting spin-dependence of the incommensurate potentials in the Hermitian regime. With non-Hermiticity introduced by spin-selective atom loss, we find the localization behaviour in this system can be suppressed by dissipation. Our work provides insights into interplay between quasi-periodicity and non-Hermitian physics, offering a new perspective on localization phenomena. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:04:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zhao",
"Entong",
""
],
[
"Liu",
"Yu-Jun",
""
],
[
"Pak",
"Ka Kwan",
""
],
[
"Ren",
"Peng",
""
],
[
"Guo",
"Mengbo",
""
],
[
"He",
"Chengdong",
""
],
[
"Jo",
"Gyu-Boong",
""
]
] |
2505.22320 | Xudong Wang | Xudong Wang, Jian Zhu, Ruichen Zhang, Lei Feng, Dusit Niyato, Jiacheng Wang, Hongyang Du, Shiwen Mao, Zhu Han | Chain-of-Thought for Large Language Model-empowered Wireless Communications | 7 pages, 5 figures | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large language models (LLMs) have opened new possibilities for automated reasoning and decision-making in wireless networks. However, applying LLMs to wireless communications presents challenges such as limited capability in handling complex logic, generalization, and reasoning. Chain-of-Thought (CoT) prompting, which guides LLMs to generate explicit intermediate reasoning steps, has been shown to significantly improve LLM performance on complex tasks. Inspired by this, this paper explores the application potential of CoT-enhanced LLMs in wireless communications. Specifically, we first review the fundamental theory of CoT and summarize various types of CoT. We then survey key CoT and LLM techniques relevant to wireless communication and networking. Moreover, we introduce a multi-layer intent-driven CoT framework that bridges high-level user intent expressed in natural language with concrete wireless control actions. Our proposed framework sequentially parses and clusters intent, selects appropriate CoT reasoning modules via reinforcement learning, then generates interpretable control policies for system configuration. Using the unmanned aerial vehicle (UAV) network as a case study, we demonstrate that the proposed framework significantly outperforms a non-CoT baseline in both communication performance and quality of generated reasoning. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:04:48 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wang",
"Xudong",
""
],
[
"Zhu",
"Jian",
""
],
[
"Zhang",
"Ruichen",
""
],
[
"Feng",
"Lei",
""
],
[
"Niyato",
"Dusit",
""
],
[
"Wang",
"Jiacheng",
""
],
[
"Du",
"Hongyang",
""
],
[
"Mao",
"Shiwen",
""
],
[
"Han",
"Zhu",
""
]
] |
2505.22321 | Markus Holzmann | Antonio Arnal and Jussi Behrndt and Markus Holzmann and Petr Siegl | Generalized boundary triples for adjoint pairs with applications to non-self-adjoint Schr\"odinger operators | 18 pages | null | null | null | math.SP math.AP math.FA | http://creativecommons.org/licenses/by/4.0/ | We extend the notion of generalized boundary triples and their Weyl functions from extension theory of symmetric operators to adjoint pairs of operators, and we provide criteria on the boundary parameters to induce closed operators with a nonempty resolvent set. The abstract results are applied to Schr\"odinger operators with complex $L^p$-potentials on bounded and unbounded Lipschitz domains with compact boundaries. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:05:28 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Arnal",
"Antonio",
""
],
[
"Behrndt",
"Jussi",
""
],
[
"Holzmann",
"Markus",
""
],
[
"Siegl",
"Petr",
""
]
] |
2505.22322 | Zhengyu Fang | Zhengyu Fang, Zhimeng Jiang, Huiyuan Chen, Xiaoge Zhang, Kaiyu Tang, Xiao Li, Jing Li | A Closer Look on Memorization in Tabular Diffusion Model: A Data-Centric Perspective | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Diffusion models have shown strong performance in generating high-quality tabular data, but they carry privacy risks by reproducing exact training samples. While prior work focuses on dataset-level augmentation to reduce memorization, little is known about which individual samples contribute most. We present the first data-centric study of memorization dynamics in tabular diffusion models. We quantify memorization for each real sample based on how many generated samples are flagged as replicas, using a relative distance ratio. Our empirical analysis reveals a heavy-tailed distribution of memorization counts: a small subset of samples contributes disproportionately to leakage, confirmed via sample-removal experiments. To understand this, we divide real samples into top- and non-top-memorized groups and analyze their training-time behaviors. We track when each sample is first memorized and monitor per-epoch memorization intensity (AUC). Memorized samples are memorized slightly earlier and show stronger signals in early training. Based on these insights, we propose DynamicCut, a two-stage, model-agnostic mitigation method: (a) rank samples by epoch-wise intensity, (b) prune a tunable top fraction, and (c) retrain on the filtered dataset. Across multiple tabular datasets and models, DynamicCut reduces memorization with minimal impact on data diversity and downstream performance. It also complements augmentation-based defenses. Furthermore, DynamicCut enables cross-model transferability: high-ranked samples identified from one model (e.g., a diffusion model) are also effective for reducing memorization when removed from others, such as GANs and VAEs. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:06:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Fang",
"Zhengyu",
""
],
[
"Jiang",
"Zhimeng",
""
],
[
"Chen",
"Huiyuan",
""
],
[
"Zhang",
"Xiaoge",
""
],
[
"Tang",
"Kaiyu",
""
],
[
"Li",
"Xiao",
""
],
[
"Li",
"Jing",
""
]
] |
2505.22323 | Haolang Lu | Hongcan Guo, Haolang Lu, Guoshun Nan, Bolun Chu, Jialin Zhuang, Yuan Yang, Wenhao Che, Sicong Leng, Qimei Cui, and Xudong Jiang | Advancing Expert Specialization for Better MoE | 33pages, 6figures | null | null | null | cs.CL cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mixture-of-Experts (MoE) models enable efficient scaling of large language models (LLMs) by activating only a subset of experts per input. However, we observe that the commonly used auxiliary load balancing loss often leads to expert overlap and overly uniform routing, which hinders expert specialization and degrades overall performance during post-training. To address this, we propose a simple yet effective solution that introduces two complementary objectives: (1) an orthogonality loss to encourage experts to process distinct types of tokens, and (2) a variance loss to encourage more discriminative routing decisions. Gradient-level analysis demonstrates that these objectives are compatible with the existing auxiliary loss and contribute to optimizing the training process. Experimental results over various model architectures and across multiple benchmarks show that our method significantly enhances expert specialization. Notably, our method improves classic MoE baselines with auxiliary loss by up to 23.79%, while also maintaining load balancing in downstream tasks, without any architectural modifications or additional components. We will release our code to contribute to the community. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:09:47 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Guo",
"Hongcan",
""
],
[
"Lu",
"Haolang",
""
],
[
"Nan",
"Guoshun",
""
],
[
"Chu",
"Bolun",
""
],
[
"Zhuang",
"Jialin",
""
],
[
"Yang",
"Yuan",
""
],
[
"Che",
"Wenhao",
""
],
[
"Leng",
"Sicong",
""
],
[
"Cui",
"Qimei",
""
],
[
"Jiang",
"Xudong",
""
]
] |
2505.22324 | Adam B. Cahaya | Adam B. Cahaya | Indirect Magnetoelectric Coupling via Skew Scattering by Orbital Angular Momentum | null | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent experimental observations of exchange bias in the La$_{0.67}$Sr$_{0.33}$MnO$_{3}$/LaAlO$_{3}$/SrTiO$_{3}$ heterostructure, which lacks an intrinsic antiferromagnetic layer, have sparked theoretical investigations into the underlying mechanisms. While traditional theories suggest that exchange bias in spin valve structures is mediated by conduction electrons in metallic spacers, the transport properties of LaAlO$3$ are dominated by its valence electrons, raising new questions about the origin of this phenomenon. In this work, we propose a theoretical model where the electronic band structure of LaAlO$_3$ is treated as a valence band perturbed by skew scattering, which is sensitive to orbital angular momentum. Our analysis reveals a significant magnetoelectric effect at the La$_{0.67}$Sr$_{0.33}$MnO$_3$/LaAlO$_3$ interface, which induces a coupling between the interface magnetization and the electric field from two dimensional electron gas at LaAlO$_{3}$/SrTiO$_{3}$ interface. This magnetoelectric coupling is found to drive the observed exchange bias, highlighting the role of electric polarization in influencing the magnetic properties of the heterostructure. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:11:29 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Cahaya",
"Adam B.",
""
]
] |
2505.22325 | Antonio Caputo | Antonio Caputo | Vector-valued Graph Signal Processing | 18 pages | null | null | null | math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Classical graph signal processing (GSP) introduces methodologies for analyzing real or complex signals defined on graph domains, moving beyond classical uniform sampling techniques, such as the graph discrete Fourier transform (GDFT), employed as a pivotal tool for transforming graph signals into their spectral representation, enabling effective signal processing techniques such as filtering and denoising. In this paper, we propose a possible generalization of the set of signals and we study some properties of the more general set of vector-valued signals, which take values into any Banach space, and some properties of the fundamental operators of vertex-frequency analysis acting on these signals, such as the Fourier transform, the convolution operator and the translation operator. In particular, we show some estimates involving their operator norm as linear operators between Banach spaces and we establish a graph version of the classical primary uncertainty principle. We also show how these estimates depend on the choice of an orthonormal basis of $\mathbb{K}^N$. The importance of considering this general set of signals derives from the possibility to study multiple signals at the same time and the correlation existing between them, since multiple scalar signals can be modelled as a unique vector-valued signal. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:13:33 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Caputo",
"Antonio",
""
]
] |
2505.22326 | James Adams | James M. Adams and Gesine Reinert and Lukasz Szpruch and Carsten Maple and Andrew Elliott | Individualised Counterfactual Examples Using Conformal Prediction Intervals | Submitted to Conformal and Probabilistic Predictions With Applications (COPA) 2025 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Counterfactual explanations for black-box models aim to pr ovide insight into an algorithmic decision to its recipient. For a binary classification problem an individual counterfactual details which features might be changed for the model to infer the opposite class. High-dimensional feature spaces that are typical of machine learning classification models admit many possible counterfactual examples to a decision, and so it is important to identify additional criteria to select the most useful counterfactuals. In this paper, we explore the idea that the counterfactuals should be maximally informative when considering the knowledge of a specific individual about the underlying classifier. To quantify this information gain we explicitly model the knowledge of the individual, and assess the uncertainty of predictions which the individual makes by the width of a conformal prediction interval. Regions of feature space where the prediction interval is wide correspond to areas where the confidence in decision making is low, and an additional counterfactual example might be more informative to an individual. To explore and evaluate our individualised conformal prediction interval counterfactuals (CPICFs), first we present a synthetic data set on a hypercube which allows us to fully visualise the decision boundary, conformal intervals via three different methods, and resultant CPICFs. Second, in this synthetic data set we explore the impact of a single CPICF on the knowledge of an individual locally around the original query. Finally, in both our synthetic data set and a complex real world dataset with a combination of continuous and discrete variables, we measure the utility of these counterfactuals via data augmentation, testing the performance on a held out set. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:13:52 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Adams",
"James M.",
""
],
[
"Reinert",
"Gesine",
""
],
[
"Szpruch",
"Lukasz",
""
],
[
"Maple",
"Carsten",
""
],
[
"Elliott",
"Andrew",
""
]
] |
2505.22327 | Antonia Karamolegkou | Antonia Karamolegkou, Angana Borah, Eunjung Cho, Sagnik Ray Choudhury, Martina Galletti, Rajarshi Ghosh, Pranav Gupta, Oana Ignat, Priyanka Kargupta, Neema Kotonya, Hemank Lamba, Sun-Joo Lee, Arushi Mangla, Ishani Mondal, Deniz Nazarova, Poli Nemkova, Dina Pisarevskaya, Naquee Rizwan, Nazanin Sabri, Dominik Stammbach, Anna Steinberg, David Tom\'as, Steven R Wilson, Bowen Yi, Jessica H Zhu, Arkaitz Zubiaga, Anders S{\o}gaard, Alexander Fraser, Zhijing Jin, Rada Mihalcea, Joel R. Tetreault, Daryna Dementieva | NLP for Social Good: A Survey of Challenges, Opportunities, and Responsible Deployment | null | null | null | null | cs.CL cs.CY | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in large language models (LLMs) have unlocked unprecedented possibilities across a range of applications. However, as a community, we believe that the field of Natural Language Processing (NLP) has a growing need to approach deployment with greater intentionality and responsibility. In alignment with the broader vision of AI for Social Good (Toma\v{s}ev et al., 2020), this paper examines the role of NLP in addressing pressing societal challenges. Through a cross-disciplinary analysis of social goals and emerging risks, we highlight promising research directions and outline challenges that must be addressed to ensure responsible and equitable progress in NLP4SG research. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:14:44 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Karamolegkou",
"Antonia",
""
],
[
"Borah",
"Angana",
""
],
[
"Cho",
"Eunjung",
""
],
[
"Choudhury",
"Sagnik Ray",
""
],
[
"Galletti",
"Martina",
""
],
[
"Ghosh",
"Rajarshi",
""
],
[
"Gupta",
"Pranav",
""
],
[
"Ignat",
"Oana",
""
],
[
"Kargupta",
"Priyanka",
""
],
[
"Kotonya",
"Neema",
""
],
[
"Lamba",
"Hemank",
""
],
[
"Lee",
"Sun-Joo",
""
],
[
"Mangla",
"Arushi",
""
],
[
"Mondal",
"Ishani",
""
],
[
"Nazarova",
"Deniz",
""
],
[
"Nemkova",
"Poli",
""
],
[
"Pisarevskaya",
"Dina",
""
],
[
"Rizwan",
"Naquee",
""
],
[
"Sabri",
"Nazanin",
""
],
[
"Stammbach",
"Dominik",
""
],
[
"Steinberg",
"Anna",
""
],
[
"Tomás",
"David",
""
],
[
"Wilson",
"Steven R",
""
],
[
"Yi",
"Bowen",
""
],
[
"Zhu",
"Jessica H",
""
],
[
"Zubiaga",
"Arkaitz",
""
],
[
"Søgaard",
"Anders",
""
],
[
"Fraser",
"Alexander",
""
],
[
"Jin",
"Zhijing",
""
],
[
"Mihalcea",
"Rada",
""
],
[
"Tetreault",
"Joel R.",
""
],
[
"Dementieva",
"Daryna",
""
]
] |
2505.22328 | Yu. G. Naidyuk | O. E. Kvitnitskaya, L. Harnagea, O. D. Feia, D. V. Efremov, B. B\"uchner, Yu. G. Naidyuk | Observation of resistive switching and diode effect in the conductivity of TiTe2 point contacts | e.g.: 9 pages, 5 figures, accepted to "Low Temperature Physics" journal | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | We measured the I(V) and dV/dI(V) characteristics of TiTe2-based point contacts (PCs) from room to helium temperatures. Features indicating the emergence of a charge density wave (CDW) were detected. They represent symmetrical relatively V=0 maxima in dV/dI(V) around 150 mV at liquid helium temperatures, which disappear above 150 K, similar to the case of sister CDW compound TiSe2. Applying higher voltages above 200 mV, we observed resistive switching in TiTe2 PCs from a metallic-like low-resistance state to non-metallic type high-resistance state with a change of resistance by an order of magnitude. A unique diode-like effect was registered in "soft" TiTe2 PCs with hysteretic I(V) at negative voltage on TiTe2. Discovering the resistive switching and diode effect adds TiTe2 to the transition-metal dichalcogenides, which could be useful in developing non-volatile ReRAM and other upcoming nanotechnologies. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:15:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Kvitnitskaya",
"O. E.",
""
],
[
"Harnagea",
"L.",
""
],
[
"Feia",
"O. D.",
""
],
[
"Efremov",
"D. V.",
""
],
[
"Büchner",
"B.",
""
],
[
"Naidyuk",
"Yu. G.",
""
]
] |
2505.22329 | Lorenzo Lamberti | Luca Esposito, Lorenzo Lamberti, Dattatreya N. N. and Prosenjit Roy | Finsler $p$-Laplacian in domains becoming unbounded | null | null | null | null | math.AP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We study the asymptotic behavior of sequences of solutions, energies functionals, and the first eigenvalues associated with the Finsler $p$-Laplace operator, also known as the anisotropic $p$-Laplace operator on a sequence of bounded cylinders whose length tends to infinity. We prove that the solutions on the bounded cylinders converge to the solution on the cross-section, with a polynomial rate of convergence in the general case and exponential convergence in some special cases. We show that energies on finite cylinders, with the multiplication of a scaling factor, converge to the energy on the cross-section. Finally, we investigate the convergence of the first eigenvalue and, for a specific subclass, we provide the optimal convergence rate. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:16:53 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Esposito",
"Luca",
""
],
[
"Lamberti",
"Lorenzo",
""
],
[
"N.",
"Dattatreya N.",
""
],
[
"Roy",
"Prosenjit",
""
]
] |
2505.22330 | Dorian Bouchet | Ilya Starshynov, Maximilian Weimar, Lukas M. Rachbauer, G\"unther Hackl, Daniele Faccio, Stefan Rotter, Dorian Bouchet | Model-free estimation of the Cram\'er-Rao bound for deep-learning microscopy in complex media | null | Nature Photonics (2025) | 10.1038/s41566-025-01657-6 | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Artificial neural networks have become important tools to harness the complexity of disordered or random photonic systems. Recent applications include the recovery of information from light that has been scrambled during propagation through a complex scattering medium, especially in the challenging case where the deterministic input-output transmission matrix cannot be measured. This naturally raises the question of what the limit is that information theory imposes on this recovery process, and whether neural networks can actually reach this limit. To answer these questions, we introduce a model-free approach to calculate the Cram\'er-Rao bound, which sets the ultimate precision limit at which artificial neural networks can operate. As an example, we apply this approach in a proof-of-principle experiment using laser light propagating through a disordered medium, evidencing that a convolutional network approaches the ultimate precision limit in the challenging task of localizing a reflective target hidden behind a dynamically-fluctuating scattering medium. The model-free method introduced here is generally applicable to benchmark the performance of any deep-learning microscope, to drive algorithmic developments and to push the precision of metrology and imaging techniques to their ultimate limit. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:17:52 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Starshynov",
"Ilya",
""
],
[
"Weimar",
"Maximilian",
""
],
[
"Rachbauer",
"Lukas M.",
""
],
[
"Hackl",
"Günther",
""
],
[
"Faccio",
"Daniele",
""
],
[
"Rotter",
"Stefan",
""
],
[
"Bouchet",
"Dorian",
""
]
] |
2505.22331 | Hanxu Jiang | Hanxu Jiang, Haiyue Yu, Xiaotong Xie, Qi Gao, Jiang Jiang, Jianbin Sun | A Multi-output Gaussian Process Regression with Negative Transfer Mitigation for Generating Boundary Test Scenarios of Multi-UAV Systems | Article,29 pages,18 figures | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adaptive sampling based on Gaussian process regression (GPR) has already been applied with considerable success to generate boundary test scenarios for multi-UAV systems (MUS). One of the key techniques in such researches is leveraging the accurate prediction of the MUS performance through GPR in different test scenarios. Due to the potential correlations among the multiple MUS performance metrics, current researches commonly utilize a multi-output GPR (MOGPR) to model the multiple performance metrics simultaneously. This approach can achieve a more accurate prediction, rather than modeling each metric individually. However, MOGPR still suffers from negative transfer. When the feature of one output variable is incorrectly learned by another, the models training process will be negatively affected, leading to a decline in prediction performance. To solve this problem, this paper proposes a novel adaptive regularization approach into the conventional MOGPR training process. Unlike existing regularization approaches for mitigating negative transfer in MOGPR, our method penalizes the inconsistencies among output-specific characteristic parameters using adaptively adjustable regularization weights. This mechanism helps each set of output parameters avoid local optima. Consequently, it yields simultaneous improvements in predictive accuracy across all outputs. Finally, we validate our approach on a numerical case and on a boundary test scenario generation case for a MUS multi-objectives search task. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:17:59 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Jiang",
"Hanxu",
""
],
[
"Yu",
"Haiyue",
""
],
[
"Xie",
"Xiaotong",
""
],
[
"Gao",
"Qi",
""
],
[
"Jiang",
"Jiang",
""
],
[
"Sun",
"Jianbin",
""
]
] |
2505.22332 | Timo L\"ohr | Timo L\"ohr, Paul Hofman, Felix Mohr, Eyke H\"ullermeier | Credal Prediction based on Relative Likelihood | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Predictions in the form of sets of probability distributions, so-called credal sets, provide a suitable means to represent a learner's epistemic uncertainty. In this paper, we propose a theoretically grounded approach to credal prediction based on the statistical notion of relative likelihood: The target of prediction is the set of all (conditional) probability distributions produced by the collection of plausible models, namely those models whose relative likelihood exceeds a specified threshold. This threshold has an intuitive interpretation and allows for controlling the trade-off between correctness and precision of credal predictions. We tackle the problem of approximating credal sets defined in this way by means of suitably modified ensemble learning techniques. To validate our approach, we illustrate its effectiveness by experiments on benchmark datasets demonstrating superior uncertainty representation without compromising predictive performance. We also compare our method against several state-of-the-art baselines in credal prediction. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:20:20 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Löhr",
"Timo",
""
],
[
"Hofman",
"Paul",
""
],
[
"Mohr",
"Felix",
""
],
[
"Hüllermeier",
"Eyke",
""
]
] |
2505.22333 | Andreas Hochenegger | Klaus Altmann, Andreas Hochenegger, Frederik Witt | Acyclic toric sheaves | 10 pages | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $\mathcal E$ be a torus-linearised reflexive sheaf over a smooth projective toric variety. Based on a theorem of Perlman-Smith, we prove an explicit sufficient condition for $\mathcal E$ to be acyclic via Weil decorations. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:20:31 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Altmann",
"Klaus",
""
],
[
"Hochenegger",
"Andreas",
""
],
[
"Witt",
"Frederik",
""
]
] |
2505.22334 | Lai Wei | Lai Wei, Yuting Li, Kaipeng Zheng, Chen Wang, Yue Wang, Linghe Kong, Lichao Sun, Weiran Huang | Advancing Multimodal Reasoning via Reinforcement Learning with Cold Start | null | null | null | null | cs.CL cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in large language models (LLMs) have demonstrated impressive chain-of-thought reasoning capabilities, with reinforcement learning (RL) playing a crucial role in this progress. While "aha moment" patterns--where models exhibit self-correction through reflection--are often attributed to emergent properties from RL, we first demonstrate that these patterns exist in multimodal LLMs (MLLMs) prior to RL training but may not necessarily correlate with improved reasoning performance. Building on these insights, we present a comprehensive study on enhancing multimodal reasoning through a two-stage approach: (1) supervised fine-tuning (SFT) as a cold start with structured chain-of-thought reasoning patterns, followed by (2) reinforcement learning via GRPO to further refine these capabilities. Our extensive experiments show that this combined approach consistently outperforms both SFT-only and RL-only methods across challenging multimodal reasoning benchmarks. The resulting models achieve state-of-the-art performance among open-source MLLMs at both 3B and 7B scales, with our 7B model showing substantial improvements over base models (e.g., 66.3 %$\rightarrow$73.4 % on MathVista, 62.9 %$\rightarrow$70.4 % on We-Math) and our 3B model achieving performance competitive with several 7B models. Overall, this work provides practical guidance for building advanced multimodal reasoning models. Our code is available at https://github.com/waltonfuture/RL-with-Cold-Start. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:21:38 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wei",
"Lai",
""
],
[
"Li",
"Yuting",
""
],
[
"Zheng",
"Kaipeng",
""
],
[
"Wang",
"Chen",
""
],
[
"Wang",
"Yue",
""
],
[
"Kong",
"Linghe",
""
],
[
"Sun",
"Lichao",
""
],
[
"Huang",
"Weiran",
""
]
] |
2505.22335 | Wancai Zheng | Wancai Zheng, Linlin Ou, Jiajie He, Libo Zhou, Xinyi Yu, Yan Wei | UP-SLAM: Adaptively Structured Gaussian SLAM with Uncertainty Prediction in Dynamic Environments | null | null | null | null | cs.RO cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent 3D Gaussian Splatting (3DGS) techniques for Visual Simultaneous Localization and Mapping (SLAM) have significantly progressed in tracking and high-fidelity mapping. However, their sequential optimization framework and sensitivity to dynamic objects limit real-time performance and robustness in real-world scenarios. We present UP-SLAM, a real-time RGB-D SLAM system for dynamic environments that decouples tracking and mapping through a parallelized framework. A probabilistic octree is employed to manage Gaussian primitives adaptively, enabling efficient initialization and pruning without hand-crafted thresholds. To robustly filter dynamic regions during tracking, we propose a training-free uncertainty estimator that fuses multi-modal residuals to estimate per-pixel motion uncertainty, achieving open-set dynamic object handling without reliance on semantic labels. Furthermore, a temporal encoder is designed to enhance rendering quality. Concurrently, low-dimensional features are efficiently transformed via a shallow multilayer perceptron to construct DINO features, which are then employed to enrich the Gaussian field and improve the robustness of uncertainty prediction. Extensive experiments on multiple challenging datasets suggest that UP-SLAM outperforms state-of-the-art methods in both localization accuracy (by 59.8%) and rendering quality (by 4.57 dB PSNR), while maintaining real-time performance and producing reusable, artifact-free static maps in dynamic environments.The project: https://aczheng-cai.github.io/up_slam.github.io/ | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:23:16 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zheng",
"Wancai",
""
],
[
"Ou",
"Linlin",
""
],
[
"He",
"Jiajie",
""
],
[
"Zhou",
"Libo",
""
],
[
"Yu",
"Xinyi",
""
],
[
"Wei",
"Yan",
""
]
] |
2505.22336 | Robert Connelly | Robert Connelly, Zhen Zhang | Spiderwebs on the Sphere and an Isoperimetric Theorem | null | null | null | null | math.MG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Here we present a rigidity result in a global (semi-global, homotopy) setting for a restrictive class of polytopes, those that can be inscribed in a unit sphere, with some additional conditions. The proof of the rigidity result for cabled frameworks on the surface of the sphere uses classical isoperimetric ideas. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:23:32 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Connelly",
"Robert",
""
],
[
"Zhang",
"Zhen",
""
]
] |
2505.22337 | Samara Ghrer | Samara Ghrer, Christophe Godin, Stefanie Wuhrer | Learning to Infer Parameterized Representations of Plants from 3D Scans | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reconstructing faithfully the 3D architecture of plants from unstructured observations is a challenging task. Plants frequently contain numerous organs, organized in branching systems in more or less complex spatial networks, leading to specific computational issues due to self-occlusion or spatial proximity between organs. Existing works either consider inverse modeling where the aim is to recover the procedural rules that allow to simulate virtual plants, or focus on specific tasks such as segmentation or skeletonization. We propose a unified approach that, given a 3D scan of a plant, allows to infer a parameterized representation of the plant. This representation describes the plant's branching structure, contains parametric information for each plant organ, and can therefore be used directly in a variety of tasks. In this data-driven approach, we train a recursive neural network with virtual plants generated using an L-systems-based procedural model. After training, the network allows to infer a parametric tree-like representation based on an input 3D point cloud. Our method is applicable to any plant that can be represented as binary axial tree. We evaluate our approach on Chenopodium Album plants, using experiments on synthetic plants to show that our unified framework allows for different tasks including reconstruction, segmentation and skeletonization, while achieving results on-par with state-of-the-art for each task. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:23:48 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ghrer",
"Samara",
""
],
[
"Godin",
"Christophe",
""
],
[
"Wuhrer",
"Stefanie",
""
]
] |
2505.22338 | Chaoyun Zhang | Hanyang Wang, Lu Wang, Chaoyun Zhang, Tianjun Mao, Si Qin, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang | Text2Grad: Reinforcement Learning from Natural Language Feedback | The code for our method is available at https://github.com/microsoft/Text2Grad | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional RLHF optimizes language models with coarse, scalar rewards that mask the fine-grained reasons behind success or failure, leading to slow and opaque learning. Recent work augments RL with textual critiques through prompting or reflection, improving interpretability but leaving model parameters untouched. We introduce Text2Grad, a reinforcement-learning paradigm that turns free-form textual feedback into span-level gradients. Given human (or programmatic) critiques, Text2Grad aligns each feedback phrase with the relevant token spans, converts these alignments into differentiable reward signals, and performs gradient updates that directly refine the offending portions of the model's policy. This yields precise, feedback-conditioned adjustments instead of global nudges. Text2Grad is realized through three components: (1) a high-quality feedback-annotation pipeline that pairs critiques with token spans; (2) a fine-grained reward model that predicts span-level reward on answer while generating explanatory critiques; and (3) a span-level policy optimizer that back-propagates natural-language gradients. Across summarization, code generation, and question answering, Text2Grad consistently surpasses scalar-reward RL and prompt-only baselines, providing both higher task metrics and richer interpretability. Our results demonstrate that natural-language feedback, when converted to gradients, is a powerful signal for fine-grained policy optimization. The code for our method is available at https://github.com/microsoft/Text2Grad | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:23:49 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wang",
"Hanyang",
""
],
[
"Wang",
"Lu",
""
],
[
"Zhang",
"Chaoyun",
""
],
[
"Mao",
"Tianjun",
""
],
[
"Qin",
"Si",
""
],
[
"Lin",
"Qingwei",
""
],
[
"Rajmohan",
"Saravan",
""
],
[
"Zhang",
"Dongmei",
""
]
] |
2505.22339 | Mengru Guo | Mengru Guo and Yang Jiao | The Dirichlet problem for Hessian quotient type curvature equations in Minkowski space | null | null | null | null | math.AP math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider the Dirichlet problem for a class of prescribed Hessian quotient type curvature equations in Minkowski space. By establishing the a priori C2 estimates, we prove the existence of smooth spacelike hypersurfaces with a class of prescribed Hessian quotient type curvature and homogeneous boundary data in the non-degenerate case. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:24:03 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Guo",
"Mengru",
""
],
[
"Jiao",
"Yang",
""
]
] |
2505.22340 | Phan Th\`anh Nam | Emanuela L. Giacomelli, Christian Hainzl, Phan Th\`anh Nam, Robert Seiringer | The Huang-Yang conjecture for the low-density Fermi gas | 65 pages | null | null | null | math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Our work establishes a three-term asymptotic expansion of the ground state energy of a dilute gas of spin $1/2$ fermions with repulsive short-range interactions, validating a formula predicted by Huang and Yang in 1957. The formula is universal in the sense that it holds for a large class of interaction potentials and depends on those only via their scattering length. We have recently proved an upper bound on the ground state energy of the desired form, and the present work completes the program by proving the matching lower bound. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:25:07 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Giacomelli",
"Emanuela L.",
""
],
[
"Hainzl",
"Christian",
""
],
[
"Nam",
"Phan Thành",
""
],
[
"Seiringer",
"Robert",
""
]
] |
2505.22341 | P{\i}nar Kirezli | P{\i}nar Kirezli and Do\u{g}ukan Ta\c{s}er | Black Bounces in $f(Q)$ Gravity with Magnetic Source | 25 pages, 4 figures | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | In this study, the source of the black bounce is discussed in the context of $f(Q)$ theory. A body of research has been dedicated to the study of symmetric black bounce solutions that are generated by a combination of a scalar field with a non-zero potential and a magnetic charge within the framework of non-linear electrodynamics. As exact solutions are not obtained, the metric functions of Simpson-Visser and Bardeen type of black bounce are studied in the field equations. black bounce solutions are obtained by violating at least one energy condition for the Simpson-Visser and Bardeen type for an ordinary scalar field in $f(Q)$ gravity, which is only achieved by a phantom scalar field in general relativity. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:26:10 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Kirezli",
"Pınar",
""
],
[
"Taşer",
"Doğukan",
""
]
] |
2505.22342 | Shreyank N Gowda | Shriram M S, Xinyue Hao, Shihao Hou, Yang Lu, Laura Sevilla-Lara, Anurag Arnab, Shreyank N Gowda | Progressive Data Dropout: An Embarrassingly Simple Approach to Faster Training | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The success of the machine learning field has reliably depended on training on large datasets. While effective, this trend comes at an extraordinary cost. This is due to two deeply intertwined factors: the size of models and the size of datasets. While promising research efforts focus on reducing the size of models, the other half of the equation remains fairly mysterious. Indeed, it is surprising that the standard approach to training remains to iterate over and over, uniformly sampling the training dataset. In this paper we explore a series of alternative training paradigms that leverage insights from hard-data-mining and dropout, simple enough to implement and use that can become the new training standard. The proposed Progressive Data Dropout reduces the number of effective epochs to as little as 12.4% of the baseline. This savings actually do not come at any cost for accuracy. Surprisingly, the proposed method improves accuracy by up to 4.82%. Our approach requires no changes to model architecture or optimizer, and can be applied across standard training pipelines, thus posing an excellent opportunity for wide adoption. Code can be found here: https://github.com/bazyagami/LearningWithRevision | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:26:52 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"S",
"Shriram M",
""
],
[
"Hao",
"Xinyue",
""
],
[
"Hou",
"Shihao",
""
],
[
"Lu",
"Yang",
""
],
[
"Sevilla-Lara",
"Laura",
""
],
[
"Arnab",
"Anurag",
""
],
[
"Gowda",
"Shreyank N",
""
]
] |
2505.22343 | Zhonghao Lyu | Zhonghao Lyu, Yulan Gao, Junting Chen, Hongyang Du, Jie Xu, Kaibin Huang, and Dong In Kim | Empowering Intelligent Low-altitude Economy with Large AI Model Deployment | null | null | null | null | eess.SP cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Low-altitude economy (LAE) represents an emerging economic paradigm that redefines commercial and social aerial activities. Large artificial intelligence models (LAIMs) offer transformative potential to further enhance the intelligence of LAE services. However, deploying LAIMs in LAE poses several challenges, including the significant gap between their computational/storage demands and the limited onboard resources of LAE entities, the mismatch between lab-trained LAIMs and dynamic physical environments, and the inefficiencies of traditional decoupled designs for sensing, communication, and computation. To address these issues, we first propose a hierarchical system architecture tailored for LAIM deployment and present representative LAE application scenarios. Next, we explore key enabling techniques that facilitate the mutual co-evolution of LAIMs and low-altitude systems, and introduce a task-oriented execution pipeline for scalable and adaptive service delivery. Then, the proposed framework is validated through real-world case studies. Finally, we outline open challenges to inspire future research. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:27:07 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Lyu",
"Zhonghao",
""
],
[
"Gao",
"Yulan",
""
],
[
"Chen",
"Junting",
""
],
[
"Du",
"Hongyang",
""
],
[
"Xu",
"Jie",
""
],
[
"Huang",
"Kaibin",
""
],
[
"Kim",
"Dong In",
""
]
] |
2505.22344 | Nikhil Behari | Nikhil Behari, Aaron Young, Akshat Dave, Ramesh Raskar | Task-Driven Implicit Representations for Automated Design of LiDAR Systems | null | null | null | null | cs.CV cs.RO | http://creativecommons.org/licenses/by/4.0/ | Imaging system design is a complex, time-consuming, and largely manual process; LiDAR design, ubiquitous in mobile devices, autonomous vehicles, and aerial imaging platforms, adds further complexity through unique spatial and temporal sampling requirements. In this work, we propose a framework for automated, task-driven LiDAR system design under arbitrary constraints. To achieve this, we represent LiDAR configurations in a continuous six-dimensional design space and learn task-specific implicit densities in this space via flow-based generative modeling. We then synthesize new LiDAR systems by modeling sensors as parametric distributions in 6D space and fitting these distributions to our learned implicit density using expectation-maximization, enabling efficient, constraint-aware LiDAR system design. We validate our method on diverse tasks in 3D vision, enabling automated LiDAR system design across real-world-inspired applications in face scanning, robotic tracking, and object detection. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:27:42 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Behari",
"Nikhil",
""
],
[
"Young",
"Aaron",
""
],
[
"Dave",
"Akshat",
""
],
[
"Raskar",
"Ramesh",
""
]
] |
2505.22345 | Alexandre Benatti | Alexandre Benatti, Roberto M. Cesar Jr., Luciano da F. Costa | A Systematic Approach for Studying How Topological Measurements Respond to Complex Networks Modifications | 38 pages and 18 figures | null | null | null | cs.SI physics.soc-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Different types of graphs and complex networks have been characterized, analyzed, and modeled based on measurements of their respective topology. However, the available networks may constitute approximations of the original structure as a consequence of sampling incompleteness, noise, and/or error in the representation of that structure. Therefore, it becomes of particular interest to quantify how successive modifications may impact a set of adopted topological measurements, and how respectively undergone changes can be interrelated, which has been addressed in this paper by considering similarity networks and hierarchical clustering approaches. These studies are developed respectively to several topological measurements (accessibility, degree, hierarchical degree, clustering coefficient, betweenness centrality, assortativity, and average shortest path) calculated from complex networks of three main types (Erd\H{o}s-R\'enyi, Barab\'asi-Albert, and geographical) with varying sizes or subjected to progressive edge removal or rewiring. The coincidence similarity index, which can implement particularly strict comparisons, is adopted for two main purposes: to quantify and visualize how the considered topological measurements respond to the considered network alterations and to represent hierarchically the relationships between the observed changes undergone by the considered topological measurements. Several results are reported and discussed, including the identification of three types of topological changes taking place as a consequence of the modifications. In addition, the changes observed for the Erd\H{o}s-R\'enyi and Barab\'asi-Albert networks resulted mutually more similarly affected by topological changes than for the geometrical networks. The latter type of network has been identified to have more heterogeneous topological features than the other two types of networks. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:27:49 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Benatti",
"Alexandre",
""
],
[
"Cesar",
"Roberto M.",
"Jr."
],
[
"Costa",
"Luciano da F.",
""
]
] |
2505.22346 | Poulomee Ghosh | Poulomee Ghosh and Shubhendu Bhasin | State Constrained Model Reference Adaptive Control with Input Amplitude and Rate Limits | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper proposes a robust model reference adaptive controller (MRAC) for uncertain multi-input multi-output (MIMO) linear time-invariant (LTI) plants with user-defined constraints on the plant states, input amplitude, and input rate. The proposed two-layer barrier Lyapunov function (BLF)-based control design considers the input and the input rate as states that are constrained using two BLFs in the first layer, while another BLF in the second layer constrains the plant states. The adaptive control law ensures that the plant states, input amplitude, and input rate remain within the user-defined safe sets despite unmatched bounded disturbances. Sufficient conditions for the existence of a feasible control policy are also provided. To the best of the authors' knowledge, this is the first optimization-free method that imposes user-defined constraints on the state, input, and input rate and also provides verifiable feasibility conditions in the presence of parametric uncertainties and disturbances. Simulation results demonstrate the effectiveness of the proposed algorithm. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:28:39 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ghosh",
"Poulomee",
""
],
[
"Bhasin",
"Shubhendu",
""
]
] |
2505.22347 | Gleb Koshevoy | Gleb Koshevoy and Vadim Schechtman | Bruhat operads | 18 pages, 2 figures | null | null | null | math.CO math.CT | http://creativecommons.org/licenses/by/4.0/ | We describe some planar operads build from the higher Bruhat orders. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:30:28 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Koshevoy",
"Gleb",
""
],
[
"Schechtman",
"Vadim",
""
]
] |
2505.22348 | Zhijian Shi | Zhijian Shi, Shengjie Xu, Jianfeng Wang, Yi Du, Weichang Hao | Emergence of Diverse Topological States in Ge Doped MnBi2Te4 | 16 pages, 5 figures | null | null | null | cond-mat.mtrl-sci cond-mat.other physics.comp-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As an ideal platform for studying interplays between symmetry, topology and magnetism, the magnetic topological insulator (MTI) MnBi2Te4 has attracted extensive attentions. However, its strong n-type intrinsic defects hinder the realizations of exotic phenomena. Stimulated by recent discoveries that Ge doping can efficiently tune the position of Fermi level, here we systematically investigate the band evolution and topological phase diagram with doping concentration from MTI MnBi2Te4 to strong topological insulator GeBi2Te4. Different from magnetically doped Bi2Se3, the topology here is determined by competition of two band inversions arising from band folding of two time-reversal invariant momenta between antiferromagnetic and nonmagnetic/ferromagnetic unit cells. By employing a band momentum mapping method, besides the known MTI phase, remarkably, we find two classes of magnetic Dirac semimetal phases at antiferromagnetic state, two classes of Weyl semimetal phases at ferromagnetic state, and an intermediate trivial state at different doping regions. Interestingly, the trivial state can be tuned into a Weyl phase with two coexisting band inversions and extraordinarily long Fermi arcs by a small strain. Our work reveals diverse topological states with intrinsic quantum phenomena can be achieved with great potential for designing future electronic devices. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:30:31 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Shi",
"Zhijian",
""
],
[
"Xu",
"Shengjie",
""
],
[
"Wang",
"Jianfeng",
""
],
[
"Du",
"Yi",
""
],
[
"Hao",
"Weichang",
""
]
] |
2505.22349 | Anjie Xu | Anjie Xu, Ruiqing Ding, Leye Wang | ChatPD: An LLM-driven Paper-Dataset Networking System | Accepted by KDD Applied Data Science Track 2025 | null | 10.1145/3711896.3737202 | null | cs.DB cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scientific research heavily depends on suitable datasets for method validation, but existing academic platforms with dataset management like PapersWithCode suffer from inefficiencies in their manual workflow. To overcome this bottleneck, we present a system, called ChatPD, that utilizes Large Language Models (LLMs) to automate dataset information extraction from academic papers and construct a structured paper-dataset network. Our system consists of three key modules: \textit{paper collection}, \textit{dataset information extraction}, and \textit{dataset entity resolution} to construct paper-dataset networks. Specifically, we propose a \textit{Graph Completion and Inference} strategy to map dataset descriptions to their corresponding entities. Through extensive experiments, we demonstrate that ChatPD not only outperforms the existing platform PapersWithCode in dataset usage extraction but also achieves about 90\% precision and recall in entity resolution tasks. Moreover, we have deployed ChatPD to continuously extract which datasets are used in papers, and provide a dataset discovery service, such as task-specific dataset queries and similar dataset recommendations. We open source ChatPD and the current paper-dataset network on this [GitHub repository]{https://github.com/ChatPD-web/ChatPD}. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:31:08 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Xu",
"Anjie",
""
],
[
"Ding",
"Ruiqing",
""
],
[
"Wang",
"Leye",
""
]
] |
2505.22350 | Michele Stecconi | Michele Stecconi and Anna Paola Todino | New chaos decomposition of Gaussian nodal volumes | 37 pages | null | null | null | math.PR math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the random variable defined by the volume of the zero set of a smooth Gaussian field, on a general Riemannian manifold possibly with boundary, a fundamental object in probability and geometry. We prove a new explicit formula for its Wiener-It\^o chaos decomposition that is notably simpler than existing alternatives and which holds in greater generality, without requiring the field to be compatible with the geometry of the manifold. A key advantage of our formulation is a significant reduction in the complexity of computing the variance of the nodal volume. Unlike the standard Hermite expansion, which requires evaluating the expectation of products of $2+2n$ Hermite polynomials, our approach reduces this task--in any dimension $n$--to computing the expectation of a product of just four Hermite polynomials. As a consequence, we establish a new exact formula for the variance, together with lower and upper bounds. Our approach introduces two parameters associated to any Gaussian field: the frequency and the eccentricity. We use them to establish a quantitative version of Berry's cancellation phenomenon for Riemannian random waves on general manifolds. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:32:19 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Stecconi",
"Michele",
""
],
[
"Todino",
"Anna Paola",
""
]
] |
2505.22351 | Daniel Paulusma | Konrad K. Dabrowski, Tala Eagling-Vose, Matthew Johnson, Giacomo Paesani, Dani\"el Paulusma | Finding $d$-Cuts in Probe $H$-Free Graphs | null | null | null | null | cs.DS cs.CC cs.DM math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For an integer $d\geq 1$, the $d$-Cut problem is that of deciding whether a graph has an edge cut in which each vertex is adjacent to at most $d$ vertices on the opposite side of the cut. The $1$-Cut problem is the well-known Matching Cut problem. The $d$-Cut problem has been extensively studied for $H$-free graphs. We extend these results to the probe graph model, where we do not know all the edges of the input graph. For a graph $H$, a partitioned probe $H$-free graph $(G,P,N)$ consists of a graph $G=(V,E)$, together with a set $P\subseteq V$ of probes and an independent set $N=V\setminus P$ of non-probes such that we can change $G$ into an $H$-free graph by adding zero or more edges between vertices in $N$. For every graph $H$ and every integer $d\geq 1$, we completely determine the complexity of $d$-Cut on partitioned probe $H$-free graphs. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:32:59 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Dabrowski",
"Konrad K.",
""
],
[
"Eagling-Vose",
"Tala",
""
],
[
"Johnson",
"Matthew",
""
],
[
"Paesani",
"Giacomo",
""
],
[
"Paulusma",
"Daniël",
""
]
] |
2505.22352 | Poulomee Ghosh | Poulomee Ghosh and Shubhendu Bhasin | State and Input Constrained Adaptive Tracking Control of Uncertain Euler-Lagrange Systems with Robustness and Feasibility Analysis | arXiv admin note: text overlap with arXiv:2212.08495 | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper proposes an adaptive tracking controller for uncertain Euler-Lagrange (E-L) systems with user-defined state and input constraints in presence of bounded external disturbances. A barrier Lyapunov function (BLF) is employed for state constraint satisfaction, integrated with a saturated controller that ensures the control input remains within pre-specified bounds. To the best of the authors' knowledge, this is the first result on tracking control of state and input-constrained uncertain E-L systems that provides verifiable conditions for the existence of a feasible control policy. The efficacy of the proposed controller in terms of constraint satisfaction and tracking performance is demonstrated through simulation on a robotic manipulator system. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:33:14 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ghosh",
"Poulomee",
""
],
[
"Bhasin",
"Shubhendu",
""
]
] |
2505.22354 | Judith Sieker | Judith Sieker, Clara Lachenmaier, Sina Zarrie{\ss} | LLMs Struggle to Reject False Presuppositions when Misinformation Stakes are High | 8 pages (including References). Accepted at CogSci 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | This paper examines how LLMs handle false presuppositions and whether certain linguistic factors influence their responses to falsely presupposed content. Presuppositions subtly introduce information as given, making them highly effective at embedding disputable or false information. This raises concerns about whether LLMs, like humans, may fail to detect and correct misleading assumptions introduced as false presuppositions, even when the stakes of misinformation are high. Using a systematic approach based on linguistic presupposition analysis, we investigate the conditions under which LLMs are more or less sensitive to adopt or reject false presuppositions. Focusing on political contexts, we examine how factors like linguistic construction, political party, and scenario probability impact the recognition of false presuppositions. We conduct experiments with a newly created dataset and examine three LLMs: OpenAI's GPT-4-o, Meta's LLama-3-8B, and MistralAI's Mistral-7B-v03. Our results show that the models struggle to recognize false presuppositions, with performance varying by condition. This study highlights that linguistic presupposition analysis is a valuable tool for uncovering the reinforcement of political misinformation in LLM responses. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:35:07 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Sieker",
"Judith",
""
],
[
"Lachenmaier",
"Clara",
""
],
[
"Zarrieß",
"Sina",
""
]
] |
2505.22355 | Yongkang Liu | Yongkang Liu and Xingle Xu and Ercong Nie and Zijing Wang and Shi Feng and Daling Wang and Qian Li and Hinrich Sch\"utze | Look Within or Look Beyond? A Theoretical Comparison Between Parameter-Efficient and Full Fine-Tuning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Parameter-Efficient Fine-Tuning (PEFT) methods achieve performance comparable to Full Fine-Tuning (FFT) while requiring significantly fewer computing resources, making it the go-to choice for researchers. We find that although PEFT can achieve competitive results on some benchmarks, its performance falls short of FFT in complex tasks, such as reasoning and instruction-based fine-tuning. In this paper, we compare the characteristics of PEFT and FFT in terms of representational capacity and robustness based on optimization theory. We theoretically demonstrate that PEFT is a strict subset of FFT. By providing theoretical upper bounds for PEFT, we show that the limited parameter space constrains the model's representational ability, making it more susceptible to perturbations. Experiments on 15 datasets encompassing classification, generation, reasoning, instruction fine-tuning tasks and 11 adversarial test sets validate our theories. We hope that these results spark further research beyond the realms of well established PEFT. The source code is in the anonymous Github repository\footnote{https://github.com/misonsky/PEFTEval}. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:35:12 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Liu",
"Yongkang",
""
],
[
"Xu",
"Xingle",
""
],
[
"Nie",
"Ercong",
""
],
[
"Wang",
"Zijing",
""
],
[
"Feng",
"Shi",
""
],
[
"Wang",
"Daling",
""
],
[
"Li",
"Qian",
""
],
[
"Schütze",
"Hinrich",
""
]
] |
2505.22356 | Mohammad Yaghini | Ang\'eline Pouget, Mohammad Yaghini, Stephan Rabanser, Nicolas Papernot | Suitability Filter: A Statistical Framework for Classifier Evaluation in Real-World Deployment Settings | Accepted to ICML 2025 | null | null | null | cs.LG cs.AI cs.CY stat.ML | http://creativecommons.org/licenses/by/4.0/ | Deploying machine learning models in safety-critical domains poses a key challenge: ensuring reliable model performance on downstream user data without access to ground truth labels for direct validation. We propose the suitability filter, a novel framework designed to detect performance deterioration by utilizing suitability signals -- model output features that are sensitive to covariate shifts and indicative of potential prediction errors. The suitability filter evaluates whether classifier accuracy on unlabeled user data shows significant degradation compared to the accuracy measured on the labeled test dataset. Specifically, it ensures that this degradation does not exceed a pre-specified margin, which represents the maximum acceptable drop in accuracy. To achieve reliable performance evaluation, we aggregate suitability signals for both test and user data and compare these empirical distributions using statistical hypothesis testing, thus providing insights into decision uncertainty. Our modular method adapts to various models and domains. Empirical evaluations across different classification tasks demonstrate that the suitability filter reliably detects performance deviations due to covariate shift. This enables proactive mitigation of potential failures in high-stakes applications. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:37:04 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Pouget",
"Angéline",
""
],
[
"Yaghini",
"Mohammad",
""
],
[
"Rabanser",
"Stephan",
""
],
[
"Papernot",
"Nicolas",
""
]
] |
2505.22357 | David Luo | David C. Luo and Shaun Stevens | On the Local Converse Theorem for Depth $\frac{1}{N}$ Supercuspidal Representations of $\text{GL}(2N, F)$ | null | null | null | null | math.RT math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we use type theory to construct a family of depth $\frac{1}{N}$ minimax supercuspidal representations of $\text{GL}(2N, F)$ which we call middle supercuspidal representations. These supercuspidals may be viewed as a natural generalization of simple supercuspidal representations, i.e. those supercuspidals of minimal positive depth. Via explicit computations of twisted gamma factors, we show that middle supercuspidal representations may be uniquely determined through twisting by quasi-characters of $F^{\times}$ and simple supercuspidal representations of $\text{GL}(N, F)$. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:37:56 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Luo",
"David C.",
""
],
[
"Stevens",
"Shaun",
""
]
] |
2505.22358 | Wanrou Du | Zhiyi Wan, Wanrou Du, Liang Li, Miao Pan, Xiaoqi Qin | Budget-Adaptive Adapter Tuning in Orthogonal Subspaces for Continual Learning in LLMs | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) often suffer from catastrophic forgetting in continual learning (CL) scenarios, where performance on previously learned tasks degrades severely while training on sequentially arriving tasks. Although pioneering CL approaches using orthogonal subspaces can mitigate task interference, they typically employ fixed budget allocation, neglecting the varying complexity across tasks and layers. Besides, recent budget-adaptive tuning methods for LLMs often adopt multi-stage paradigms that decouple optimization and budget allocation. Such decoupling results in potential misalignment, which hinders those approaches' practical application in CL scenarios. To address these limitations, we propose OA-Adapter, a novel parameter-efficient approach for continual learning in LLMs that unifies dynamic budget adaptation with orthogonal subspace learning in a single end-to-end training stage. Specifically, OA-Adapter introduces a dynamic bottleneck dimension adaptation mechanism that simultaneously allocates an efficient parameter budget and optimizes task objectives without misalignment. To effectively preserve previously acquired knowledge while coordinating with the dynamic budget allocation, orthogonal constraints are applied specifically between the parameter subspace of the current task and the dynamically allocated parameter subspaces of historical tasks. Experimental results on continual learning benchmarks demonstrate that OA-Adapter outperforms state-of-the-art methods in both accuracy and parameter efficiency, achieving higher average accuracy while using 58.5% fewer parameters on the standard CL benchmark. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:38:21 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Wan",
"Zhiyi",
""
],
[
"Du",
"Wanrou",
""
],
[
"Li",
"Liang",
""
],
[
"Pan",
"Miao",
""
],
[
"Qin",
"Xiaoqi",
""
]
] |
2505.22359 | Matan Schliserman | Matan Schliserman and Tomer Koren | Multiclass Loss Geometry Matters for Generalization of Gradient Descent in Separable Classification | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the generalization performance of unregularized gradient methods for separable linear classification. While previous work mostly deal with the binary case, we focus on the multiclass setting with $k$ classes and establish novel population risk bounds for Gradient Descent for loss functions that decay to zero. In this setting, we show risk bounds that reveal that convergence rates are crucially influenced by the geometry of the loss template, as formalized by Wang and Scott (2024), rather than of the loss function itself. Particularly, we establish risk upper bounds that holds for any decay rate of the loss whose template is smooth with respect to the $p$-norm. In the case of exponentially decaying losses, our results indicates a contrast between the $p=\infty$ case, where the risk exhibits a logarithmic dependence on $k$, and $p=2$ where the risk scales linearly with $k$. To establish this separation formally, we also prove a lower bound in the latter scenario, demonstrating that the polynomial dependence on $k$ is unavoidable. Central to our analysis is a novel bound on the Rademacher complexity of low-noise vector-valued linear predictors with a loss template smooth w.r.t.~general $p$-norms. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:39:14 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Schliserman",
"Matan",
""
],
[
"Koren",
"Tomer",
""
]
] |
2505.22360 | Kewen Chen | Kewen Chen, Xiaobin Hu, Wenqi Ren | Identity-Preserving Text-to-Image Generation via Dual-Level Feature Decoupling and Expert-Guided Fusion | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large-scale text-to-image generation models have led to a surge in subject-driven text-to-image generation, which aims to produce customized images that align with textual descriptions while preserving the identity of specific subjects. Despite significant progress, current methods struggle to disentangle identity-relevant information from identity-irrelevant details in the input images, resulting in overfitting or failure to maintain subject identity. In this work, we propose a novel framework that improves the separation of identity-related and identity-unrelated features and introduces an innovative feature fusion mechanism to improve the quality and text alignment of generated images. Our framework consists of two key components: an Implicit-Explicit foreground-background Decoupling Module (IEDM) and a Feature Fusion Module (FFM) based on a Mixture of Experts (MoE). IEDM combines learnable adapters for implicit decoupling at the feature level with inpainting techniques for explicit foreground-background separation at the image level. FFM dynamically integrates identity-irrelevant features with identity-related features, enabling refined feature representations even in cases of incomplete decoupling. In addition, we introduce three complementary loss functions to guide the decoupling process. Extensive experiments demonstrate the effectiveness of our proposed method in enhancing image generation quality, improving flexibility in scene adaptation, and increasing the diversity of generated outputs across various textual descriptions. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:40:46 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Chen",
"Kewen",
""
],
[
"Hu",
"Xiaobin",
""
],
[
"Ren",
"Wenqi",
""
]
] |
2505.22361 | Yining Wang | Xiangyu Chang, Xi Chen, Yining Wang, Zhiyi Zeng | Continuum-armed Bandit Optimization with Batch Pairwise Comparison Oracles | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | This paper studies a bandit optimization problem where the goal is to maximize a function $f(x)$ over $T$ periods for some unknown strongly concave function $f$. We consider a new pairwise comparison oracle, where the decision-maker chooses a pair of actions $(x, x')$ for a consecutive number of periods and then obtains an estimate of $f(x)-f(x')$. We show that such a pairwise comparison oracle finds important applications to joint pricing and inventory replenishment problems and network revenue management. The challenge in this bandit optimization is twofold. First, the decision-maker not only needs to determine a pair of actions $(x, x')$ but also a stopping time $n$ (i.e., the number of queries based on $(x, x')$). Second, motivated by our inventory application, the estimate of the difference $f(x)-f(x')$ is biased, which is different from existing oracles in stochastic optimization literature. To address these challenges, we first introduce a discretization technique and local polynomial approximation to relate this problem to linear bandits. Then we developed a tournament successive elimination technique to localize the discretized cell and run an interactive batched version of LinUCB algorithm on cells. We establish regret bounds that are optimal up to poly-logarithmic factors. Furthermore, we apply our proposed algorithm and analytical framework to the two operations management problems and obtain results that improve state-of-the-art results in the existing literature. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:41:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Chang",
"Xiangyu",
""
],
[
"Chen",
"Xi",
""
],
[
"Wang",
"Yining",
""
],
[
"Zeng",
"Zhiyi",
""
]
] |
2505.22362 | Aihu Zhang | Aihu Zhang, Jiaxing Xu, Mengcheng Lan, Shili Xiang, Yiping Ke | Directed Homophily-Aware Graph Neural Network | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Graph Neural Networks (GNNs) have achieved significant success in various learning tasks on graph-structured data. Nevertheless, most GNNs struggle to generalize to heterophilic neighborhoods. Additionally, many GNNs ignore the directional nature of real-world graphs, resulting in suboptimal performance on directed graphs with asymmetric structures. In this work, we propose Directed Homophily-aware Graph Neural Network (DHGNN), a novel framework that addresses these limitations by incorporating homophily-aware and direction-sensitive components. DHGNN employs a resettable gating mechanism to adaptively modulate message contributions based on homophily levels and informativeness, and a structure-aware noise-tolerant fusion module to effectively integrate node representations from the original and reverse directions. Extensive experiments on both homophilic and heterophilic directed graph datasets demonstrate that DHGNN outperforms state-of-the-art methods in node classification and link prediction. In particular, DHGNN improves over the best baseline by up to 15.07% in link prediction. Our analysis further shows that the gating mechanism captures directional homophily gaps and fluctuating homophily across layers, providing deeper insights into message-passing behavior on complex graph structures. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:41:04 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zhang",
"Aihu",
""
],
[
"Xu",
"Jiaxing",
""
],
[
"Lan",
"Mengcheng",
""
],
[
"Xiang",
"Shili",
""
],
[
"Ke",
"Yiping",
""
]
] |
2505.22363 | Amir Shahhosseini | Amir Shahhosseini, Thomas Chaffey and Rodolphe Sepulchre | Operator-Splitting Methods for Neuromorphic Circuit Simulation | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | A novel splitting algorithm is proposed for the numerical simulation of neuromorphic circuits. The algorithm is grounded in the operator-theoretic concept of monotonicity, which bears both physical and algorithmic significance. The splitting exploits this correspondence to translate the circuit architecture into the algorithmic architecture. The paper illustrates the many advantages of the proposed operator-theoretic framework over conventional numerical integration for the simulation of multiscale hierarchical events that characterize neuromorphic behaviors. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:42:26 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Shahhosseini",
"Amir",
""
],
[
"Chaffey",
"Thomas",
""
],
[
"Sepulchre",
"Rodolphe",
""
]
] |
2505.22364 | Gabriele Visentin | Gabriele Visentin, Patrick Cheridito | Computing Optimal Transport Maps and Wasserstein Barycenters Using Conditional Normalizing Flows | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a novel method for efficiently computing optimal transport maps and Wasserstein barycenters in high-dimensional spaces. Our approach uses conditional normalizing flows to approximate the input distributions as invertible pushforward transformations from a common latent space. This makes it possible to directly solve the primal problem using gradient-based minimization of the transport cost, unlike previous methods that rely on dual formulations and complex adversarial optimization. We show how this approach can be extended to compute Wasserstein barycenters by solving a conditional variance minimization problem. A key advantage of our conditional architecture is that it enables the computation of barycenters for hundreds of input distributions, which was computationally infeasible with previous methods. Our numerical experiments illustrate that our approach yields accurate results across various high-dimensional tasks and compares favorably with previous state-of-the-art methods. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:46:07 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Visentin",
"Gabriele",
""
],
[
"Cheridito",
"Patrick",
""
]
] |
2505.22365 | Lorenzo Lamberti | Lorenzo Lamberti and Antoine Lemenant | Quantitative regularity properties for the optimal design problem | null | null | null | null | math.OC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper we slightly improve the regularity theory for the so called optimal design problem. We first establish the uniform rectifiability of the boundary of the optimal set, for a larger class of minimizers, in any dimension. As an application, we improve the bound obtained by Larsen in dimension~2 about the mutual distance between two connected components. Finally we also prove that the full regularity in dimension 2 holds true provided that the ratio between the two constants in front of the Dirichlet energy is not larger than 4, which partially answers to a question raised by Larsen. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:48:55 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Lamberti",
"Lorenzo",
""
],
[
"Lemenant",
"Antoine",
""
]
] |
2505.22366 | Lukas Liedtke | Lukas Liedtke, Per Gunnar Kjeldsberg, Frank Alexander Kraemer, Magnus Jahre | EStacker: Explaining Battery-Less IoT System Performance with Energy Stacks | null | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | The number of Internet of Things (IoT) devices is increasing exponentially, and it is environmentally and economically unsustainable to power all these devices with batteries. The key alternative is energy harvesting, but battery-less IoT systems require extensive evaluation to demonstrate that they are sufficiently performant across the full range of expected operating conditions. IoT developers thus need an evaluation platform that (i) ensures that each evaluated application and configuration is exposed to exactly the same energy environment and events, and (ii) provides a detailed account of what the application spends the harvested energy on. We therefore developed the EStacker evaluation platform which (i) provides fair and repeatable evaluation, and (ii) generates energy stacks. Energy stacks break down the total energy consumption of an application across hardware components and application activities, thereby explaining what the application specifically uses energy on. We augment EStacker with the ST-SP optimization which, in our experiments, reduces evaluation time by 6.3x on average while retaining the temporal behavior of the battery-less IoT system (average throughput error of 7.7%) by proportionally scaling time and power. We demonstrate the utility of EStacker through two case studies. In the first case study, we use energy stack profiles to identify a performance problem that, once addressed, improves performance by 3.3x. The second case study focuses on ST-SP, and we use it to explore the design space required to dimension the harvester and energy storage sizes of a smart parking application in roughly one week (7.7 days). Without ST-SP, sweeping this design space would have taken well over one month (41.7 days). | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:49:23 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Liedtke",
"Lukas",
""
],
[
"Kjeldsberg",
"Per Gunnar",
""
],
[
"Kraemer",
"Frank Alexander",
""
],
[
"Jahre",
"Magnus",
""
]
] |
2505.22367 | Raghunath Sahoo | Aswathy Menon K R, Suraj Prasad, Neelkamal Mallick, Raghunath Sahoo, and Gergely G\'abor Barnaf\"oldi | Probing nuclear structure in relativistic p-O and O-O collisions at the LHC through the measurement of anisotropic flow coefficients | 11 pages and 5 captioned figures. Submitted for publication | null | null | null | hep-ph hep-ex hep-th nucl-ex nucl-th | http://creativecommons.org/licenses/by-sa/4.0/ | RHIC and LHC plan to inject $^{16}\rm O$ nuclei with a focus to investigate collectivity and the origin of quark-gluon plasma signatures in small collision systems. $^{16}\rm O$ nuclei is known to possess clusters of $\alpha$-particles ($^{4}\rm He$) inside the nucleus. In this paper, we study the anisotropic flow coefficients such as elliptic flow ($v_2$) and triangular flow ($v_3$), which are sensitive to the nuclear geometry of colliding nuclei, for p-O and O-O collisions at $\sqrt{s_{\rm NN}}=9.61$ TeV and 7 TeV respectively. The study is performed employing a hybrid model encompassing IPGlasma + MUSIC + iSS + UrQMD. The results of the clustered nuclear geometry are compared with those of the Woods-Saxon nuclear profile. Both initial and final state anisotropies are explored. This study is thus one of its first kind, where the study of anisotropic flow coefficients for p-O and O-O collisions is presented using a hybrid hydrodynamics model. We find a small effect of $\alpha$-clustering in p-O, while a significant one for O-O collisions. It is also observed that the magnitude of the effect correlates with the size of the $^{4}$He. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:55:02 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"R",
"Aswathy Menon K",
""
],
[
"Prasad",
"Suraj",
""
],
[
"Mallick",
"Neelkamal",
""
],
[
"Sahoo",
"Raghunath",
""
],
[
"Barnaföldi",
"Gergely Gábor",
""
]
] |
2505.22368 | Yujun Cheng | Enfang Cui, Yujun Cheng, Rui She, Dan Liu, Zhiyuan Liang, Minxin Guo, Tianzheng Li, Qian Wei, Wenjuan Xing, Zhijie Zhong | AgentDNS: A Root Domain Naming System for LLM Agents | 7 pages, 6 figures | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | The rapid evolution of Large Language Model (LLM) agents has highlighted critical challenges in cross-vendor service discovery, interoperability, and communication. Existing protocols like model context protocol and agent-to-agent protocol have made significant strides in standardizing interoperability between agents and tools, as well as communication among multi-agents. However, there remains a lack of standardized protocols and solutions for service discovery across different agent and tool vendors. In this paper, we propose AgentDNS, a root domain naming and service discovery system designed to enable LLM agents to autonomously discover, resolve, and securely invoke third-party agent and tool services across organizational and technological boundaries. Inspired by the principles of the traditional DNS, AgentDNS introduces a structured mechanism for service registration, semantic service discovery, secure invocation, and unified billing. We detail the architecture, core functionalities, and use cases of AgentDNS, demonstrating its potential to streamline multi-agent collaboration in real-world scenarios. The source code will be published on https://github.com/agentdns. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:56:22 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Cui",
"Enfang",
""
],
[
"Cheng",
"Yujun",
""
],
[
"She",
"Rui",
""
],
[
"Liu",
"Dan",
""
],
[
"Liang",
"Zhiyuan",
""
],
[
"Guo",
"Minxin",
""
],
[
"Li",
"Tianzheng",
""
],
[
"Wei",
"Qian",
""
],
[
"Xing",
"Wenjuan",
""
],
[
"Zhong",
"Zhijie",
""
]
] |
2505.22369 | Xin Zhang | Jia-Le Ling, Guo-Hong Du, Tian-Nuo Li, Jing-Fei Zhang, Shao-Jiang Wang, Xin Zhang | Model-independent cosmological inference after the DESI DR2 data with improved inverse distance ladder | 13 pages, 4 figures | null | null | null | astro-ph.CO gr-qc hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, the baryon acoustic oscillations (BAO) measurements from the DESI survey have suggested hints of dynamical dark energy, challenging the standard $\Lambda $CDM model. In this work, we adopt an improved inverse distance ladder approach based on the latest cosmological data to provide a model-independent perspective, employing a global parametrization based on cosmic age (PAge). Our analysis incorporates DESI DR2 BAO measurements, cosmic chronometer (CC) data, and type Ia supernovae (SNe) observations from either the DESY5 or PantheonPlus datasets. For the DESY5+DESI DR2+CC datasets, we obtain $H_0 = 67.91 \pm 2.33~\mathrm{km~s^{-1}~Mpc^{-1}}$. This value is consistent with the Planck 2018 result, while shows $2.0 \sigma$ tension with the SH0ES measurement. Furthermore, by mapping specific cosmological models into PAge approximation parameter space $(p_{\mathrm{age}}, \eta)$, our model-independent analysis reveals a notable deviation from the $\Lambda \mathrm{CDM}$ model, as indicated by the DESY5 and DESI DR2 datasets. Finally, DESY5+DESI DR2+CC datasets provide nearly decisive evidence favoring the PAge model over the standard $\Lambda \mathrm{CDM}$ model. These findings highlight the need for further investigation into the expansion history to better understand the deviations from the $\Lambda \mathrm{CDM}$ model. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:56:48 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ling",
"Jia-Le",
""
],
[
"Du",
"Guo-Hong",
""
],
[
"Li",
"Tian-Nuo",
""
],
[
"Zhang",
"Jing-Fei",
""
],
[
"Wang",
"Shao-Jiang",
""
],
[
"Zhang",
"Xin",
""
]
] |
2505.22372 | Miha E. Habi\v{c} | Miha E. Habi\v{c} and Charles Weng and Cathy Zhang | More nonamalgamable forcing extensions | 17 pages | null | null | null | math.LO | http://creativecommons.org/licenses/by-sa/4.0/ | We extend the results of arXiv:1808.01509 on nonamalgamable forcing extensions to families of posets with wide projections. We also use a different coding method to obtain nonamalgamable extensions by filter-based Mathias forcing. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:58:38 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Habič",
"Miha E.",
""
],
[
"Weng",
"Charles",
""
],
[
"Zhang",
"Cathy",
""
]
] |
2505.22373 | Alessandro Torcini Prof | Denis S. Goldobin, Maria V. Ageeva, Matteo di Volo, Ferdinand Tixidre, Alessandro Torcini | Synaptic shot-noise triggers fast and slow global oscillations in balanced neural networks | 28 pages, 14 figures, submitted to Physical Review E | null | null | null | cond-mat.dis-nn q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | Neural dynamics is determined by the transmission of discrete synaptic pulses (synaptic shot-noise) among neurons. However, the neural responses are usually obtained within the diffusion approximation modeling synaptic inputs as continuous Gaussian noise. Here, we present a rigorous mean-field theory that encompasses synaptic shot-noise for sparse balanced inhibitory neural networks driven by an excitatory drive. Our theory predicts new dynamical regimes, in agreement with numerical simulations, which are not captured by the classical diffusion approximation. Notably, these regimes feature self-sustained global oscillations emerging at low connectivity (in-degree) via either continuous or hysteretic transitions and characterized by irregular neural activity, as expected for balanced dynamics. For sufficiently weak (strong) excitatory drive (inhibitory feedback) the transition line displays a peculiar re-entrant shape revealing the existence of global oscillations at low and high in-degrees, separated by an asynchronous regime at intermediate levels of connectivity. The mechanisms leading to the emergence of these global oscillations are distinct: drift-driven at high connectivity and cluster activation at low connectivity. The frequency of these two kinds of global oscillations can be varied from slow (around 1 Hz) to fast (around 100 Hz), without altering their microscopic and macroscopic features, by adjusting the excitatory drive and the synaptic inhibition strength in a prescribed way. Furthermore, the cluster-activated oscillations at low in-degrees could correspond to the gamma rhythms reported in mammalian cortex and hippocampus and attributed to ensembles of inhibitory neurons sharing few synaptic connections [G. Buzsaki and X.-J. Wang, Annual Review of Neuroscience 35, 203 (2012)]. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 13:59:01 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Goldobin",
"Denis S.",
""
],
[
"Ageeva",
"Maria V.",
""
],
[
"di Volo",
"Matteo",
""
],
[
"Tixidre",
"Ferdinand",
""
],
[
"Torcini",
"Alessandro",
""
]
] |
2505.22374 | Yurii Nikonorov Gennadyevich | E.V. Nikitenko, Yu.G. Nikonorov | On face angles of tetrahedra with a given base | 40 pages, 39 figures. Comments are welcome! | null | null | null | math.MG math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let us consider the set $\Omega (\triangle ABC)$ of all tetrahedra $ABCD$ with a given non-degenerate base $ABC$ in $\mathbb{E}^3$ and $D$ lying outside the plane $ABC$. Let us denote by $\Sigma(\triangle ABC)$ the set $\left\{\Bigl(\cos \overline{\alpha},\cos \overline{\beta},\cos \overline{\gamma} \Bigr)\in \mathbb{R}^3\,|\, ABCD \in \Omega (\triangle ABC)\right\}$, where $\overline{\alpha}=\angle BDC$, $\overline{\beta}=\angle ADC$, and $\overline{\gamma}=\angle ADB$. The paper is devoted to the problem of determining of the closure of $\Sigma(\triangle ABC)$ in $\mathbb{R}^3$ and its boundary. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:02:46 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Nikitenko",
"E. V.",
""
],
[
"Nikonorov",
"Yu. G.",
""
]
] |
2505.22376 | Ba\c{s}ak K\"u\c{c}\"uk | Ba\c{s}ak K\"u\c{c}\"uk | Functoriality of the Klein-Williams Invariant and Universality Theory | 33 pages | null | null | null | math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Both the Klein-Williams invariant $\ell_G(f)$ from \cite{KW2} and the generalized equivariant Lefschetz invariant $\lambda_G(f)$ from \cite{weber07} serve as complete obstructions to the fixed point problem in the equivariant setting. The latter is functorial in the sense of Definition \ref{functorial}. The first part of this paper aims to demonstrate that $\ell_G(f)$ is also functorial. The second part summarizes the ``universality" theory of such functorial invariants, developed in \cites{lueck1999, Weber06}, and explicitly computes the group in which the universal invariant lies, under a certain hypothesis. The final part explores the relationship between $\ell_G(f)$ and $\lambda_G(f)$, and presents examples to compare $\ell_G(f)$, $\lambda_G(f)$, and the universal invariant. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:03:22 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Küçük",
"Başak",
""
]
] |
2505.22377 | Minghua Chen Professor | Na Xue, Minghua Chen | Multiprecision computing for multistage fractional physics-informed neural networks | 17 pages | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fractional physics-informed neural networks (fPINNs) have been successfully introduced in [Pang, Lu and Karniadakis, SIAM J. Sci. Comput. 41 (2019) A2603-A2626], which observe relative errors of $10^{-3} \, \sim \, 10^{-4}$ for the subdiffusion equations. However their high-precision (multiprecision) numerical solution remains challenging, due to the limited regularity of the subdiffusion model caused by the nonlocal operator. To fill in the gap, we present the multistage fPINNs based on traditional multistage PINNs [Wang and Lai, J. Comput. Phys. 504 (2024) 112865]. Numerical experiments show that the relative errors improve to $10^{-7} \, \sim \, 10^{-8}$ for the subdiffusion equations on uniform or nouniform meshes. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:06:11 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Xue",
"Na",
""
],
[
"Chen",
"Minghua",
""
]
] |
2505.22378 | Michael Hertneck | Michael Hertneck and David Meister and Frank Allg\"ower | Current trends and future directions in event-based control | Preprint submitted to the European Journal of Control | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The defining characteristic of event-based control is that feedback loops are only closed when indicated by a triggering condition that takes recent information about the system into account. This stands in contrast to periodic control where the feedback loop is closed periodically. Benefits of event-based control arise when sampling comes at a cost, which occurs, e.g., for Networked Control Systems or in other setups with resource constraints. A rapidly growing number of publications deals with event-based control. Nevertheless, some fundamental questions about event-based control are still unsolved. In this article, we provide an overview of current research trends in event-based control. We focus on results that aim for a better understanding of effects that occur in feedback loops with event-based control. Based on this summary, we identify important open directions for future research. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:06:32 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Hertneck",
"Michael",
""
],
[
"Meister",
"David",
""
],
[
"Allgöwer",
"Frank",
""
]
] |
2505.22379 | Souradip Chattopadhyay | Souradip Chattopadhyay, Zihao Yu, Y. Sungtaek Ju, Hangjie Ji | Dynamics of thin film flows on a vertical fibre with vapor absorption | 32 pages, 13 figures | null | null | null | physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Water vapor capture through free surface flows plays a crucial role in various industrial applications, such as liquid desiccant air conditioning systems, water harvesting, and dewatering. This paper studies the dynamics of a silicone liquid sorbent (also known as water-absorbing silicone oil) flowing down a vertical cylindrical fibre while absorbing water vapor. We propose a one-sided thin-film-type model for these dynamics, where the governing equations form a coupled system of nonlinear fourth-order partial differential equations for the liquid film thickness and oil concentration. The model incorporates gravity, surface tension, Marangoni effects induced by concentration gradients, and non-mass-conserving effects due to absorption flux. Interfacial instabilities, driven by the competition between mass-conserving and non-mass-conserving effects, are investigated via stability analysis. We numerically show that water absorption can lead to the formation of irregular wavy patterns and trigger droplet coalescence downstream. Systematic simulations further identify parameter ranges for the Marangoni number and absorption parameter that lead to the onset of droplet coalescence dynamics and regime transitions. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:07:56 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Chattopadhyay",
"Souradip",
""
],
[
"Yu",
"Zihao",
""
],
[
"Ju",
"Y. Sungtaek",
""
],
[
"Ji",
"Hangjie",
""
]
] |
2505.22380 | Bernd Siebert | Michel van Garrel, Helge Ruddat, Bernd Siebert | Intrinsic enumerative mirror symmetry: Takahashi's log mirror symmetry for $(\mathbb{P}^2,E)$ revisited | 31 pages, 6 figures | null | null | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | Let $E$ be a smooth cubic in the projective plane $\mathbb{P}^2$. Nobuyoshi Takahashi formulated a conjecture that expresses counts of rational curves of varying degree in $\mathbb{P}^2\setminus E$ as the Taylor coefficients of a particular period integral of a pencil of affine plane cubics after reparametrizing the pencil using the exponential of a second period integral.
The intrinsic mirror construction introduced by Mark Gross and the third author associates to a degeneration of $(\mathbb{P}^2, E)$ a canonical wall structure from which one constructs a family of projective plane cubics that is birational to Takahashi's pencil in its reparametrized form. By computing the period integral of the positive real locus explicitly, we find that it equals the logarithm of the product of all asymptotic wall functions. The coefficients of these asymptotic wall functions are logarithmic Gromov-Witten counts of the central fiber of the degeneration that agree with the algebraic curve counts in $(\mathbb{P}^2,E)$ in question. We conclude that Takahashi's conjecture is a natural consequence of intrinsic mirror symmetry. Our method generalizes to give similar results for log Calabi-Yau varieties of arbitrary dimension. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:08:32 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"van Garrel",
"Michel",
""
],
[
"Ruddat",
"Helge",
""
],
[
"Siebert",
"Bernd",
""
]
] |
2505.22381 | Lukas Kirchdorfer | Lukas Kirchdorfer, Konrad \"Ozdemir, Stjepan Kusenic, Han van der Aa, Heiner Stuckenschmidt | A Divide-and-Conquer Approach for Modeling Arrival Times in Business Process Simulation | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Business Process Simulation (BPS) is a critical tool for analyzing and improving organizational processes by estimating the impact of process changes. A key component of BPS is the case-arrival model, which determines the pattern of new case entries into a process. Although accurate case-arrival modeling is essential for reliable simulations, as it influences waiting and overall cycle times, existing approaches often rely on oversimplified static distributions of inter-arrival times. These approaches fail to capture the dynamic and temporal complexities inherent in organizational environments, leading to less accurate and reliable outcomes. To address this limitation, we propose Auto Time Kernel Density Estimation (AT-KDE), a divide-and-conquer approach that models arrival times of processes by incorporating global dynamics, day-of-week variations, and intraday distributional changes, ensuring both precision and scalability. Experiments conducted across 20 diverse processes demonstrate that AT-KDE is far more accurate and robust than existing approaches while maintaining sensible execution time efficiency. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:09:51 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Kirchdorfer",
"Lukas",
""
],
[
"Özdemir",
"Konrad",
""
],
[
"Kusenic",
"Stjepan",
""
],
[
"van der Aa",
"Han",
""
],
[
"Stuckenschmidt",
"Heiner",
""
]
] |
2505.22382 | Jean Kieffer | Noam D. Elkies and Jean Kieffer | Fast evaluation of Riemann theta functions in any dimension | null | null | null | null | math.NT cs.NA math.AG math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe an algorithm to numerically evaluate Riemann theta functions in any dimension in quasi-linear time in terms of the required precision, uniformly on reduced input. This algorithm is implemented in the FLINT number theory library and vastly outperforms existing software. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:09:57 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Elkies",
"Noam D.",
""
],
[
"Kieffer",
"Jean",
""
]
] |
2505.22383 | Mikaela Salom\'e Carrasco Hidalgo | M. Carrasco-H, E. Contreras, E. Fuenmayor and P. Le\'on | Hyperbolic polytrope | 15 pages, 26 figures | Annals of Physics (2025), Volume 473, Article 169909 | 10.1016/j.aop.2024.169909 | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we study self-gravitating objects that obey a polytropic equation of state in hyperbolic symmetry. Specifically, we describe in detail the steps to derive the Lane-Emden equation from the structure equations of the system. To integrate the equations numerically, we propose the Cosenza-Herrera-Esculpi-Witten anisotropy and study the cases $\gamma \ne 1$ and $\gamma = 1$ in the parameter space of the models. We find that the matter sector exhibits the usual and expected behavior for certain values in this parameter space: energy density (in absolute value) and radial pressure are decreasing functions and vanish at the surface, while the mass function is increasing toward the surface. We find that the anisotropy of the system is positive and decreasing, consistent with the behavior of the radial pressure, which reaches a local minimum at the surface (i.e., the pressure gradient is zero at the surface). We also study the compactness of the dense objects as a function of the polytropic index and obtain that it has an upper bound given by the maximum value it reaches for a certain $n$. Some extensions of the work and future proposals are discussed. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:10:03 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Carrasco-H",
"M.",
""
],
[
"Contreras",
"E.",
""
],
[
"Fuenmayor",
"E.",
""
],
[
"León",
"P.",
""
]
] |
2505.22384 | Harmender Gahlawat | Foivos Fioravantes and Harmender Gahlawat and Nikolaos Melissinos | Exact Algorithms and Lower Bounds for Forming Coalitions of Constrained Maximum Size | a preliminary version appeared in AAAI 2025 | null | null | null | cs.DS cs.AI | http://creativecommons.org/licenses/by/4.0/ | Imagine we want to split a group of agents into teams in the most \emph{efficient} way, considering that each agent has their own preferences about their teammates. This scenario is modeled by the extensively studied \textsc{Coalition Formation} problem. Here, we study a version of this problem where each team must additionally be of bounded size.
We conduct a systematic algorithmic study, providing several intractability results as well as multiple exact algorithms that scale well as the input grows (FPT), which could prove useful in practice.
Our main contribution is an algorithm that deals efficiently with tree-like structures (bounded \emph{treewidth}) for ``small'' teams. We complement this result by proving that our algorithm is asymptotically optimal. Particularly, there can be no algorithm that vastly outperforms the one we present, under reasonable theoretical assumptions, even when considering star-like structures (bounded \emph{vertex cover number}). | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:11:14 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Fioravantes",
"Foivos",
""
],
[
"Gahlawat",
"Harmender",
""
],
[
"Melissinos",
"Nikolaos",
""
]
] |
2505.22385 | Soumyajit Seth | Samaira Tibrewal and Soumyajit Seth | Does the Forced Van der Pol Oscillator Exhibit Irregular Behavior? | null | null | null | null | nlin.CD physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nonlinear oscillators are commonly encountered in a wide range of physical and engineering systems, exhibiting rich and complex dynamics. Among these, the Van der Pol oscillator is well known for its self-sustained limit cycle behavior. However, when subjected to external sinusoidal forcing, its dynamics can deviate significantly from this regular behavior. This study explores the emergence of complex dynamical regimes in the sinusoidally forced Van der Pol oscillator, focusing on identifying the transition from regular to aperiodic behavior. The primary research objective is to determine whether the system exhibits irregular dynamics such as quasi-periodicity or chaos and to identify the conditions under which these arise. We hypothesize that varying the amplitude and frequency ratio of the external forcing relative to the natural frequency of the oscillator can induce a spectrum of dynamical responses, including higher-order periodic and chaotic regimes. We perform numerical simulations using Python and examine the behavior of the system through time series analysis, phase portraits, and bifurcation diagrams to test this. The results demonstrate that as the forcing amplitude and frequency ratio are varied, the system undergoes transitions through periodic, quasi-periodic, and chaotic states. These findings highlight the complex nonlinear interactions in forced oscillatory systems and have significant implications for applications in biological rhythms, electronic circuits, and astrophysical phenomena. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:12:38 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Tibrewal",
"Samaira",
""
],
[
"Seth",
"Soumyajit",
""
]
] |
2505.22386 | Dinesha Vasanta Hegde | Dinesha V. Hegde (1 and 2), Tae K. Kim (2), Nikolai V. Pogorelov (1 and 2), Shaela I. Jones (3 and 4), Charles N. Arge (3) ((1) Department of Space Science, The University of Alabama in Huntsville, Huntsville, USA (2) Center for Space Plasma and Aeronomic Research, The University of Alabama in Huntsville, Huntsville, USA (3) NASA Goddard Space Flight Center, Greenbelt, USA (4) Catholic University of America, Washington, USA) | Ensemble Modeling of the Solar Wind Flow with Boundary Conditions Governed by Synchronic Photospheric Magnetograms. I. Multi-point Validation in the Inner Heliosphere | 36 Pages, 12 Figures, Accepted for publication in The Astrophysical Journal (ApJ) | null | null | null | astro-ph.SR physics.comp-ph physics.space-ph | http://creativecommons.org/licenses/by/4.0/ | The solar wind (SW) is a vital component of space weather, providing a background for solar transients such as coronal mass ejections, stream interaction regions, and energetic particles propagating toward Earth. Accurate prediction of space weather events requires a precise description and thorough understanding of physical processes occurring in the ambient SW plasma. Ensemble simulations of the three-dimensional SW flow are performed using an empirically-driven magnetohydrodynamic heliosphere model implemented in the Multi-Scale Fluid-Kinetic Simulation Suite (MS-FLUKSS). The effect of uncertainties in the photospheric boundary conditions on the simulation outcome is investigated. The results are in good overall agreement with the observations from the Parker Solar Probe, Solar Orbiter, Solar Terrestrial Relations Observatory, and OMNI data at Earth, specifically during 2020-2021. This makes it possible to shed more light on the properties of the SW propagating through the heliosphere and perspectives for improving space weather forecasts. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:13:09 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Hegde",
"Dinesha V.",
"",
"1 and 2"
],
[
"Kim",
"Tae K.",
"",
"Center for Space Plasma and Aeronomic Research, The University of Alabama in Huntsville, Huntsville, USA"
],
[
"Pogorelov",
"Nikolai V.",
"",
"1 and 2"
],
[
"Jones",
"Shaela I.",
"",
"3 and 4"
],
[
"Arge",
"Charles N.",
"",
"NASA Goddard Space Flight Center, Greenbelt, USA"
]
] |
2505.22387 | Jaehyun Choi | Jaehyun Choi, Gyojin Han, Dong-Jae Lee, Sunghyun Baek, Junmo Kim | DAM: Domain-Aware Module for Multi-Domain Dataset Condensation | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Dataset Condensation (DC) has emerged as a promising solution to mitigate the computational and storage burdens associated with training deep learning models. However, existing DC methods largely overlook the multi-domain nature of modern datasets, which are increasingly composed of heterogeneous images spanning multiple domains. In this paper, we extend DC and introduce Multi-Domain Dataset Condensation (MDDC), which aims to condense data that generalizes across both single-domain and multi-domain settings. To this end, we propose the Domain-Aware Module (DAM), a training-time module that embeds domain-related features into each synthetic image via learnable spatial masks. As explicit domain labels are mostly unavailable in real-world datasets, we employ frequency-based pseudo-domain labeling, which leverages low-frequency amplitude statistics. DAM is only active during the condensation process, thus preserving the same images per class (IPC) with prior methods. Experiments show that DAM consistently improves in-domain, out-of-domain, and cross-architecture performance over baseline dataset condensation methods. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:13:38 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Choi",
"Jaehyun",
""
],
[
"Han",
"Gyojin",
""
],
[
"Lee",
"Dong-Jae",
""
],
[
"Baek",
"Sunghyun",
""
],
[
"Kim",
"Junmo",
""
]
] |
2505.22388 | Jin Xi | Zhentao Shi, Jin Xi, Haitian Xie | A Synthetic Business Cycle Approach to Counterfactual Analysis with Nonstationary Macroeconomic Data | null | null | null | null | econ.EM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates the use of synthetic control methods for causal inference in macroeconomic settings when dealing with possibly nonstationary data. While the synthetic control approach has gained popularity for estimating counterfactual outcomes, we caution researchers against assuming a common nonstationary trend factor across units for macroeconomic outcomes, as doing so may result in misleading causal estimation-a pitfall we refer to as the spurious synthetic control problem. To address this issue, we propose a synthetic business cycle framework that explicitly separates trend and cyclical components. By leveraging the treated unit's historical data to forecast its trend and using control units only for cyclical fluctuations, our divide-and-conquer strategy eliminates spurious correlations and improves the robustness of counterfactual prediction in macroeconomic applications. As empirical illustrations, we examine the cases of German reunification and the handover of Hong Kong, demonstrating the advantages of the proposed approach. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:14:14 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Shi",
"Zhentao",
""
],
[
"Xi",
"Jin",
""
],
[
"Xie",
"Haitian",
""
]
] |
2505.22390 | Daojin Fan | Daojin Fan, Guoding Liu, Shaowei Li, Ming Gong, Dachao Wu, Yiming Zhang, Chen Zha, Fusheng Chen, Sirui Cao, Yangsen Ye, Qingling Zhu, Chong Ying, Shaojun Guo, Haoran Qian, Yulin Wu, Hui Deng, Gang Wu, Cheng-Zhi Peng, Xiongfeng Ma, Xiaobo Zhu, Jian-Wei Pan | Calibrating quantum gates up to 52 qubits in a superconducting processor | 25 pages, 10 figures | npj Quantum Information,2025,11:33 | 10.1038/s41534-025-00983-5 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Benchmarking large-scale quantum gates, typically involving multiple native two-qubit and singlequbit gates, is crucial in quantum computing. Global fidelity, encompassing information about intergate correlations, offers a comprehensive metric for evaluating and optimizing gate performance, unlike the fidelities of individual local native gates. In this work, utilizing the character-average benchmarking protocol implementable in a shallow circuit, we successfully benchmark gate fidelities up to 52 qubits. Notably, we achieved a fidelity of 63.09$\pm $0.23% for a 44-qubit parallel CZ gate. Utilizing the global fidelity of the parallel CZ gate, we explore the correlations among local CZ gates by introducing an inter-gate correlation metric, enabling one to simultaneously quantify crosstalk error when benchmarking gate fidelity. Finally, we apply our methods in gate optimization. By leveraging global fidelity for optimization, we enhance the fidelity of a 6-qubit parallel CZ gate from 87.65% to 92.04% and decrease the gate correlation from 3.53% to 3.22%, compared to local gate fidelitybased optimization. The experimental results align well with our established composite noise model, incorporating depolarizing and ZZ-coupling noises, and provide valuable insight into further study and mitigation of correlated noise. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:17:00 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Fan",
"Daojin",
""
],
[
"Liu",
"Guoding",
""
],
[
"Li",
"Shaowei",
""
],
[
"Gong",
"Ming",
""
],
[
"Wu",
"Dachao",
""
],
[
"Zhang",
"Yiming",
""
],
[
"Zha",
"Chen",
""
],
[
"Chen",
"Fusheng",
""
],
[
"Cao",
"Sirui",
""
],
[
"Ye",
"Yangsen",
""
],
[
"Zhu",
"Qingling",
""
],
[
"Ying",
"Chong",
""
],
[
"Guo",
"Shaojun",
""
],
[
"Qian",
"Haoran",
""
],
[
"Wu",
"Yulin",
""
],
[
"Deng",
"Hui",
""
],
[
"Wu",
"Gang",
""
],
[
"Peng",
"Cheng-Zhi",
""
],
[
"Ma",
"Xiongfeng",
""
],
[
"Zhu",
"Xiaobo",
""
],
[
"Pan",
"Jian-Wei",
""
]
] |
2505.22391 | Difan Zou | Yi Zhang, Difan Zou | Physics-Informed Distillation of Diffusion Models for PDE-Constrained Generation | 23 pages, 5 figures, 4 tables | null | null | null | cs.LG cs.AI cs.CE cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | Modeling physical systems in a generative manner offers several advantages, including the ability to handle partial observations, generate diverse solutions, and address both forward and inverse problems. Recently, diffusion models have gained increasing attention in the modeling of physical systems, particularly those governed by partial differential equations (PDEs). However, diffusion models only access noisy data $\boldsymbol{x}_t$ at intermediate steps, making it infeasible to directly enforce constraints on the clean sample $\boldsymbol{x}_0$ at each noisy level. As a workaround, constraints are typically applied to the expectation of clean samples $\mathbb{E}[\boldsymbol{x}_0|\boldsymbol{x}_t]$, which is estimated using the learned score network. However, imposing PDE constraints on the expectation does not strictly represent the one on the true clean data, known as Jensen's Gap. This gap creates a trade-off: enforcing PDE constraints may come at the cost of reduced accuracy in generative modeling. To address this, we propose a simple yet effective post-hoc distillation approach, where PDE constraints are not injected directly into the diffusion process, but instead enforced during a post-hoc distillation stage. We term our method as Physics-Informed Distillation of Diffusion Models (PIDDM). This distillation not only facilitates single-step generation with improved PDE satisfaction, but also support both forward and inverse problem solving and reconstruction from randomly partial observation. Extensive experiments across various PDE benchmarks demonstrate that PIDDM significantly improves PDE satisfaction over several recent and competitive baselines, such as PIDM, DiffusionPDE, and ECI-sampling, with less computation overhead. Our approach can shed light on more efficient and effective strategies for incorporating physical constraints into diffusion models. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:17:58 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Zhang",
"Yi",
""
],
[
"Zou",
"Difan",
""
]
] |
2505.22392 | San-Dong Guo | San-Dong Guo, Liguo Zhang, Xiao-Shu Guo and Gangqiang Zhu | Weak valley-layer coupling and valley polarization in centrosymmetric $\mathrm{FeCl_2}$ monolayer | 7 pages, 6 figures | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Using the valley degree of freedom as a carrier of information for storage and processing, valley polarization plays a crucial role. A variety of mechanisms for valley polarization have been proposed, among which the valley-layer coupling mechanism involves the induction of valley polarization by an out-of-plane electric field. Here, through first-principles calculations, it is found that the weak valley-layer coupling can exist in centrosymmetric $\mathrm{FeCl_2}$ monolayer. It is crucial to note that valley-layer coupling only occurs with out-of-plane magnetization and vanishes with in-plane magnetization. Compared to monolayers with strong valley-layer coupling, $\mathrm{FeCl_2}$ requires an extremely strong electric field to achieve the same magnitude of valley splitting. Valley polarization switching can be achieved by manipulating the directions of magnetization and electric field. Reversing only one of these directions switches the valley polarization, whereas reversing both simultaneously leaves it unchanged. Moreover, the simply stacked bilayer $\mathrm{FeCl_2}$, as a $PT$-antiferromagnet, can spontaneously achieve valley polarization without an external electric field, highlighting its potential for miniaturization, ultradensity, and ultrafast performance. Our work provides guidelines for identifying materials with weak valley-layer coupling, and further enables the regulation of valley polarization through electric field and stacking engineering. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:18:31 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Guo",
"San-Dong",
""
],
[
"Zhang",
"Liguo",
""
],
[
"Guo",
"Xiao-Shu",
""
],
[
"Zhu",
"Gangqiang",
""
]
] |
2505.22393 | Chetan Lodha | Chetan lodha, Manak Parmar and Ajay Kumar Rai | Exploring Charm Bound States: Mass Spectra and Decay Dynamics of D Mesons and $Cq\bar{q}\bar{q}$ Tetraquarks | 30 pages; 10 figures | null | null | null | hep-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Motivated by the discovery of several charm states exhibiting tetraquark-like characteristics at BESIII and LHCb, this study investigates the spectroscopy and decay properties of D mesons and tetraquark states with quark content $Cq\bar{q}\bar{q}$ within the diquark - antidiquark framework. The analysis is performed using a potential model based on the Cornell potential, considering color antitriplet - triplet configurations. Mass spectra are computed for both mesons and tetraquarks, while their decay behaviors are examined using the factorization approach for D mesons and Fierz rearrangement for tetraquark decays. Theoretical results are compared with experimentally observed resonances to improve our understanding of charm quark bound systems. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:21:42 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"lodha",
"Chetan",
""
],
[
"Parmar",
"Manak",
""
],
[
"Rai",
"Ajay Kumar",
""
]
] |
2505.22394 | Fan Fei | Fan Fei, Jiajun Tang, Fei-Peng Tian, Boxin Shi, Ping Tan | PacTure: Efficient PBR Texture Generation on Packed Views with Visual Autoregressive Models | 20 pages, 7 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present PacTure, a novel framework for generating physically-based rendering (PBR) material textures from an untextured 3D mesh, a text description, and an optional image prompt. Early 2D generation-based texturing approaches generate textures sequentially from different views, resulting in long inference times and globally inconsistent textures. More recent approaches adopt multi-view generation with cross-view attention to enhance global consistency, which, however, limits the resolution for each view. In response to these weaknesses, we first introduce view packing, a novel technique that significantly increases the effective resolution for each view during multi-view generation without imposing additional inference cost, by formulating the arrangement of multi-view maps as a 2D rectangle bin packing problem. In contrast to UV mapping, it preserves the spatial proximity essential for image generation and maintains full compatibility with current 2D generative models. To further reduce the inference cost, we enable fine-grained control and multi-domain generation within the next-scale prediction autoregressive framework to create an efficient multi-view multi-domain generative backbone. Extensive experiments show that PacTure outperforms state-of-the-art methods in both quality of generated PBR textures and efficiency in training and inference. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:23:30 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Fei",
"Fan",
""
],
[
"Tang",
"Jiajun",
""
],
[
"Tian",
"Fei-Peng",
""
],
[
"Shi",
"Boxin",
""
],
[
"Tan",
"Ping",
""
]
] |
2505.22395 | Asghar Ullah | Asghar Ullah, \"Ozg\"ur E. M\"ustecapl{\i}o\u{g}lu, Matteo G. A. Paris | Configuration-dependent precision in magnetometry and thermometry using multi-qubit quantum sensors | 15 pages, 10 figures; comments are welcome | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We study the performance of quantum sensors composed of four qubits arranged in different geometries for magnetometry and thermometry. The qubits interact via the transverse-field Ising model with both ferromagnetic and antiferromagnetic couplings, maintained in thermal equilibrium with a heat bath under an external magnetic field. Using quantum Fisher information (QFI), we evaluate the metrological precision of these sensors. For ferromagnetic couplings, weakly connected graphs (e.g., the chain graph, P_4) perform optimally in estimating weak magnetic fields, whereas highly connected graphs (e.g., the complete graph, K_4) excel at strong fields. Conversely, K_4 achieves the highest sensitivity for temperature estimation in the weak-field regime. In the antiferromagnetic case, we uncover a fundamental trade-off dictated by spectral degeneracy: configurations with non-degenerate energy spectra - such as the pan-like graph (three qubits in a triangle with the fourth attached) - exhibit strong magnetic field sensitivity due to their pronounced response to perturbations. In contrast, symmetric structures like the square graph, featuring degenerate energy levels (particularly ground-state degeneracy), are better suited for precise thermometry. Notably, our four-qubit sensors achieve peak precision in the low-temperature, weak-field regime. Finally, we introduce a spectral sensitivity measure that quantifies energy spectrum deformations under small perturbations, offering a tool to optimize magnetometric performance. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:23:51 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ullah",
"Asghar",
""
],
[
"Müstecaplıoğlu",
"Özgür E.",
""
],
[
"Paris",
"Matteo G. A.",
""
]
] |
2505.22396 | Xudong Li | Xudong Li, Mengdan Zhang, Peixian Chen, Xiawu Zheng, Yan Zhang, Jingyuan Zheng, Yunhang Shen, Ke Li, Chaoyou Fu, Xing Sun, Rongrong Ji | Zooming from Context to Cue: Hierarchical Preference Optimization for Multi-Image MLLMs | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-modal Large Language Models (MLLMs) excel at single-image tasks but struggle with multi-image understanding due to cross-modal misalignment, leading to hallucinations (context omission, conflation, and misinterpretation). Existing methods using Direct Preference Optimization (DPO) constrain optimization to a solitary image reference within the input sequence, neglecting holistic context modeling. We propose Context-to-Cue Direct Preference Optimization (CcDPO), a multi-level preference optimization framework that enhances per-image perception in multi-image settings by zooming into visual clues -- from sequential context to local details. It features: (i) Context-Level Optimization : Re-evaluates cognitive biases underlying MLLMs' multi-image context comprehension and integrates a spectrum of low-cost global sequence preferences for bias mitigation. (ii) Needle-Level Optimization : Directs attention to fine-grained visual details through region-targeted visual prompts and multimodal preference supervision. To support scalable optimization, we also construct MultiScope-42k, an automatically generated dataset with high-quality multi-level preference pairs. Experiments show that CcDPO significantly reduces hallucinations and yields consistent performance gains across general single- and multi-image tasks. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:24:02 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Li",
"Xudong",
""
],
[
"Zhang",
"Mengdan",
""
],
[
"Chen",
"Peixian",
""
],
[
"Zheng",
"Xiawu",
""
],
[
"Zhang",
"Yan",
""
],
[
"Zheng",
"Jingyuan",
""
],
[
"Shen",
"Yunhang",
""
],
[
"Li",
"Ke",
""
],
[
"Fu",
"Chaoyou",
""
],
[
"Sun",
"Xing",
""
],
[
"Ji",
"Rongrong",
""
]
] |
2505.22397 | Jules Tilly Dr | Christoph Brunken, Olivier Peltre, Heloise Chomet, Lucien Walewski, Manus McAuliffe, Valentin Heyraud, Solal Attias, Martin Maarand, Yessine Khanfir, Edan Toledo, Fabio Falcioni, Marie Bluntzer, Silvia Acosta-Guti\'errez, Jules Tilly | Machine Learning Interatomic Potentials: library for efficient training, model development and simulation of molecular systems | 18 pages, 6 tables, 3 figures | null | null | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine Learning Interatomic Potentials (MLIP) are a novel in silico approach for molecular property prediction, creating an alternative to disrupt the accuracy/speed trade-off of empirical force fields and density functional theory (DFT). In this white paper, we present our MLIP library which was created with two core aims: (1) provide to industry experts without machine learning background a user-friendly and computationally efficient set of tools to experiment with MLIP models, (2) provide machine learning developers a framework to develop novel approaches fully integrated with molecular dynamics tools. The library includes in this release three model architectures (MACE, NequIP, and ViSNet), and two molecular dynamics (MD) wrappers (ASE, and JAX-MD), along with a set of pre-trained organics models. The seamless integration with JAX-MD, in particular, facilitates highly efficient MD simulations, bringing MLIP models significantly closer to industrial application. The library is available on GitHub and on PyPI under the Apache license 2.0. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:24:47 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Brunken",
"Christoph",
""
],
[
"Peltre",
"Olivier",
""
],
[
"Chomet",
"Heloise",
""
],
[
"Walewski",
"Lucien",
""
],
[
"McAuliffe",
"Manus",
""
],
[
"Heyraud",
"Valentin",
""
],
[
"Attias",
"Solal",
""
],
[
"Maarand",
"Martin",
""
],
[
"Khanfir",
"Yessine",
""
],
[
"Toledo",
"Edan",
""
],
[
"Falcioni",
"Fabio",
""
],
[
"Bluntzer",
"Marie",
""
],
[
"Acosta-Gutiérrez",
"Silvia",
""
],
[
"Tilly",
"Jules",
""
]
] |
2505.22399 | Emiliano Dall'Anese | Damola Ajeyemi, Yiting Chen, Antonin Colot, Jorge Cortes, Emiliano Dall'Anese | Learning to Pursue AC Optimal Power Flow Solutions with Feasibility Guarantees | null | null | null | null | math.OC cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | This paper focuses on an AC optimal power flow (OPF) problem for distribution feeders equipped with controllable distributed energy resources (DERs). We consider a solution method that is based on a continuous approximation of the projected gradient flow - referred to as the safe gradient flow - that incorporates voltage and current information obtained either through real-time measurements or power flow computations. These two setups enable both online and offline implementations. The safe gradient flow involves the solution of convex quadratic programs (QPs). To enhance computational efficiency, we propose a novel framework that employs a neural network approximation of the optimal solution map of the QP. The resulting method has two key features: (a) it ensures that the DERs' setpoints are practically feasible, even for an online implementation or when an offline algorithm has an early termination; (b) it ensures convergence to a neighborhood of a strict local optimizer of the AC OPF. The proposed method is tested on a 93-node distribution system with realistic loads and renewable generation. The test shows that our method successfully regulates voltages within limits during periods with high renewable generation. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:26:21 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Ajeyemi",
"Damola",
""
],
[
"Chen",
"Yiting",
""
],
[
"Colot",
"Antonin",
""
],
[
"Cortes",
"Jorge",
""
],
[
"Dall'Anese",
"Emiliano",
""
]
] |
2505.22400 | Zehao Li | Zehao Li, Hao Jiang, Yujun Cai, Jianing Chen, Baolong Bi, Shuqin Gao, Honglong Zhao, Yiwei Wang, Tianlu Mao, Zhaoqi Wang | STDR: Spatio-Temporal Decoupling for Real-Time Dynamic Scene Rendering | null | null | null | null | cs.GR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although dynamic scene reconstruction has long been a fundamental challenge in 3D vision, the recent emergence of 3D Gaussian Splatting (3DGS) offers a promising direction by enabling high-quality, real-time rendering through explicit Gaussian primitives. However, existing 3DGS-based methods for dynamic reconstruction often suffer from \textit{spatio-temporal incoherence} during initialization, where canonical Gaussians are constructed by aggregating observations from multiple frames without temporal distinction. This results in spatio-temporally entangled representations, making it difficult to model dynamic motion accurately. To overcome this limitation, we propose \textbf{STDR} (Spatio-Temporal Decoupling for Real-time rendering), a plug-and-play module that learns spatio-temporal probability distributions for each Gaussian. STDR introduces a spatio-temporal mask, a separated deformation field, and a consistency regularization to jointly disentangle spatial and temporal patterns. Extensive experiments demonstrate that incorporating our module into existing 3DGS-based dynamic scene reconstruction frameworks leads to notable improvements in both reconstruction quality and spatio-temporal consistency across synthetic and real-world benchmarks. | [
{
"version": "v1",
"created": "Wed, 28 May 2025 14:26:41 GMT"
}
] | 2025-05-29T00:00:00 | [
[
"Li",
"Zehao",
""
],
[
"Jiang",
"Hao",
""
],
[
"Cai",
"Yujun",
""
],
[
"Chen",
"Jianing",
""
],
[
"Bi",
"Baolong",
""
],
[
"Gao",
"Shuqin",
""
],
[
"Zhao",
"Honglong",
""
],
[
"Wang",
"Yiwei",
""
],
[
"Mao",
"Tianlu",
""
],
[
"Wang",
"Zhaoqi",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.