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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2502.13497 | Piyawat Lertvittayakumjorn | Piyawat Lertvittayakumjorn, David Kinney, Vinodkumar Prabhakaran, Donald Martin Jr., Sunipa Dev | Towards Geo-Culturally Grounded LLM Generations | ACL 2025 (main conference) | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Generative large language models (LLMs) have demonstrated gaps in diverse cultural awareness across the globe. We investigate the effect of retrieval augmented generation and search-grounding techniques on LLMs' ability to display familiarity with various national cultures. Specifically, we compare the performance of standard LLMs, LLMs augmented with retrievals from a bespoke knowledge base (i.e., KB grounding), and LLMs augmented with retrievals from a web search (i.e., search grounding) on multiple cultural awareness benchmarks. We find that search grounding significantly improves the LLM performance on multiple-choice benchmarks that test propositional knowledge (e.g., cultural norms, artifacts, and institutions), while KB grounding's effectiveness is limited by inadequate knowledge base coverage and a suboptimal retriever. However, search grounding also increases the risk of stereotypical judgments by language models and fails to improve evaluators' judgments of cultural familiarity in a human evaluation with adequate statistical power. These results highlight the distinction between propositional cultural knowledge and open-ended cultural fluency when it comes to evaluating LLMs' cultural awareness. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 07:29:58 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Feb 2025 06:38:50 GMT"
},
{
"version": "v3",
"created": "Mon, 16 Jun 2025 20:58:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lertvittayakumjorn",
"Piyawat",
""
],
[
"Kinney",
"David",
""
],
[
"Prabhakaran",
"Vinodkumar",
""
],
[
"Martin",
"Donald",
"Jr."
],
[
"Dev",
"Sunipa",
""
]
] |
2502.13662 | Nikita Puchkin | Konstantin Yakovlev and Nikita Puchkin | Generalization error bound for denoising score matching under relaxed manifold assumption | Accepted for presentation at the 38th Annual Conference on Learning Theory (COLT 2025) | null | null | null | cs.LG math.ST stat.ML stat.TH | http://creativecommons.org/licenses/by/4.0/ | We examine theoretical properties of the denoising score matching estimate. We model the density of observations with a nonparametric Gaussian mixture. We significantly relax the standard manifold assumption allowing the samples step away from the manifold. At the same time, we are still able to leverage a nice distribution structure. We derive non-asymptotic bounds on the approximation and generalization errors of the denoising score matching estimate. The rates of convergence are determined by the intrinsic dimension. Furthermore, our bounds remain valid even if we allow the ambient dimension grow polynomially with the sample size. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 12:14:52 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Feb 2025 15:51:11 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 12:04:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yakovlev",
"Konstantin",
""
],
[
"Puchkin",
"Nikita",
""
]
] |
2502.13695 | Enam Chowdhury | Joseph R. Smith, Ziyao Su, Simin Zhang, Charles Varin, Vitaly E. Gruzdev, and Enam A. Chowdhury | A Fully Three-Dimensional Kinetic Particle-In-Cell Framework for Modeling Laser-Dielectric Interactions: Few-Cycle Pulse Damage | null | null | 10.1017/hpl.2025.10041 | null | physics.optics physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a fully three-dimensional kinetic framework for modeling intense short pulse lasers interacting with dielectric materials. Our work modifies the open-source Particle-In-Cell (PIC) code EPOCH to include new models for molecular photoionization and dielectric optical response. We use this framework to model the laser-induced damage of dielectric materials by few-cycle laser pulses. The framework is benchmarked against experimental results for bulk silica targets and then applied to model multi-layer dielectric mirrors with a sequence of simulations with varying laser fluence. This allows us to better understand the laser damage process by providing new insight into energy absorption, excited particle dynamics, and nonthermal excited particle dist | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 13:09:52 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Smith",
"Joseph R.",
""
],
[
"Su",
"Ziyao",
""
],
[
"Zhang",
"Simin",
""
],
[
"Varin",
"Charles",
""
],
[
"Gruzdev",
"Vitaly E.",
""
],
[
"Chowdhury",
"Enam A.",
""
]
] |
2502.14445 | Lorenzo Pacchiardi | Lorenzo Pacchiardi, Konstantinos Voudouris, Ben Slater, Fernando Mart\'inez-Plumed, Jos\'e Hern\'andez-Orallo, Lexin Zhou, Wout Schellaert | PredictaBoard: Benchmarking LLM Score Predictability | Accepted at ACL Findings 2025 | null | null | null | cs.CL cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Despite possessing impressive skills, Large Language Models (LLMs) often fail unpredictably, demonstrating inconsistent success in even basic common sense reasoning tasks. This unpredictability poses a significant challenge to ensuring their safe deployment, as identifying and operating within a reliable "safe zone" is essential for mitigating risks. To address this, we present PredictaBoard, a novel collaborative benchmarking framework designed to evaluate the ability of score predictors (referred to as assessors) to anticipate LLM errors on specific task instances (i.e., prompts) from existing datasets. PredictaBoard evaluates pairs of LLMs and assessors by considering the rejection rate at different tolerance errors. As such, PredictaBoard stimulates research into developing better assessors and making LLMs more predictable, not only with a higher average performance. We conduct illustrative experiments using baseline assessors and state-of-the-art LLMs. PredictaBoard highlights the critical need to evaluate predictability alongside performance, paving the way for safer AI systems where errors are not only minimised but also anticipated and effectively mitigated. Code for our benchmark can be found at https://github.com/Kinds-of-Intelligence-CFI/PredictaBoard | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 10:52:38 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:34:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pacchiardi",
"Lorenzo",
""
],
[
"Voudouris",
"Konstantinos",
""
],
[
"Slater",
"Ben",
""
],
[
"Martínez-Plumed",
"Fernando",
""
],
[
"Hernández-Orallo",
"José",
""
],
[
"Zhou",
"Lexin",
""
],
[
"Schellaert",
"Wout",
""
]
] |
2502.15077 | Sanghyun Yi | Sanghyun Yi, Qingfeng Liu, Mostafa El-Khamy | Hardware-Friendly Static Quantization Method for Video Diffusion Transformers | Accepted to MIPR 2025 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Diffusion Transformers for video generation have gained significant research interest since the impressive performance of SORA. Efficient deployment of such generative-AI models on GPUs has been demonstrated with dynamic quantization. However, resource-constrained devices cannot support dynamic quantization, and need static quantization of the models for their efficient deployment on AI processors. In this paper, we propose a novel method for the post-training quantization of OpenSora\cite{opensora}, a Video Diffusion Transformer, without relying on dynamic quantization techniques. Our approach employs static quantization, achieving video quality comparable to FP16 and dynamically quantized ViDiT-Q methods, as measured by CLIP, and VQA metrics. In particular, we utilize per-step calibration data to adequately provide a post-training statically quantized model for each time step, incorporating channel-wise quantization for weights and tensor-wise quantization for activations. By further applying the smooth-quantization technique, we can obtain high-quality video outputs with the statically quantized models. Extensive experimental results demonstrate that static quantization can be a viable alternative to dynamic quantization for video diffusion transformers, offering a more efficient approach without sacrificing performance. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 22:29:24 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Mar 2025 05:17:19 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 05:03:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yi",
"Sanghyun",
""
],
[
"Liu",
"Qingfeng",
""
],
[
"El-Khamy",
"Mostafa",
""
]
] |
2502.15400 | Yuri Fukaya | Yuri Fukaya, Bo Lu, Keiji Yada, Yukio Tanaka, Jorge Cayao | Superconducting phenomena in systems with unconventional magnets | 58 pages, 23 figures. Invited topical review. Comments are welcome | null | null | null | cond-mat.supr-con cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | In this work we review the recent advances on superconducting phenomena in junctions formed by superconductors and unconventional magnets. Conventional magnets, such as ferromagnets and antiferromagnets, are characterized by broken time-reversal symmetry but only ferromagnets produce a finite net magnetization due to parallel spin alignment and spin-split bands in momentum. Very recently, a new type of magnets has been reported and here we refer to them as \emph{unconventional magnets} because they exhibit special properties of both ferromagnets and antiferromagnets: they exhibit zero net magnetization (like antiferromagnets) and a nonrelativistic spin splitting of energy bands (like ferromagnets), both leading to anisotropic spin-polarized Fermi surfaces. An interesting property of unconventional magnets is that their magnetic order can be even or odd with respect to momentum, where \emph{$d$-wave altermagnets} and \emph{$p$-wave magnets} are the most representative examples. In this regard, $d$-wave altermagnets and $p$-wave magnets are seen as counterparts in magnetism of the unconventional $d$- and $p$-wave superconducting states, respectively. While the impact of conventional magnetism on superconductivity has been largely studied, the combination of unconventional magnets and superconductivity has only lately attracted considerably attention. This work provides a comprehensive review of the recent progress on the interplay between superconductivity and unconventional magnets. In particular, we focus on the fundamental emerging superconducting phenomena and also discuss the potential implications towards quantum applications. | [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 11:59:46 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 00:42:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fukaya",
"Yuri",
""
],
[
"Lu",
"Bo",
""
],
[
"Yada",
"Keiji",
""
],
[
"Tanaka",
"Yukio",
""
],
[
"Cayao",
"Jorge",
""
]
] |
2502.15750 | Emil Vogt | Emil Vogt and \'Alvaro Fern\'andez Corral and Yahya Saleh and Andrey Yachmenev | Transferability and interpretability of vibrational normalizing-flow coordinates | null | null | null | null | physics.chem-ph physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | The choice of vibrational coordinates is crucial for the accuracy, efficiency, and interpretability of molecular vibrational dynamics and spectra calculations. We explore the recently proposed normalizing-flow vibrational coordinates, which are learned molecule-specific coordinate transformations optimized for a given basis set. Much like how spherical coordinates naturally simplify the hydrogen atom by embedding physical insight into the coordinate system, normalizing-flow coordinates offload complexity from the basis functions into the coordinate transformation itself. This shift not only improves basis-set convergence, but also enhances interpretability of vibrational motions. We provide an analysis of the utility, interpretation and associated constraints of normalizing-flow vibrational coordinates. Moreover, we demonstrate that these coordinates can be generalized across different isotopologues, and even structurally related molecules, achieved with minimal fine-tuning of selected output parameters. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 14:16:02 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Feb 2025 07:58:55 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 12:31:08 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Vogt",
"Emil",
""
],
[
"Corral",
"Álvaro Fernández",
""
],
[
"Saleh",
"Yahya",
""
],
[
"Yachmenev",
"Andrey",
""
]
] |
2502.15910 | Zheyuan Liu | Zheyuan Liu, Guangyao Dou, Xiangchi Yuan, Chunhui Zhang, Zhaoxuan Tan, Meng Jiang | Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models | ACL 2025 Main Conference | null | null | null | cs.CL | http://creativecommons.org/publicdomain/zero/1.0/ | Generative models such as Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) trained on massive datasets can lead them to memorize and inadvertently reveal sensitive information, raising ethical and privacy concerns. While some prior works have explored this issue in the context of LLMs, it presents a unique challenge for MLLMs due to the entangled nature of knowledge across modalities, making comprehensive unlearning more difficult. To address this challenge, we propose Modality Aware Neuron Unlearning (MANU), a novel unlearning framework for MLLMs designed to selectively clip neurons based on their relative importance to the targeted forget data, curated for different modalities. Specifically, MANU consists of two stages: important neuron selection and selective pruning. The first stage identifies and collects the most influential neurons across modalities relative to the targeted forget knowledge, while the second stage is dedicated to pruning those selected neurons. MANU effectively isolates and removes the neurons that contribute most to the forget data within each modality, while preserving the integrity of retained knowledge. Our experiments conducted across various MLLM architectures illustrate that MANU can achieve a more balanced and comprehensive unlearning in each modality without largely affecting the overall model utility. | [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 19:54:46 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:29:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Zheyuan",
""
],
[
"Dou",
"Guangyao",
""
],
[
"Yuan",
"Xiangchi",
""
],
[
"Zhang",
"Chunhui",
""
],
[
"Tan",
"Zhaoxuan",
""
],
[
"Jiang",
"Meng",
""
]
] |
2502.16043 | Teanna Barrett | Teanna Barrett, Chinasa T. Okolo, B. Biira, Eman Sherif, Amy X. Zhang, and Leilani Battle | African Data Ethics: A Discursive Framework for Black Decolonial Data Science | null | null | null | null | cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The shift towards pluralism in global data ethics acknowledges the importance of including perspectives from the Global Majority to develop responsible data science practices that mitigate systemic harms in the current data science ecosystem. Sub-Saharan African (SSA) practitioners, in particular, are disseminating progressive data ethics principles and best practices for identifying and navigating anti-blackness and data colonialism. To center SSA voices in the global data ethics discourse, we present a framework for African data ethics informed by the thematic analysis of an interdisciplinary corpus of 50 documents. Our framework features six major principles: 1) Challenge Power Asymmetries, 2) Assert Data Self-Determination, 3) Invest in Local Data Institutions & Infrastructures, 4) Utilize Communalist Practices, 5) Center Communities on the Margins, and 6) Uphold Common Good. We compare our framework to seven particularist data ethics frameworks to find similar conceptual coverage but diverging interpretations of shared values. Finally, we discuss how African data ethics demonstrates the operational value of data ethics frameworks. Our framework highlights Sub-Saharan Africa as a pivotal site of responsible data science by promoting the practice of communalism, self-determination, and cultural preservation. | [
{
"version": "v1",
"created": "Sat, 22 Feb 2025 02:34:46 GMT"
},
{
"version": "v2",
"created": "Wed, 26 Feb 2025 23:30:56 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 17:05:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Barrett",
"Teanna",
""
],
[
"Okolo",
"Chinasa T.",
""
],
[
"Biira",
"B.",
""
],
[
"Sherif",
"Eman",
""
],
[
"Zhang",
"Amy X.",
""
],
[
"Battle",
"Leilani",
""
]
] |
2502.16379 | Christopher Bartel | Nathan J. Szymanski, Alexander Smith, Prodromos Daoutidis, Christopher J. Bartel | Topological descriptors for the electron density of inorganic solids | null | ACS Materials Lett. 2025, 7, 6, 2158-2164 | 10.1021/acsmaterialslett.5c00390 | null | cond-mat.mtrl-sci physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Descriptors play an important role in data-driven materials design. While most descriptors of crystalline materials emphasize structure and composition, they often neglect the electron density - a complex yet fundamental quantity that governs material properties. Here, we introduce Betti curves as topological descriptors that compress electron densities into compact representations. Derived from persistent homology, Betti curves capture bonding characteristics by encoding components, cycles, and voids across varied electron density thresholds. Machine learning models trained on Betti curves outperform those trained on raw electron densities by an average of 33 percentage points in classifying structure prototypes, predicting thermodynamic stability, and distinguishing metals from non-metals. Shannon entropy calculations reveal that Betti curves retain comparable information content to electron density while requiring two orders of magnitude less data. By combining expressive power with compact representation, Betti curves highlight the potential of topological data analysis to advance materials design. | [
{
"version": "v1",
"created": "Sat, 22 Feb 2025 22:59:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:55:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Szymanski",
"Nathan J.",
""
],
[
"Smith",
"Alexander",
""
],
[
"Daoutidis",
"Prodromos",
""
],
[
"Bartel",
"Christopher J.",
""
]
] |
2502.16611 | Shitong Xu | Shitong Xu, Yiyuan Yang, Niki Trigoni, Andrew Markham | Target Speaker Extraction through Comparing Noisy Positive and Negative Audio Enrollments | 11 pages, 6 figures | null | null | null | cs.SD cs.AI eess.AS | http://creativecommons.org/licenses/by/4.0/ | Target speaker extraction focuses on isolating a specific speaker's voice from an audio mixture containing multiple speakers. To provide information about the target speaker's identity, prior works have utilized clean audio samples as conditioning inputs. However, such clean audio examples are not always readily available. For instance, obtaining a clean recording of a stranger's voice at a cocktail party without leaving the noisy environment is generally infeasible. Limited prior research has explored extracting the target speaker's characteristics from noisy enrollments, which may contain overlapping speech from interfering speakers. In this work, we explore a novel enrollment strategy that encodes target speaker information from the noisy enrollment by comparing segments where the target speaker is talking (Positive Enrollments) with segments where the target speaker is silent (Negative Enrollments). Experiments show the effectiveness of our model architecture, which achieves over 2.1 dB higher SI-SNRi compared to prior works in extracting the monaural speech from the mixture of two speakers. Additionally, the proposed two-stage training strategy accelerates convergence, reducing the number of optimization steps required to reach 3 dB SNR by 60\%. Overall, our method achieves state-of-the-art performance in the monaural target speaker extraction conditioned on noisy enrollments. | [
{
"version": "v1",
"created": "Sun, 23 Feb 2025 15:33:44 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:10:07 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Shitong",
""
],
[
"Yang",
"Yiyuan",
""
],
[
"Trigoni",
"Niki",
""
],
[
"Markham",
"Andrew",
""
]
] |
2502.16787 | Tomohiro Harada | Tomohiro Harada, Chiang-Mei Chen, Rituparna Mandal | Singularity resolution and regular black hole formation in gravitational collapse in asymptotically safe gravity | 34 pages, 6 figures, accepted for publication in Physical Review D | null | null | RUP-25-3 | gr-qc astro-ph.CO hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We adopt an effective action inspired by asymptotically safe gravity, in which the effective gravitational constant is parametrized as $G(\epsilon) = G_{N} /[1 + \tilde{\omega} (G_{N}^{2} \epsilon)^{\alpha}]$, where $G_{N}$ and $\epsilon$ denote Newton's gravitational constant and the energy density of the matter field, respectively, with two dimensionless model parameters, $\tilde{\omega}$ and $\alpha$. Within this framework, we investigate the complete gravitational collapse of a homogeneous ball of perfect fluid and find that singularity is completely resolved for $\alpha > 1$ but not for $1/2 \le \alpha \le 1$. The case of $0 < \alpha < 1/2$ is inconsistent with asymptotic safety. Moreover, we note that although the singularity cannot be fully resolved for $\alpha = 1$, it is significantly weakened by quantum gravity effects. Furthermore, we successfully construct a static exterior metric which, together with the interior solution, describes the dynamical formation of regular black holes in an asymptotically flat spacetime. The resulting regular black hole, obtained as the final static state, contains a de Sitter core and admits a static metric fully expressible in terms of the Lerch transcendent for general cases and in elementary functions for certain values of $\alpha$, including $\alpha = 2$. We also discuss the formation of gravastars and the late-time evaporation process of the regular black holes. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 02:45:29 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 03:00:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Harada",
"Tomohiro",
""
],
[
"Chen",
"Chiang-Mei",
""
],
[
"Mandal",
"Rituparna",
""
]
] |
2502.17060 | Andreas Loizou | Andreas Loizou and Dimitrios Tsoumakos | Analytics Modelling over Multiple Datasets using Vector Embeddings | null | The 36th International Conference on Database and Expert Systems Applications, DEXA 2025 | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The massive increase in the data volume and dataset availability for analysts compels researchers to focus on data content and select high-quality datasets to enhance the performance of analytics operators. While selecting high-quality data significantly boosts analytical accuracy and efficiency, the exact process is very challenging given large-scale dataset availability. To address this issue, we propose a novel methodology that infers the outcome of analytics operators by creating a model from the available datasets. Each dataset is transformed to a vector embedding representation generated by our proposed deep learning model NumTabData2Vec, where similarity search are employed. Through experimental evaluation, we compare the prediction performance and the execution time of our framework to another state-of-the-art modelling operator framework, illustrating that our approach predicts analytics outcomes accurately, and increases speedup. Furthermore, our vectorization model can project different real-world scenarios to a lower vector embedding representation accurately and distinguish them. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 11:21:08 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Apr 2025 08:36:47 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 15:45:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Loizou",
"Andreas",
""
],
[
"Tsoumakos",
"Dimitrios",
""
]
] |
2502.17421 | Penghui Yang | Penghui Yang, Cunxiao Du, Fengzhuo Zhang, Haonan Wang, Tianyu Pang, Chao Du, Bo An | LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | As Large Language Models (LLMs) can now process extremely long contexts, efficient inference over these extended inputs has become increasingly important, especially for emerging applications like LLM agents that highly depend on this capability. Speculative decoding (SD) offers a promising lossless acceleration technique compared to lossy alternatives such as quantization and model cascades. However, most state-of-the-art SD methods are trained on short texts (typically fewer than 4k tokens), making them unsuitable for long-context scenarios. Specifically, adapting these methods to long contexts presents three key challenges: (1) the excessive memory demands posed by draft models due to large Key-Value (KV) cache; (2) performance degradation resulting from the mismatch between short-context training and long-context inference; and (3) inefficiencies in tree attention mechanisms when managing long token sequences. This work introduces LongSpec, a framework that addresses these challenges through three core innovations: a memory-efficient draft model with a constant-sized KV cache; novel position indices that mitigate the training-inference mismatch; and an attention aggregation strategy that combines fast prefix computation with standard tree attention to enable efficient decoding. Experimental results confirm the effectiveness of LongSpec, achieving up to a 3.26x speedup over strong Flash Attention baselines across five long-context understanding datasets, as well as a 2.25x reduction in wall-clock time on the AIME24 long reasoning task with the QwQ model, demonstrating significant latency improvements for long-context applications. The code is available at https://github.com/sail-sg/LongSpec. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 18:53:31 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 05:58:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yang",
"Penghui",
""
],
[
"Du",
"Cunxiao",
""
],
[
"Zhang",
"Fengzhuo",
""
],
[
"Wang",
"Haonan",
""
],
[
"Pang",
"Tianyu",
""
],
[
"Du",
"Chao",
""
],
[
"An",
"Bo",
""
]
] |
2502.17514 | Hantao Lou | Hantao Lou, Changye Li, Jiaming Ji, Yaodong Yang | SAE-V: Interpreting Multimodal Models for Enhanced Alignment | 17 pages, 13 figures | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the integration of image modality, the semantic space of multimodal large language models (MLLMs) is more complex than text-only models, making their interpretability more challenging and their alignment less stable, particularly susceptible to low-quality data, which can lead to inconsistencies between modalities, hallucinations, and biased outputs. As a result, developing interpretability methods for MLLMs is crucial for improving alignment quality and efficiency. In text-only LLMs, Sparse Autoencoders (SAEs) have gained attention for their ability to interpret latent representations. However, extending SAEs to multimodal settings presents new challenges due to modality fusion and the difficulty of isolating cross-modal representations. To address these challenges, we introduce SAE-V, a mechanistic interpretability framework that extends the SAE paradigm to MLLMs. By identifying and analyzing interpretable features along with their corresponding data, SAE-V enables fine-grained interpretation of both model behavior and data quality, facilitating a deeper understanding of cross-modal interactions and alignment dynamics. Moreover, by utilizing cross-modal feature weighting, SAE-V provides an intrinsic data filtering mechanism to enhance model alignment without requiring additional models. Specifically, when applied to the alignment process of MLLMs, SAE-V-based data filtering methods could achieve more than 110% performance with less than 50% data. Our results highlight SAE-V's ability to enhance interpretability and alignment in MLLMs, providing insights into their internal mechanisms. | [
{
"version": "v1",
"created": "Sat, 22 Feb 2025 14:20:07 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 01:51:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lou",
"Hantao",
""
],
[
"Li",
"Changye",
""
],
[
"Ji",
"Jiaming",
""
],
[
"Yang",
"Yaodong",
""
]
] |
2502.17820 | Phuong Nam Vu | Nam P. Vu, Daniel Dong, Xiaohan Dan, Ningyi Lyu, Victor Batista, Yuan Liu | A Computational Framework for Simulations of Dissipative Non-Adiabatic Dynamics on Hybrid Oscillator-Qubit Quantum Devices | 66 pages including references and appendices, 14 figures, minor revisions from previous version | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a computational framework for simulating non-adiabatic vibronic dynamics on circuit quantum electrodynamics (cQED) platforms. Our approach leverages hybrid oscillator-qubit quantum hardware with mid-circuit measurements and resets, enabling the incorporation of environmental effects such as dissipation and dephasing. To demonstrate its capabilities, we simulate energy transfer dynamics in a triad model of photosynthetic chromophores inspired by natural antenna systems. We specifically investigate the role of dissipation during the relaxation dynamics following photoexcitation, where electronic transitions are coupled to the evolution of quantum vibrational modes. Our results indicate that hybrid oscillator-qubit devices, operating with noise levels below the intrinsic dissipation rates of typical molecular antenna systems, can achieve the simulation fidelity required for practical computations on near-term and early fault-tolerant quantum computing platforms. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 03:56:45 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:06:51 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Vu",
"Nam P.",
""
],
[
"Dong",
"Daniel",
""
],
[
"Dan",
"Xiaohan",
""
],
[
"Lyu",
"Ningyi",
""
],
[
"Batista",
"Victor",
""
],
[
"Liu",
"Yuan",
""
]
] |
2502.18032 | Jinrong Hu | Jinrong Hu | The dual Minkowski problem for positive indices | null | null | null | null | math.AP math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We derive the stability result of the dual curvature measure with near constant density in the even case. As an application, the existence and uniqueness of solutions to the even dual Minkowski problem for positive indices in $\mathbb{R}^{n+1}$ are obtained with $n\geq 1$, provided the density of the given measure is close to 1 in the $C^{\alpha}$ norm with $\alpha\in (0,1)$. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 09:46:52 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:50:52 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 09:18:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hu",
"Jinrong",
""
]
] |
2502.18145 | Ziyue Lin | Ziyue Lin, Siqi Shen, Zichen Cheng, Cheok Lam Lai, Siming Chen | Carbon and Silicon, Coexist or Compete? A Survey on Human-AI Interactions in Agent-based Modeling and Simulation | 36 pages, 3 figures | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent interest in human-AI interactions in agent-based modeling and simulation (ABMS) has grown rapidly due to the widespread utilization of large language models (LLMs). ABMS is an intelligent approach that simulates autonomous agents' behaviors within a defined environment to research emergent phenomena. Integrating LLMs into ABMS enables natural language interaction between humans and models. Meanwhile, it introduces new challenges that rely on human interaction to address. Human involvement can assist ABMS in adapting to flexible and complex research demands. However, systematic reviews of interactions that examine how humans and AI interact in ABMS are lacking. In this paper, we investigate existing works and propose a novel taxonomy to categorize the interactions derived from them. Specifically, human users refer to researchers who utilize ABMS tools to conduct their studies in our survey. We decompose interactions into five dimensions: the goals that users want to achieve (Why), the phases that users are involved (When), the components of the system (What), the roles of users (Who), and the means of interactions (How). Our analysis summarizes the findings that reveal existing interaction patterns. They provide researchers who develop interactions with comprehensive guidance on how humans and AI interact. We further discuss the unexplored interactions and suggest future research directions. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 12:15:22 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:23:37 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lin",
"Ziyue",
""
],
[
"Shen",
"Siqi",
""
],
[
"Cheng",
"Zichen",
""
],
[
"Lai",
"Cheok Lam",
""
],
[
"Chen",
"Siming",
""
]
] |
2502.18770 | Jiayi Fu | Jiayi Fu, Xuandong Zhao, Chengyuan Yao, Heng Wang, Qi Han, Yanghua Xiao | Reward Shaping to Mitigate Reward Hacking in RLHF | 24 pages | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Reinforcement Learning from Human Feedback (RLHF) is essential for aligning large language models (LLMs) with human values. However, RLHF is susceptible to \emph{reward hacking}, where the agent exploits flaws in the reward function rather than learning the intended behavior, thus degrading alignment. Although reward shaping helps stabilize RLHF and partially mitigate reward hacking, a systematic investigation into shaping techniques and their underlying principles remains lacking. To bridge this gap, we present a comprehensive study of the prevalent reward shaping methods. Our analysis suggests two key design principles: (1) the RL reward should be bounded, and (2) the RL reward benefits from rapid initial growth followed by gradual convergence. Guided by these insights, we propose Preference As Reward (PAR), a novel approach that leverages the latent preferences embedded within the reward model as the signal for reinforcement learning. We evaluated PAR on two base models, Gemma2-2B, and Llama3-8B, using two datasets, Ultrafeedback-Binarized and HH-RLHF. Experimental results demonstrate PAR's superior performance over other reward shaping methods. On the AlpacaEval 2.0 benchmark, PAR achieves a win rate of at least 5 percentage points higher than competing approaches. Furthermore, PAR exhibits remarkable data efficiency, requiring only a single reference reward for optimal performance, and maintains robustness against reward hacking even after two full epochs of training. The code is available at https://github.com/PorUna-byte/PAR, and the Work done during the internship at StepFun by Jiayi Fu. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 02:57:59 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Feb 2025 04:49:19 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 05:05:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fu",
"Jiayi",
""
],
[
"Zhao",
"Xuandong",
""
],
[
"Yao",
"Chengyuan",
""
],
[
"Wang",
"Heng",
""
],
[
"Han",
"Qi",
""
],
[
"Xiao",
"Yanghua",
""
]
] |
2502.18843 | An-Jun Liu | An-Jun Liu, Bryan K. Clark | Efficient optimization of neural network backflow for ab-initio quantum chemistry | This significant update features a major revision to the core algorithm by replacing the preselection with an efficient method constructing and intermittently updating the target space, new benchmark results for all-electron H2O and N2 systems, and two additional ablation studies providing deeper insights into the algorithmic enhancements | null | null | null | physics.chem-ph cond-mat.dis-nn physics.comp-ph quant-ph | http://creativecommons.org/licenses/by/4.0/ | The ground state of second-quantized quantum chemistry Hamiltonians is key to determining molecular properties. Neural quantum states (NQS) offer flexible and expressive wavefunction ansatze for this task but face two main challenges: highly peaked ground-state wavefunctions hinder efficient sampling, and local energy evaluations scale quartically with system size, incurring significant computational costs. In this work, we overcome these challenges by introducing a suite of algorithmic enhancements, which includes efficient periodic compact subspace construction, truncated local energy evaluations, improved stochastic sampling, and physics-informed modifications.
Applying these techniques to the neural network backflow (NNBF) ansatz, we demonstrate significant gains in both accuracy and scalability. Our enhanced method surpasses traditional quantum chemistry methods like CCSD and CCSD(T), outperforms other NQS approaches, and achieves competitive energies with state-of-the-art ab initio techniques such as HCI, ASCI, FCIQMC, and DMRG. A series of ablation and comparative studies quantifies the contribution of each enhancement to the observed improvements in accuracy and efficiency. Furthermore, we investigate the representational capacity of the ansatz, finding that its performance correlates with the inverse participation ratio (IPR), with more delocalized states being more challenging to approximate. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 05:31:08 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 17:48:52 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"An-Jun",
""
],
[
"Clark",
"Bryan K.",
""
]
] |
2502.19110 | Zhengping Jiang | Zhengping Jiang, Anqi Liu, Benjamin Van Durme | Conformal Linguistic Calibration: Trading-off between Factuality and Specificity | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Language model outputs are not always reliable, thus prompting research into how to adapt model responses based on uncertainty. Common approaches include: \emph{abstention}, where models refrain from generating responses when uncertain; and \emph{linguistic calibration}, where models hedge their statements using uncertainty quantifiers. However, abstention can withhold valuable information, while linguistically calibrated responses are often challenging to leverage in downstream tasks. We propose a unified view, Conformal Linguistic Calibration (CLC), which reinterprets linguistic calibration as \emph{answer set prediction}. First we present a framework connecting abstention and linguistic calibration through the lens of linguistic pragmatics. We then describe an implementation of CLC that allows for controlling the level of imprecision in model responses. Results demonstrate our method produces calibrated outputs with conformal guarantees on factual accuracy. Further, our approach enables fine-tuning models to perform uncertainty-aware adaptive claim rewriting, offering a controllable balance between factuality and specificity. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 13:01:49 GMT"
},
{
"version": "v2",
"created": "Fri, 13 Jun 2025 04:49:28 GMT"
},
{
"version": "v3",
"created": "Mon, 16 Jun 2025 18:13:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jiang",
"Zhengping",
""
],
[
"Liu",
"Anqi",
""
],
[
"Van Durme",
"Benjamin",
""
]
] |
2502.19113 | Thomas Nussle | Thomas Nussle, Stam Nicolis, Iason Sofos and Joseph Barker | Path integral spin dynamics with exchange and external field | 12 pages, 5 figures | null | null | null | quant-ph cond-mat.mtrl-sci physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | In this work, we propose a path integral-inspired formalism for computing the quantum thermal expectation values of spin systems, when subject to magnetic fields that can be time-dependent and can accommodate the presence of Heisenberg exchange interactions between the spins. This is done by deriving an effective magnetic field from the quantum partition function of the system to use in classical atomistic spin dynamics simulations and generalises the formalism presented in our previous work [Phys. Rev. Research 5, 043075 (2023)]. In special cases where the effective field can be computed exactly we compare our results with exact/numerical diagonalisation methods for both ferromagnetic and antiferromagnetic coupling. We show that our method works well across a large temperature range and can reproduce quantum expectation values for antiferromagnetic coupling which is usually not possible with classical models. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 13:09:10 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 10:49:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nussle",
"Thomas",
""
],
[
"Nicolis",
"Stam",
""
],
[
"Sofos",
"Iason",
""
],
[
"Barker",
"Joseph",
""
]
] |
2502.19834 | Guanzhou Ke | Guanzhou Ke, Shengfeng He, Xiao Li Wang, Bo Wang, Guoqing Chao, Yuanyang Zhang, Yi Xie, and HeXing Su | Knowledge Bridger: Towards Training-free Missing Modality Completion | Accepted to CVPR 2025 | null | null | null | cs.LG cs.CV cs.MM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Previous successful approaches to missing modality completion rely on carefully designed fusion techniques and extensive pre-training on complete data, which can limit their generalizability in out-of-domain (OOD) scenarios. In this study, we pose a new challenge: can we develop a missing modality completion model that is both resource-efficient and robust to OOD generalization? To address this, we present a training-free framework for missing modality completion that leverages large multimodal models (LMMs). Our approach, termed the "Knowledge Bridger", is modality-agnostic and integrates generation and ranking of missing modalities. By defining domain-specific priors, our method automatically extracts structured information from available modalities to construct knowledge graphs. These extracted graphs connect the missing modality generation and ranking modules through the LMM, resulting in high-quality imputations of missing modalities. Experimental results across both general and medical domains show that our approach consistently outperforms competing methods, including in OOD generalization. Additionally, our knowledge-driven generation and ranking techniques demonstrate superiority over variants that directly employ LMMs for generation and ranking, offering insights that may be valuable for applications in other domains. | [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 07:14:11 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 04:45:52 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 01:45:10 GMT"
},
{
"version": "v4",
"created": "Fri, 28 Mar 2025 13:38:14 GMT"
},
{
"version": "v5",
"created": "Tue, 17 Jun 2025 11:56:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ke",
"Guanzhou",
""
],
[
"He",
"Shengfeng",
""
],
[
"Wang",
"Xiao Li",
""
],
[
"Wang",
"Bo",
""
],
[
"Chao",
"Guoqing",
""
],
[
"Zhang",
"Yuanyang",
""
],
[
"Xie",
"Yi",
""
],
[
"Su",
"HeXing",
""
]
] |
2502.20510 | Paul B. Larson | Paul B. Larson, Grigor Sargsyan, Trevor Wilson | A model of the Axiom of Determinacy in which every set of reals is universally Baire | null | Forum of Mathematics, Sigma 13 (2025) e94 | 10.1017/fms.2025.10053 | null | math.LO | http://creativecommons.org/licenses/by/4.0/ | The consistency of the theory $\mathsf{ZF} + \mathsf{AD}_{\mathbb{R}} + {}$``every set of reals is universally Baire'' is proved relative to $\mathsf{ZFC} + {}$``there is a cardinal that is a limit of Woodin cardinals and of strong cardinals.'' The proof is based on the derived model construction, which was used by Woodin to show that the theory $\mathsf{ZF} + \mathsf{AD}_{\mathbb{R}} + {}$``every set of reals is Suslin'' is consistent relative to $\mathsf{ZFC} + {}$``there is a cardinal $\lambda$ that is a limit of Woodin cardinals and of $\mathord{<}\lambda$-strong cardinals.'' The $\Sigma^2_1$ reflection property of our model is proved using genericity iterations as used by Neeman and Steel. | [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 20:40:18 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 22:08:54 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Mar 2025 15:24:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Larson",
"Paul B.",
""
],
[
"Sargsyan",
"Grigor",
""
],
[
"Wilson",
"Trevor",
""
]
] |
2502.20620 | Ayana Niwa | Ayana Niwa and Masahiro Kaneko and Kentaro Inui | Rectifying Belief Space via Unlearning to Harness LLMs' Reasoning | Accepted at ACL2025 Findings (long) | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) can exhibit advanced reasoning yet still generate incorrect answers. We hypothesize that such errors frequently stem from spurious beliefs, propositions the model internally considers true but are incorrect. To address this, we propose a method to rectify the belief space by suppressing these spurious beliefs while simultaneously enhancing true ones, thereby enabling more reliable inferences. Our approach first identifies the beliefs that lead to incorrect or correct answers by prompting the model to generate textual explanations, using our Forward-Backward Beam Search (FBBS). We then apply unlearning to suppress the identified spurious beliefs and enhance the true ones, effectively rectifying the model's belief space. Empirical results on multiple QA datasets and LLMs show that our method corrects previously misanswered questions without harming overall model performance. Furthermore, our approach yields improved generalization on unseen data, suggesting that rectifying a model's belief space is a promising direction for mitigating errors and enhancing overall reliability. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 00:57:45 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:26:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Niwa",
"Ayana",
""
],
[
"Kaneko",
"Masahiro",
""
],
[
"Inui",
"Kentaro",
""
]
] |
2502.21094 | Max van Haren | Max van Haren, Roy S. Smith, and Tom Oomen | System Identification Beyond the Nyquist Frequency: A Kernel-Regularized Approach | null | Control Engineering Practice 164:106425 (2025) | 10.1016/j.conengprac.2025.106425 | null | eess.SY cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Models that contain intersample behavior are important for control design of systems with slow-rate outputs. The aim of this paper is to develop a system identification technique for fast-rate models of systems where only slow-rate output measurements are available, e.g., vision-in-the-loop systems. In this paper, the intersample response is estimated by identifying fast-rate models through least-squares criteria, and the limitations of these models are determined. In addition, a method is developed that surpasses these limitations and is capable of estimating unique fast-rate models of arbitrary order by regularizing the least-squares estimate. The developed method utilizes fast-rate inputs and slow-rate outputs and identifies fast-rate models accurately in a single identification experiment. Finally, both simulation and experimental validation on a prototype wafer stage demonstrate the effectiveness of the framework. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 14:32:46 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 07:29:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"van Haren",
"Max",
""
],
[
"Smith",
"Roy S.",
""
],
[
"Oomen",
"Tom",
""
]
] |
2503.01050 | Weilun Li | Weilun Li, Qimu Yuan, Yinan Chen, Joshua R. S. Lilly, Marina R. Filip, Laura M. Herz, Michael B. Johnston, Joanne Etheridge | Ruddlesden-Popper defects act as a free surface: role in formation and photophysical properties of CsPbI3 | null | null | 10.1002/adma.202501788 | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The perovskite semiconductor, CsPbI3, holds excellent promise for solar cell applications due to its suitable bandgap. However, achieving phase-stable CsPbI3 solar cells with high power conversion efficiency remains a major challenge. Ruddlesden-Popper (RP) defects have been identified in a range of perovskite semiconductors, including CsPbI3. However, there is limited understanding as to why they form or their impact on stability and photophysical properties. Here we increase the prevalence of RP defects with increased Cs-excess in vapour-deposited CsPbI3 thin films and observe superior structural stability but inferior photophysical properties. Significantly, using electron microscopy, we find that the atomic positions at the planar defect are comparable to those of a free surface, revealing their role in phase stabilisation. Density functional theory (DFT) calculations reveal the RP planes are electronically benign, however, antisites observed at RP turning points are likely to be malign. We therefore propose that increasing RP planes while reducing RP turning points could offer a breakthrough for improving both phase stability and photophysical performance. The formation mechanism revealed here may well apply more generally to RP structures in other perovskite systems. | [
{
"version": "v1",
"created": "Sun, 02 Mar 2025 22:44:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Weilun",
""
],
[
"Yuan",
"Qimu",
""
],
[
"Chen",
"Yinan",
""
],
[
"Lilly",
"Joshua R. S.",
""
],
[
"Filip",
"Marina R.",
""
],
[
"Herz",
"Laura M.",
""
],
[
"Johnston",
"Michael B.",
""
],
[
"Etheridge",
"Joanne",
""
]
] |
2503.01120 | Shubham Kejriwal | Shubham Kejriwal, Francisco Duque, Alvin J. K. Chua, Jonathan Gair | Bias-Corrected Importance Sampling for Inferring Beyond-Vacuum-GR Effects in Gravitational-Wave Sources | (Before-proofs-accepted) 13 pages, 6 figures | null | null | null | gr-qc | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The upcoming gravitational wave (GW) observatory LISA will measure the parameters of sources like extreme-mass-ratio inspirals (EMRIs) to exquisite precision. These measurements will also be sensitive to perturbations to the vacuum, GR-consistent evolution of sources, which might be caused by astrophysical environments or deviations from general relativity (GR). Previous studies have shown such ``beyond-vacuum-GR'' perturbations to potentially induce severe biases ($\gtrsim 10\sigma$) on recovered parameters under the ``null'' vacuum-GR hypothesis. While Bayesian inference can be performed under the null hypothesis using Markov Chain Monte Carlo (MCMC) samplers, it is computationally infeasible to repeat for more than a modest subset of all possible beyond-vacuum-GR hypotheses. We introduce bias-corrected importance sampling, a generic inference technique for nested models that is informed by the null hypothesis posteriors and the linear signal approximation to correct any induced inference biases. For a typical EMRI source that is significantly influenced by its environment but has been inferred only under the null hypothesis, the proposed method efficiently recovers the injected (unbiased) source parameters and the true posterior at a fraction of the expense of redoing MCMC inference under the full hypothesis. In future GW data analysis using the output of the proposed LISA global-fit pipeline, such methods may be necessary for the feasible and systematic inference of beyond-vacuum-GR effects. | [
{
"version": "v1",
"created": "Mon, 03 Mar 2025 02:56:39 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 17:24:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kejriwal",
"Shubham",
""
],
[
"Duque",
"Francisco",
""
],
[
"Chua",
"Alvin J. K.",
""
],
[
"Gair",
"Jonathan",
""
]
] |
2503.01590 | Lorenzo Bastonero | Lorenzo Bastonero, Cristiano Malica, Eric Macke, Marnik Bercx, Sebastian P. Huber, Iurii Timrov, Nicola Marzari | First-principles Hubbard parameters with automated and reproducible workflows | null | npj Comput Mater 11, 183 (2025) | 10.1038/s41524-025-01685-4 | null | cond-mat.mtrl-sci physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | We introduce an automated, flexible framework (aiida-hubbard) to self-consistently calculate Hubbard $U$ and $V$ parameters from first-principles. By leveraging density-functional perturbation theory, the computation of the Hubbard parameters is efficiently parallelized using multiple concurrent and inexpensive primitive cell calculations. Furthermore, the intersite $V$ parameters are defined on-the-fly during the iterative procedure to account for atomic relaxations and diverse coordination environments. We demonstrate the scalability and reliability of the framework by computing in high-throughput fashion the self-consistent onsite $U$ and intersite $V$ parameters for 115 Li-containing bulk solids. Our analysis of the Hubbard parameters calculated reveals a significant correlation of the onsite $U$ values on the oxidation state and coordination environment of the atom on which the Hubbard manifold is centered, while intersite $V$ values exhibit a general decay with increasing interatomic distance. We find, e.g., that the numerical values of $U$ for Fe and Mn 3d orbitals can vary up to 3 eV and 6 eV, respectively; their distribution is characterized by typical shifts of about 0.5 eV and 1.0 eV upon change in oxidation state, or local coordination environment. For the intersite $V$ a narrower spread is found, with values ranging between 0.2 eV and 1.6 eV when considering transition metal and oxygen interactions. This framework paves the way for the exploration of redox materials chemistry and high-throughput screening of $d$ and $f$ compounds across diverse research areas, including the discovery and design of novel energy storage materials, as well as other technologically-relevant applications. | [
{
"version": "v1",
"created": "Mon, 03 Mar 2025 14:28:51 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bastonero",
"Lorenzo",
""
],
[
"Malica",
"Cristiano",
""
],
[
"Macke",
"Eric",
""
],
[
"Bercx",
"Marnik",
""
],
[
"Huber",
"Sebastian P.",
""
],
[
"Timrov",
"Iurii",
""
],
[
"Marzari",
"Nicola",
""
]
] |
2503.03536 | Michael Powers Ph.D. | Michael R. Powers, Jiaxin Xu | A Criterion for Extending Continuous-Mixture Identifiability Results | null | null | null | null | stat.ML math.PR stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mixture distributions provide a versatile and widely used framework for modeling random phenomena, and are particularly well-suited to the analysis of geoscientific processes and their attendant risks to society. For continuous mixtures of random variables, we specify a simple criterion - generating-function accessibility - to extend previously known kernel-based identifiability (or unidentifiability) results to new kernel distributions. This criterion, based on functional relationships between the relevant kernels' moment-generating functions or Laplace transforms, may be applied to continuous mixtures of both discrete and continuous random variables. To illustrate the proposed approach, we present results for several specific kernels, in each case briefly noting its relevance to research in the geosciences and/or related risk analysis. | [
{
"version": "v1",
"created": "Wed, 05 Mar 2025 14:19:29 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 18:39:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Powers",
"Michael R.",
""
],
[
"Xu",
"Jiaxin",
""
]
] |
2503.03880 | Jonathan F. MacArt | Priyesh Kakka, Jonathan F. MacArt | Neural network-augmented eddy viscosity closures for turbulent premixed jet flames | null | Combustion and Flame 278 (2025) 114241 | 10.1016/j.combustflame.2025.114241 | null | physics.flu-dyn physics.comp-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Extending gradient-type turbulence closures to turbulent premixed flames is challenging due to the significant influence of combustion heat release. We incorporate a deep neural network (DNN) into Reynolds-averaged Navier--Stokes (RANS) models for the turbulent viscosity and thermal conductivity as nonlinear functions of the local flow state and thermochemical gradients. Our models are optimized over the RANS partial differential equations (PDEs) using an adjoint-based data assimilation procedure. Because we directly target the RANS solution, as opposed to the unclosed terms, successfully trained models are guaranteed to improve the in-sample accuracy of the DNN-augmented RANS predictions. We demonstrate the learned closures for in- and out-of-sample $\textit{a posteriori}$ RANS predictions of compressible, premixed, turbulent jet flames with turbulent Damk\"ohler numbers spanning the gradient- and counter-gradient transport regimes. The DNN-augmented RANS predictions have one to two orders of magnitude lower spatiotemporal mean-squared error than those using a baseline $k$--$\epsilon$ model, even for Damk\"ohler numbers far from those used for training. This demonstrates the accuracy, stability, and generalizability of the PDE-constrained modeling approach for turbulent jet flames over this relatively wide Damk\"ohler number range. | [
{
"version": "v1",
"created": "Wed, 05 Mar 2025 20:25:35 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 18:27:27 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kakka",
"Priyesh",
""
],
[
"MacArt",
"Jonathan F.",
""
]
] |
2503.04429 | Michael Lan | Narmeen Oozeer, Dhruv Nathawani, Nirmalendu Prakash, Michael Lan, Abir Harrasse, Amirali Abdullah | Activation Space Interventions Can Be Transferred Between Large Language Models | 75 pages | null | null | null | cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | The study of representation universality in AI models reveals growing convergence across domains, modalities, and architectures. However, the practical applications of representation universality remain largely unexplored. We bridge this gap by demonstrating that safety interventions can be transferred between models through learned mappings of their shared activation spaces. We demonstrate this approach on two well-established AI safety tasks: backdoor removal and refusal of harmful prompts, showing successful transfer of steering vectors that alter the models' outputs in a predictable way. Additionally, we propose a new task, \textit{corrupted capabilities}, where models are fine-tuned to embed knowledge tied to a backdoor. This tests their ability to separate useful skills from backdoors, reflecting real-world challenges. Extensive experiments across Llama, Qwen and Gemma model families show that our method enables using smaller models to efficiently align larger ones. Furthermore, we demonstrate that autoencoder mappings between base and fine-tuned models can serve as reliable ``lightweight safety switches", allowing dynamic toggling between model behaviors. | [
{
"version": "v1",
"created": "Thu, 06 Mar 2025 13:38:44 GMT"
},
{
"version": "v2",
"created": "Mon, 05 May 2025 16:39:22 GMT"
},
{
"version": "v3",
"created": "Mon, 16 Jun 2025 18:43:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Oozeer",
"Narmeen",
""
],
[
"Nathawani",
"Dhruv",
""
],
[
"Prakash",
"Nirmalendu",
""
],
[
"Lan",
"Michael",
""
],
[
"Harrasse",
"Abir",
""
],
[
"Abdullah",
"Amirali",
""
]
] |
2503.04619 | Xin Zhang | Xin Zhang, Qiyu Wei, Yingjie Zhu, Linhai Zhang, Deyu Zhou, Sophia Ananiadou | SynGraph: A Dynamic Graph-LLM Synthesis Framework for Sparse Streaming User Sentiment Modeling | Accepted at ACL 2025 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | User reviews on e-commerce platforms exhibit dynamic sentiment patterns driven by temporal and contextual factors. Traditional sentiment analysis methods focus on static reviews, failing to capture the evolving temporal relationship between user sentiment rating and textual content. Sentiment analysis on streaming reviews addresses this limitation by modeling and predicting the temporal evolution of user sentiments. However, it suffers from data sparsity, manifesting in temporal, spatial, and combined forms. In this paper, we introduce SynGraph, a novel framework designed to address data sparsity in sentiment analysis on streaming reviews. SynGraph alleviates data sparsity by categorizing users into mid-tail, long-tail, and extreme scenarios and incorporating LLM-augmented enhancements within a dynamic graph-based structure. Experiments on real-world datasets demonstrate its effectiveness in addressing sparsity and improving sentiment modeling in streaming reviews. | [
{
"version": "v1",
"created": "Thu, 06 Mar 2025 17:05:33 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:22:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Xin",
""
],
[
"Wei",
"Qiyu",
""
],
[
"Zhu",
"Yingjie",
""
],
[
"Zhang",
"Linhai",
""
],
[
"Zhou",
"Deyu",
""
],
[
"Ananiadou",
"Sophia",
""
]
] |
2503.04804 | Arturs Kanepajs | Arturs Kanepajs, Aditi Basu, Sankalpa Ghose, Constance Li, Akshat Mehta, Ronak Mehta, Samuel David Tucker-Davis, Eric Zhou, Bob Fischer, Jacy Reese Anthis | What do Large Language Models Say About Animals? Investigating Risks of Animal Harm in Generated Text | null | null | null | null | cs.CY cs.CL | http://creativecommons.org/licenses/by/4.0/ | As machine learning systems become increasingly embedded in society, their impact on human and nonhuman life continues to escalate. Technical evaluations have addressed a variety of potential harms from large language models (LLMs) towards humans and the environment, but there is little empirical work regarding harms towards nonhuman animals. Following the growing recognition of animal protection in regulatory and ethical AI frameworks, we present AnimalHarmBench (AHB), a benchmark for risks of animal harm in LLM-generated text. Our benchmark dataset comprises 1,850 curated questions from Reddit post titles and 2,500 synthetic questions based on 50 animal categories (e.g., cats, reptiles) and 50 ethical scenarios with a 70-30 public-private split. Scenarios include open-ended questions about how to treat animals, practical scenarios with potential animal harm, and willingness-to-pay measures for the prevention of animal harm. Using the LLM-as-a-judge framework, responses are evaluated for their potential to increase or decrease harm, and evaluations are debiased for the tendency of judges to judge their own outputs more favorably. AHB reveals significant differences across frontier LLMs, animal categories, scenarios, and subreddits. We conclude with future directions for technical research and addressing the challenges of building evaluations on complex social and moral topics. | [
{
"version": "v1",
"created": "Mon, 03 Mar 2025 15:32:18 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 03:02:59 GMT"
},
{
"version": "v3",
"created": "Sun, 15 Jun 2025 16:09:26 GMT"
},
{
"version": "v4",
"created": "Tue, 17 Jun 2025 07:44:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kanepajs",
"Arturs",
""
],
[
"Basu",
"Aditi",
""
],
[
"Ghose",
"Sankalpa",
""
],
[
"Li",
"Constance",
""
],
[
"Mehta",
"Akshat",
""
],
[
"Mehta",
"Ronak",
""
],
[
"Tucker-Davis",
"Samuel David",
""
],
[
"Zhou",
"Eric",
""
],
[
"Fischer",
"Bob",
""
],
[
"Anthis",
"Jacy Reese",
""
]
] |
2503.05576 | Florent Mertens | F. G. Mertens, M. Mevius, L. V. E. Koopmans, A. R. Offringa, S. Zaroubi, A. Acharya, S. A. Brackenhoff, E. Ceccotti, E. Chapman, K. Chege, B. Ciardi, R. Ghara, S. Ghosh, S. K. Giri, I. Hothi, C. H\"ofer, I. T. Iliev, V. Jeli\'c, Q. Ma, G. Mellema, S. Munshi, V. N. Pandey, and S. Yatawatta | Deeper multi-redshift upper limits on the Epoch of Reionization 21-cm signal power spectrum from LOFAR between z=8.3 and z=10.1 | Accepted in A&A | A&A 698, A186 (2025) | 10.1051/0004-6361/202554158 | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We present new upper limits on the 21-cm signal power spectrum from the Epoch of Reionization (EoR), at redshifts $z \approx 10.1, 9.1, \text{ and } 8.3$, based on reprocessed observations from the Low-Frequency Array (LOFAR). The analysis incorporates significant enhancements in calibration methods, sky model subtraction, radio-frequency interference (RFI) mitigation, and an improved signal separation technique using machine learning to develop a physically motivated covariance model for the 21-cm signal. These advancements have markedly reduced previously observed excess power due to residual systematics, bringing the measurements closer to the theoretical thermal noise limit across the entire $k$-space. Using comparable observational data, we achieve a 2 to 4-fold improvement over our previous LOFAR limits, with best upper limits of $\Delta_{21}^2 < (68.7\,\mathrm{mK})^2$ at $k = 0.076\,h\,\mathrm{cMpc}^{-1}$, $\Delta_{21}^2 < (54.3\,\mathrm{mK})^2$ at $k = 0.076\,h\,\mathrm{cMpc}^{-1}$ and $\Delta_{21}^2 < (65.5\,\mathrm{mK})^2$ at $k = 0.083\,h\,\mathrm{cMpc}^{-1}$ at redshifts $z \approx 10.1, 9.1$, and $8.3$, respectively. These new multi-redshift upper limits provide new constraints that can be used to refine our understanding of the astrophysical processes during the EoR. Comprehensive validation tests, including signal injection, were performed to ensure the robustness of our methods. The remaining excess power is attributed to residual foreground emissions from distant sources, beam model inaccuracies, and low-level RFI. We discuss ongoing and future improvements to the data processing pipeline aimed at further reducing these residuals, thereby enhancing the sensitivity of LOFAR observations in the quest to detect the 21-cm signal from the EoR. | [
{
"version": "v1",
"created": "Fri, 07 Mar 2025 16:58:45 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 13:17:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mertens",
"F. G.",
""
],
[
"Mevius",
"M.",
""
],
[
"Koopmans",
"L. V. E.",
""
],
[
"Offringa",
"A. R.",
""
],
[
"Zaroubi",
"S.",
""
],
[
"Acharya",
"A.",
""
],
[
"Brackenhoff",
"S. A.",
""
],
[
"Ceccotti",
"E.",
""
],
[
"Chapman",
"E.",
""
],
[
"Chege",
"K.",
""
],
[
"Ciardi",
"B.",
""
],
[
"Ghara",
"R.",
""
],
[
"Ghosh",
"S.",
""
],
[
"Giri",
"S. K.",
""
],
[
"Hothi",
"I.",
""
],
[
"Höfer",
"C.",
""
],
[
"Iliev",
"I. T.",
""
],
[
"Jelić",
"V.",
""
],
[
"Ma",
"Q.",
""
],
[
"Mellema",
"G.",
""
],
[
"Munshi",
"S.",
""
],
[
"Pandey",
"V. N.",
""
],
[
"Yatawatta",
"S.",
""
]
] |
2503.06294 | Dibakar Roychowdhury | Dibakar Roychowdhury | Field theory aspects of $ \eta $-deformed superstring background | v2, accepted in JHEP | JHEP 06 (2025) 161 | 10.1007/JHEP06(2025)161 | null | hep-th gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We explore various field theory aspects of integrable $ \eta $-deformed geometry in type IIB supergravity by employing several holographic probes. These include the computation of holographic timelike entanglement entropy and estimation of various other field theory observables for example, the flow central charge and the quantum complexity. We also discuss the associated brane set up and compute Page charges. We further use them to calculate the coupling constant in the dual QFTs considering both small and large deformation limits. | [
{
"version": "v1",
"created": "Sat, 08 Mar 2025 18:07:16 GMT"
},
{
"version": "v2",
"created": "Sun, 11 May 2025 05:08:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Roychowdhury",
"Dibakar",
""
]
] |
2503.06315 | Yakov Shnir | Luiz A. Ferreira, Aliaksei Mikhaliuk and Yakov Shnir | $U(1)$ gauged non-topological solitons in the 3+1 dimensional $O(3)$ sigma-model | 18 pages, 6 figures, discussions and references have been added, accepted by Phys Rev D | null | null | null | hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present and study new non-topological soliton solutions in the $U(1)$ gauged non-linear $O(3)$ sigma model with a symmetry breaking potential in 3+1 dimensional flat space-time. The configurations are endowed with an electric and magnetic field and also carry a nonvanishing angular momentum density. We discuss properties of these solitons and investigate the domains of their existence. | [
{
"version": "v1",
"created": "Sat, 08 Mar 2025 19:19:25 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 19:02:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ferreira",
"Luiz A.",
""
],
[
"Mikhaliuk",
"Aliaksei",
""
],
[
"Shnir",
"Yakov",
""
]
] |
2503.06377 | Akihiro Munemasa | Yen-chi Roger Lin, Akihiro Munemasa, Tetsuji Taniguchi, and Kiyoto Yoshino | Sets of equiangular lines in dimension $18$ constructed from $A_9 \oplus A_9 \oplus A_1$ | 17 pages, corrected typo | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | In 2023, Greaves et~al.\ constructed several sets of 57 equiangular lines in dimension 18. Using the concept of switching root introduced by Cao et~al.\ in 2021, these sets of equiangular lines are embedded in a lattice of rank 19 spanned by norm 3 vectors together with a switching root. We characterize this lattice as an overlattice of the root lattice $A_9\oplus A_9\oplus A_1$, and show that there are at least $246896$ sets of 57 equiangular lines in dimension $18$ arising in this way, up to isometry. Additionally, we prove that all of these sets of equiangular lines are strongly maximal. Here, a set of equiangular lines is said to be strongly maximal if there is no set of equiangular lines properly containing it even if the dimension of the underlying space is increased. Among these sets, there are ones with only six distinct Seidel eigenvalues. | [
{
"version": "v1",
"created": "Sun, 09 Mar 2025 01:06:08 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:51:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lin",
"Yen-chi Roger",
""
],
[
"Munemasa",
"Akihiro",
""
],
[
"Taniguchi",
"Tetsuji",
""
],
[
"Yoshino",
"Kiyoto",
""
]
] |
2503.06404 | Jieci Wang | Xiaofang Liu, Wentao Liu, Zhilong Liu, Jieci Wang | Harvesting correlations from BTZ black hole coupled to a Lorentz-violating vector field | 14 pages, 31 figures | null | null | null | gr-qc quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we investigate the effects of Lorentz violation on correlations harvesting, specifically focusing on the harvested entanglement and harvested mutual information between two Unruh-DeWitt detectors interacting with a quantum field in the Lorentz-violating BTZ-like black hole spacetime. Our findings reveal that Lorentz symmetry breaking has contrasting impacts on entanglement harvesting and mutual information harvesting in BTZ backgrounds: it enhances mutual information harvesting while suppressing entanglement harvesting. This phenomenon suggests that the increase in total correlations in Lorentz-violating vector field backgrounds with gravitational coupling is predominantly driven by classical components, with quantum correlations contributing less to the overall mutual information. These results indicate that Lorentz violation, as a quantum property of spacetime, may impose intrinsic constraints on the quantum information capacity encoded in spacetime due to competition among quantum degrees of freedom for resources. Furthermore, Lorentz symmetry breaking expands the \textit{entanglement shadow} region, further demonstrating its disruptive effect on quantum correlations. | [
{
"version": "v1",
"created": "Sun, 09 Mar 2025 02:50:15 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 11:44:38 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 01:18:18 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Xiaofang",
""
],
[
"Liu",
"Wentao",
""
],
[
"Liu",
"Zhilong",
""
],
[
"Wang",
"Jieci",
""
]
] |
2503.06432 | Xiao Yu Chen | Xiaoyu Chen, HongSheng Hu | The boundness of Lusztig's $a$-function for Coxeter groups of finite rank | 44 pages, Rewrite subsection 3.1, including indeas, intuitions, and toy examples | null | null | null | math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove that the Lusztig's $a$-function is bounded for any Coxeter group of finite rank. | [
{
"version": "v1",
"created": "Sun, 09 Mar 2025 04:09:17 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 01:44:46 GMT"
},
{
"version": "v3",
"created": "Thu, 13 Mar 2025 15:24:39 GMT"
},
{
"version": "v4",
"created": "Wed, 19 Mar 2025 03:54:26 GMT"
},
{
"version": "v5",
"created": "Sat, 22 Mar 2025 14:47:36 GMT"
},
{
"version": "v6",
"created": "Sat, 19 Apr 2025 06:41:10 GMT"
},
{
"version": "v7",
"created": "Wed, 07 May 2025 14:16:20 GMT"
},
{
"version": "v8",
"created": "Tue, 17 Jun 2025 11:59:18 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chen",
"Xiaoyu",
""
],
[
"Hu",
"HongSheng",
""
]
] |
2503.06926 | Jeffrey Willard | Yuchen Han, Yucheng Wu, Jeffrey Willard | Effect of Selection Format on LLM Performance | null | null | null | null | cs.CL cs.AI cs.CE cs.ET cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper investigates a critical aspect of large language model (LLM) performance: the optimal formatting of classification task options in prompts. Through an extensive experimental study, we compared two selection formats -- bullet points and plain English -- to determine their impact on model performance. Our findings suggest that presenting options via bullet points generally yields better results, although there are some exceptions. Furthermore, our research highlights the need for continued exploration of option formatting to drive further improvements in model performance. | [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 05:11:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:41:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Han",
"Yuchen",
""
],
[
"Wu",
"Yucheng",
""
],
[
"Willard",
"Jeffrey",
""
]
] |
2503.07631 | Viet Lai | Thuy Nguyen, Dang Nguyen, Hoang Nguyen, Thuan Luong, Long Hoang Dang, Viet Dac Lai | OWLViz: An Open-World Benchmark for Visual Question Answering | ICML 2025 Workshop on Multi-Agent Systems in the Era of Foundation Models: Opportunities, Challenges, and Futures. (8 pages + appendix) | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a challenging benchmark for the Open WorLd VISual question answering (OWLViz) task. OWLViz presents concise, unambiguous queries that require integrating multiple capabilities, including visual understanding, web exploration, and specialized tool usage. While humans achieve 69.2% accuracy on these intuitive tasks, even state-of-the-art VLMs struggle, with the best model, Gemini 2.0, achieving only 26.6% accuracy. Current agentic VLMs, which rely on limited vision and vision-language models as tools, perform even worse. This performance gap reveals significant limitations in multimodal systems' ability to select appropriate tools and execute complex reasoning sequences, establishing new directions for advancing practical AI research. | [
{
"version": "v1",
"created": "Tue, 04 Mar 2025 19:37:33 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 03:34:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nguyen",
"Thuy",
""
],
[
"Nguyen",
"Dang",
""
],
[
"Nguyen",
"Hoang",
""
],
[
"Luong",
"Thuan",
""
],
[
"Dang",
"Long Hoang",
""
],
[
"Lai",
"Viet Dac",
""
]
] |
2503.07866 | Scott Willenbrock | Scott Willenbrock | A Brief History of Mass | 16 pages, 2 figures. Minor additions | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | It has been known since the 1950's that an unstable particle is associated with a complex pole in the propagator. This had to be rediscovered twice: in the early 1970's in the context of hadronic resonances, and in the early 1990's in the context of the $Z$ boson. The physical mass of the particle is the real part of the pole in the complex energy plane. In hadronic physics, this replaced the ``Breit-Wigner mass,'' which was found to depend on the parameterization of the ``energy-dependent width.'' In $Z$ physics, it replaced the ``on-shell'' mass, which was found to be gauge dependent. Although the mass defined from the complex pole position has been widely discussed in the literature, it has not yet made its way into quantum field theory textbooks. | [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 21:27:27 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 20:18:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Willenbrock",
"Scott",
""
]
] |
2503.08669 | Zekun Li | Zekun Li, Shinda Huang, Jiangtian Wang, Nathan Zhang, Antonis Antoniades, Wenyue Hua, Kaijie Zhu, Sirui Zeng, Chi Wang, William Yang Wang, Xifeng Yan | SOPBench: Evaluating Language Agents at Following Standard Operating Procedures and Constraints | Code, data, and over 24k agent trajectories are released at https://github.com/Leezekun/SOPBench | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | As language agents increasingly automate critical tasks, their ability to follow domain-specific standard operating procedures (SOPs), policies, and constraints when taking actions and making tool calls becomes essential yet remains underexplored. To address this gap, we develop an automated evaluation pipeline SOPBench with: (1) executable environments containing 167 tools/functions across seven customer service domains with service-specific SOPs and rule-based verifiers, (2) an automated test generation framework producing over 900 verified test cases, and (3) an automated evaluation framework to rigorously assess agent adherence from multiple dimensions. Our approach transforms each service-specific SOP code program into a directed graph of executable functions and requires agents to call these functions based on natural language SOP descriptions. The original code serves as oracle rule-based verifiers to assess compliance, reducing reliance on manual annotations and LLM-based evaluations. We evaluate 18 leading models, and results show the task is challenging even for top-tier models (like GPT-4o, Claude-3.7-Sonnet), with variances across domains. Reasoning models like o4-mini-high show superiority while other powerful models perform less effectively (pass rates of 30%-50%), and small models (7B, 8B) perform significantly worse. Additionally, language agents can be easily jailbroken to overlook SOPs and constraints. Code, data, and over 24k agent trajectories are released at https://github.com/Leezekun/SOPBench. | [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 17:53:02 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 17:50:44 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Zekun",
""
],
[
"Huang",
"Shinda",
""
],
[
"Wang",
"Jiangtian",
""
],
[
"Zhang",
"Nathan",
""
],
[
"Antoniades",
"Antonis",
""
],
[
"Hua",
"Wenyue",
""
],
[
"Zhu",
"Kaijie",
""
],
[
"Zeng",
"Sirui",
""
],
[
"Wang",
"Chi",
""
],
[
"Wang",
"William Yang",
""
],
[
"Yan",
"Xifeng",
""
]
] |
2503.08679 | Iv\'an Arcuschin | Iv\'an Arcuschin, Jett Janiak, Robert Krzyzanowski, Senthooran Rajamanoharan, Neel Nanda, Arthur Conmy | Chain-of-Thought Reasoning In The Wild Is Not Always Faithful | Accepted to the Reasoning and Planning for LLMs Workshop (ICLR 25), 10 main paper pages, 39 appendix pages | null | null | null | cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Chain-of-Thought (CoT) reasoning has significantly advanced state-of-the-art AI capabilities. However, recent studies have shown that CoT reasoning is not always faithful when models face an explicit bias in their prompts, i.e., the CoT can give an incorrect picture of how models arrive at conclusions. We go further and show that unfaithful CoT can also occur on realistic prompts with no artificial bias. We find that when separately presented with the questions "Is X bigger than Y?" and "Is Y bigger than X?", models sometimes produce superficially coherent arguments to justify systematically answering Yes to both questions or No to both questions, despite such responses being logically contradictory. We show preliminary evidence that this is due to models' implicit biases towards Yes or No, thus labeling this unfaithfulness as Implicit Post-Hoc Rationalization. Our results reveal that several production models exhibit surprisingly high rates of post-hoc rationalization in our settings: GPT-4o-mini (13%) and Haiku 3.5 (7%). While frontier models are more faithful, especially thinking ones, none are entirely faithful: Gemini 2.5 Flash (2.17%), ChatGPT-4o (0.49%), DeepSeek R1 (0.37%), Gemini 2.5 Pro (0.14%), and Sonnet 3.7 with thinking (0.04%). We also investigate Unfaithful Illogical Shortcuts, where models use subtly illogical reasoning to try to make a speculative answer to hard maths problems seem rigorously proven. Our findings raise challenges for strategies for detecting undesired behavior in LLMs via the chain of thought. | [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 17:56:30 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 17:49:58 GMT"
},
{
"version": "v3",
"created": "Wed, 19 Mar 2025 19:20:42 GMT"
},
{
"version": "v4",
"created": "Tue, 17 Jun 2025 17:59:57 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Arcuschin",
"Iván",
""
],
[
"Janiak",
"Jett",
""
],
[
"Krzyzanowski",
"Robert",
""
],
[
"Rajamanoharan",
"Senthooran",
""
],
[
"Nanda",
"Neel",
""
],
[
"Conmy",
"Arthur",
""
]
] |
2503.08703 | Yimeng Shan | Yimeng Shan, Zhenbang Ren, Haodi Wu, Wenjie Wei, Rui-Jie Zhu, Shuai Wang, Dehao Zhang, Yichen Xiao, Jieyuan Zhang, Kexin Shi, Jingzhinan Wang, Jason K. Eshraghian, Haicheng Qu, Jiqing Zhang, Malu Zhang, Yang Yang | SDTrack: A Baseline for Event-based Tracking via Spiking Neural Networks | 11 pages,7 figures,4 tables | null | null | null | cs.NE cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Event cameras provide superior temporal resolution, dynamic range, power efficiency, and pixel bandwidth. Spiking Neural Networks (SNNs) naturally complement event data through discrete spike signals, making them ideal for event-based tracking. However, current approaches that combine Artificial Neural Networks (ANNs) and SNNs, along with suboptimal architectures, compromise energy efficiency and limit tracking performance. To address these limitations, we propose the first Transformer-based spike-driven tracking pipeline. Our Global Trajectory Prompt (GTP) method effectively captures global trajectory information and aggregates it with event streams into event images to enhance spatiotemporal representation. We then introduce SDTrack, a Transformer-based spike-driven tracker comprising a Spiking MetaFormer backbone and a simple tracking head that directly predicts normalized coordinates using spike signals. The framework is end-to-end, does not require data augmentation or post-processing. Extensive experiments demonstrate that SDTrack achieves state-of-the-art performance while maintaining the lowest parameter count and energy consumption across multiple event-based tracking benchmarks, establishing a solid baseline for future research in the field of neuromorphic vision. | [
{
"version": "v1",
"created": "Sun, 09 Mar 2025 02:01:40 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:08:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shan",
"Yimeng",
""
],
[
"Ren",
"Zhenbang",
""
],
[
"Wu",
"Haodi",
""
],
[
"Wei",
"Wenjie",
""
],
[
"Zhu",
"Rui-Jie",
""
],
[
"Wang",
"Shuai",
""
],
[
"Zhang",
"Dehao",
""
],
[
"Xiao",
"Yichen",
""
],
[
"Zhang",
"Jieyuan",
""
],
[
"Shi",
"Kexin",
""
],
[
"Wang",
"Jingzhinan",
""
],
[
"Eshraghian",
"Jason K.",
""
],
[
"Qu",
"Haicheng",
""
],
[
"Zhang",
"Jiqing",
""
],
[
"Zhang",
"Malu",
""
],
[
"Yang",
"Yang",
""
]
] |
2503.09354 | Christoph Huber | Christoph Huber, Dino Knoll and Michael Guthe | Fully-Synthetic Training for Visual Quality Inspection in Automotive Production | Accepted for publication in Procedia CIRP | Procedia CIRP 134 (2025) 777-782 | 10.1016/j.procir.2025.02.205 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Visual Quality Inspection plays a crucial role in modern manufacturing environments as it ensures customer safety and satisfaction. The introduction of Computer Vision (CV) has revolutionized visual quality inspection by improving the accuracy and efficiency of defect detection. However, traditional CV models heavily rely on extensive datasets for training, which can be costly, time-consuming, and error-prone. To overcome these challenges, synthetic images have emerged as a promising alternative. They offer a cost-effective solution with automatically generated labels. In this paper, we propose a pipeline for generating synthetic images using domain randomization. We evaluate our approach in three real inspection scenarios and demonstrate that an object detection model trained solely on synthetic data can outperform models trained on real images. | [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:58:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Huber",
"Christoph",
""
],
[
"Knoll",
"Dino",
""
],
[
"Guthe",
"Michael",
""
]
] |
2503.09483 | Andreas Kofler | Andreas Kofler, Luca Calatroni, Christoph Kolbitsch, Kostas Papafitsoros | Learning Spatially Adaptive $\ell_1$-Norms Weights for Convolutional Synthesis Regularization | Accepted for publication in the proceedings of the EUSIPCO 2025 conference | null | null | null | cs.LG cs.CV math.OC | http://creativecommons.org/licenses/by/4.0/ | We propose an unrolled algorithm approach for learning spatially adaptive parameter maps in the framework of convolutional synthesis-based $\ell_1$ regularization. More precisely, we consider a family of pre-trained convolutional filters and estimate deeply parametrized spatially varying parameters applied to the sparse feature maps by means of unrolling a FISTA algorithm to solve the underlying sparse estimation problem. The proposed approach is evaluated for image reconstruction of low-field MRI and compared to spatially adaptive and non-adaptive analysis-type procedures relying on Total Variation regularization and to a well-established model-based deep learning approach. We show that the proposed approach produces visually and quantitatively comparable results with the latter approaches and at the same time remains highly interpretable. In particular, the inferred parameter maps quantify
the local contribution of each filter in the reconstruction, which provides valuable insight into the algorithm mechanism and could potentially be used to discard unsuited filters. | [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 15:38:11 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Mar 2025 10:38:39 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 08:50:23 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kofler",
"Andreas",
""
],
[
"Calatroni",
"Luca",
""
],
[
"Kolbitsch",
"Christoph",
""
],
[
"Papafitsoros",
"Kostas",
""
]
] |
2503.09991 | Qi-Yue Yu | Qi-yue Yu, Shi-wen Lin, and Ting-wei Yang | Finite Field Multiple Access II:from Symbol-wise to Codeword-wise | 50 pages, 9 figures | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A finite-field multiple-access (FFMA) system separates users within a finite field by utilizing different element-pairs (EPs) as virtual resources. The Cartesian product of distinct EPs forms an EP code, which serves as the input to a finite-field multiplexing module (FF-MUX). This allows the FFMA technique to reorder the channel coding and multiplexing modules, enabling the superimposed signals to function as codewords that can be decoded by a channel code. This flexibility allows the FFMA system to efficiently support a large number of users with short packet traffic, addressing the finite blocklength (FBL) challenge in multiuser reliable transmission. Designing EP codes is a central challenge in FFMA systems. In this paper, we construct EP codes based on a bit(s)-to-codeword transformation approach and define the corresponding EP code as a codeword-wise EP (CWEP) code. We then investigate the encoding process of EP codes, and propose unique sum-pattern mapping (USPM) structural property constraints to design uniquely decodable CWEP codes. Next, we present the $\kappa$-fold ternary orthogonal matrix ${\bf T}_{\rm o}(2^{\kappa}, 2^{\kappa})$ over GF$(3^m)$, where $m = 2^{\kappa}$, and the ternary non-orthogonal matrix ${\bf T}_{\rm no}(M,m)$ over GF$(3^m)$, for constructing specific CWEP codes. Based on the proposed CWEP codes, we introduce three FFMA modes: channel codeword multiple access (FF-CCMA), code division multiple access (FF-CDMA), and non-orthogonal multiple access (FF-NOMA). Simulation results demonstrate that all three modes effectively support massive user transmissions with well-behaved error performance. | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 03:00:50 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:07:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yu",
"Qi-yue",
""
],
[
"Lin",
"Shi-wen",
""
],
[
"Yang",
"Ting-wei",
""
]
] |
2503.10069 | Xiangyu Shi | Xiangyu Shi, Zerui Li, Wenqi Lyu, Jiatong Xia, Feras Dayoub, Yanyuan Qiao, Qi Wu | SmartWay: Enhanced Waypoint Prediction and Backtracking for Zero-Shot Vision-and-Language Navigation | Accepted by IROS 2025. Project website: https://sxyxs.github.io/smartway/ | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision-and-Language Navigation (VLN) in continuous environments requires agents to interpret natural language instructions while navigating unconstrained 3D spaces. Existing VLN-CE frameworks rely on a two-stage approach: a waypoint predictor to generate waypoints and a navigator to execute movements. However, current waypoint predictors struggle with spatial awareness, while navigators lack historical reasoning and backtracking capabilities, limiting adaptability. We propose a zero-shot VLN-CE framework integrating an enhanced waypoint predictor with a Multi-modal Large Language Model (MLLM)-based navigator. Our predictor employs a stronger vision encoder, masked cross-attention fusion, and an occupancy-aware loss for better waypoint quality. The navigator incorporates history-aware reasoning and adaptive path planning with backtracking, improving robustness. Experiments on R2R-CE and MP3D benchmarks show our method achieves state-of-the-art (SOTA) performance in zero-shot settings, demonstrating competitive results compared to fully supervised methods. Real-world validation on Turtlebot 4 further highlights its adaptability. | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 05:32:57 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 05:47:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shi",
"Xiangyu",
""
],
[
"Li",
"Zerui",
""
],
[
"Lyu",
"Wenqi",
""
],
[
"Xia",
"Jiatong",
""
],
[
"Dayoub",
"Feras",
""
],
[
"Qiao",
"Yanyuan",
""
],
[
"Wu",
"Qi",
""
]
] |
2503.10309 | Elspeth Lee Dr | Elspeth K.H. Lee | Beyond monoculture: polydisperse moment methods for sub-stellar atmosphere cloud microphysics I. Examining properties of the exponential distribution | V1 - Submitted to A&A (13 March 2025). V2 - typos and corrections. V3 - accepted version (7 May 2025) | A&A 698, A220 (2025) | 10.1051/0004-6361/202554511 | null | astro-ph.EP astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Observational data provided by JWST instruments continue to challenge theories and models of cloud formation in sub-stellar atmospheres, requiring more sophisticated approaches in an effort to understand their spatial complexity. However, to date, most cloud microphysical models using the moment method for sub-stellar atmospheres have assumed a monodisperse size distribution, neglecting polydisperse properties. We aim to extend beyond the common assumption of a monodisperse size distribution and analyse cloud microphysical processes assuming an exponential distribution. We derive expressions for the zeroth and first moments of condensation/evaporation and collisional growth processes under the assumption of an exponential size distribution. We then compare the differences between monodisperse and exponential distribution microphysics using a simple one-dimensional (1D) column model applied to a Y-dwarf KCl cloud scenario. We find that adopting an exponential distribution modifies condensation/evaporation rates by a factor of $\approx$0.9 and collisional growth rates by factors of $>$1.1 (Kn $\ll$ 1) and $\approx$1.37 (Kn $\gg$ 1) for Brownian coagulation and $\approx$0.85 for gravitational coalescence, compared to the monodisperse case. In our specific test cases, we find maximal relative differences of $>$200\% in total number density and $>$40\% in mean radius of the cloud particles between the monodisperse and exponential distributions. Our framework offer a simple way to take into account polydispersity with an assumed exponential size distribution for sub-stellar atmospheric cloud microphysics using a two-moment method. In follow up studies, we will examine more complex distributions, such as the log-normal and gamma distributions, that require more than two moments to characterise self-consistently. | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 12:39:01 GMT"
},
{
"version": "v2",
"created": "Sun, 30 Mar 2025 14:15:36 GMT"
},
{
"version": "v3",
"created": "Wed, 07 May 2025 08:04:58 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lee",
"Elspeth K. H.",
""
]
] |
2503.10381 | Qian Xiao | Qian Xiao | On Dirichlet non-improvable numbers and shrinking target problems | Accepted by Ergodic Theory and Dynamical Systems | null | 10.1017/etds.2025.10139 | null | math.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In one-dimensional Diophantine approximation, the Diophantine properties of a real number are characterized by its partial quotients, especially the growth of its large partial quotients. Notably, Kleinbock and Wadleigh [Proc. Amer. Math. Soc. 2018] made a seminal contribution by linking the improvability of Dirichlet's theorem to the growth of the product of consecutive partial quotients. In this paper, we extend the concept of Dirichlet non-improvable sets within the framework of shrinking target problems. Specifically, consider the dynamical system $([0,1), T)$ of continued fractions. Let $\{z_n\}_{n \ge 1}$ be a sequence of real numbers in $[0,1]$ and let $B > 1$. We determine the Hausdorff dimension of the following set:
\[
\begin{split}
\{x\in[0,1):|T^nx-z_n||T^{n+1}x-Tz_n|<B^{-n}\text{ infinitely often}\}.
\end{split}
\] | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 13:57:03 GMT"
},
{
"version": "v2",
"created": "Sun, 11 May 2025 02:13:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xiao",
"Qian",
""
]
] |
2503.10512 | Devjeet Roy | Hooman Shahrokhi, Devjeet Raj Roy, Yan Yan, Venera Arnaoudova and Janaradhan Rao Doppa | Conformal Prediction Sets for Deep Generative Models via Reduction to Conformal Regression | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | We consider the problem of generating valid and small prediction sets by sampling outputs (e.g., software code and natural language text) from a black-box deep generative model for a given input (e.g., textual prompt). The validity of a prediction set is determined by a user-defined binary admissibility function depending on the target application. For example, requiring at least one program in the set to pass all test cases in code generation application. To address this problem, we develop a simple and effective conformal inference algorithm referred to as Generative Prediction Sets (GPS). Given a set of calibration examples and black-box access to a deep generative model, GPS can generate prediction sets with provable guarantees. The key insight behind GPS is to exploit the inherent structure within the distribution over the minimum number of samples needed to obtain an admissible output to develop a simple conformal regression approach over the minimum number of samples. Experiments on multiple datasets for code and math word problems using different large language models demonstrate the efficacy of GPS over state-of-the-art methods. | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 16:16:23 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 23:59:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shahrokhi",
"Hooman",
""
],
[
"Roy",
"Devjeet Raj",
""
],
[
"Yan",
"Yan",
""
],
[
"Arnaoudova",
"Venera",
""
],
[
"Doppa",
"Janaradhan Rao",
""
]
] |
2503.10929 | Moses Stewart | Moses Stewart | Constructing an Instrument as a Function of Covariates | null | null | null | null | econ.EM | http://creativecommons.org/licenses/by/4.0/ | Researchers often use instrumental variables (IV) models to investigate the causal relationship between an endogenous variable and an outcome while controlling for covariates. When an exogenous variable is unavailable to serve as the instrument for an endogenous treatment, a recurring empirical practice is to construct one from a nonlinear transformation of the covariates. We investigate how reliable these estimates are under mild forms of misspecification. Our main result shows that for instruments constructed from covariates, the IV estimand can be arbitrarily biased under mild forms of misspecification, even when imposing constant linear treatment effects. We perform a semi-synthetic exercise by calibrating data to alternative models proposed in the literature and estimating the average treatment effect. Our results show that IV specifications that use instruments constructed from covariates are non-robust to nonlinearity in the true structural function. | [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 22:32:16 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 22:08:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Stewart",
"Moses",
""
]
] |
2503.11413 | Bowy La Rivi\`ere | Niels T. Pronk, Bowy M. La Rivi\`ere, Natalia Chepiga | Deconfined quantum criticality in a frustrated Haldane chain with single-ion anisotropy | 6 pages, 5 figures | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We report a phase diagram of the antiferromagnetic spin-1 chain with nearest-neighbor Heisenberg and three-site interactions in the presence of single-ion anisotropy. We show that the Gaussian and Ising transitions that separate the topological Haldane phase from the two anisotropic phases eventually fuse into a higher symmetry point characterized by the Wess-Zumino-Witten (WZW) SU(2)$_2$ critical theory providing a lattice realization of the conformal embedding. On the other side of the WZW multi-critical point, the Ising critical line reappears together with the eight-vertex transition. This transition is a one-dimensional realization of a deconfined quantum criticality separating the dimerized and Ising antiferromagnetic phases - two ordered phases with incompatible order parameters. | [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 13:54:14 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 07:24:46 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pronk",
"Niels T.",
""
],
[
"La Rivière",
"Bowy M.",
""
],
[
"Chepiga",
"Natalia",
""
]
] |
2503.11593 | Bastian Bunzeck | Bastian Bunzeck, Daniel Duran, Sina Zarrie{\ss} | Do Construction Distributions Shape Formal Language Learning In German BabyLMs? | Accepted at CoNNL 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We analyze the influence of utterance-level construction distributions in German child-directed/child-available speech on the resulting word-level, syntactic and semantic competence (and their underlying learning trajectories) in small LMs, which we train on a novel collection of developmentally plausible language data for German. We find that trajectories are surprisingly robust for markedly different distributions of constructions in the training data, which have little effect on final accuracies and almost no effect on global learning trajectories. While syntax learning benefits from more complex utterances, word-level learning culminates in better scores with more fragmentary utterances. We argue that LMs trained on developmentally plausible data can contribute to debates on how conducive different kinds of linguistic stimuli are to language learning. | [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 17:02:45 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:44:18 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bunzeck",
"Bastian",
""
],
[
"Duran",
"Daniel",
""
],
[
"Zarrieß",
"Sina",
""
]
] |
2503.11808 | Alisa Sheinkman | Alisa Sheinkman, Sara Wade | Understanding the Trade-offs in Accuracy and Uncertainty Quantification: Architecture and Inference Choices in Bayesian Neural Networks | 24 pages | null | null | null | cs.LG stat.ME stat.ML | http://creativecommons.org/licenses/by/4.0/ | As modern neural networks get more complex, specifying a model with high predictive performance and sound uncertainty quantification becomes a more challenging task. Despite some promising theoretical results on the true posterior predictive distribution of Bayesian neural networks, the properties of even the most commonly used posterior approximations are often questioned. Computational burdens and intractable posteriors expose miscalibrated Bayesian neural networks to poor accuracy and unreliable uncertainty estimates. Approximate Bayesian inference aims to replace unknown and intractable posterior distributions with some simpler but feasible distributions. The dimensions of modern deep models, coupled with the lack of identifiability, make Markov chain Monte Carlo (MCMC) tremendously expensive and unable to fully explore the multimodal posterior. On the other hand, variational inference benefits from improved computational complexity but lacks the asymptotical guarantees of sampling-based inference and tends to concentrate around a single mode. The performance of both approaches heavily depends on architectural choices; this paper aims to shed some light on this by considering the computational costs, accuracy and uncertainty quantification in different scenarios including large width and out-of-sample data. To improve posterior exploration, different model averaging and ensembling techniques are studied, along with their benefits on predictive performance. In our experiments, variational inference overall provided better uncertainty quantification than MCMC; further, stacking and ensembles of variational approximations provided comparable accuracy to MCMC at a much-reduced cost. | [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 18:55:48 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 16:07:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sheinkman",
"Alisa",
""
],
[
"Wade",
"Sara",
""
]
] |
2503.12553 | Xianzu Wu | Xianzu Wu, Zhenxin Ai, Harry Yang, Ser-Nam Lim, Jun Liu and Huan Wang | Niagara: Normal-Integrated Geometric Affine Field for Scene Reconstruction from a Single View | null | null | null | null | cs.GR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in single-view 3D scene reconstruction have highlighted the challenges in capturing fine geometric details and ensuring structural consistency, particularly in high-fidelity outdoor scene modeling. This paper presents Niagara, a new single-view 3D scene reconstruction framework that can faithfully reconstruct challenging outdoor scenes from a single input image for the first time.
Our approach integrates monocular depth and normal estimation as input, which substantially improves its ability to capture fine details, mitigating common issues like geometric detail loss and deformation.
Additionally, we introduce a geometric affine field (GAF) and 3D self-attention as geometry-constraint, which combines the structural properties of explicit geometry with the adaptability of implicit feature fields, striking a balance between efficient rendering and high-fidelity reconstruction.
Our framework finally proposes a specialized encoder-decoder architecture, where a depth-based 3D Gaussian decoder is proposed to predict 3D Gaussian parameters, which can be used for novel view synthesis. Extensive results and analyses suggest that our Niagara surpasses prior SoTA approaches such as Flash3D in both single-view and dual-view settings, significantly enhancing the geometric accuracy and visual fidelity, especially in outdoor scenes. | [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 15:50:18 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 05:47:02 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wu",
"Xianzu",
""
],
[
"Ai",
"Zhenxin",
""
],
[
"Yang",
"Harry",
""
],
[
"Lim",
"Ser-Nam",
""
],
[
"Liu",
"Jun",
""
],
[
"Wang",
"Huan",
""
]
] |
2503.12607 | Fengxing Zhu | Fengxing Zhu | Bootstrap percolation on a generalized Hamming cube \MakeUppercase{\romannumeral 2} | null | null | null | null | math.CO math.PR | http://creativecommons.org/licenses/by/4.0/ | In this paper we investigate the critical probability $p_c(Q_n,r)$ for bootstrap percolation with the infection threshold $r$ on the $n$-dimensional hypercube $Q_n$ with vertex set $V(Q_n)=\{0,1\}^n$ and edges connecting the pairs at Hamming distance $1$. More precisely, by utilizing the techniques developed by Balogh, Bollob{\'a}s, and Morris (2009), we determine the first-order term of $p_c(Q_n,n^a)$ where $\frac{2}{3}<a< 1$. Additionally, we obtain the critical probability $p_c(Q_{k,n},r)$ for bootstrap percolation with the infection threshold $r=\frac{N}{2}$ on the generalized $n$-dimensional hypercube $Q_{k,n}$ with vertex set $V(Q_{k,n})=\{0,1\}^n$ and edges connecting the pairs at Hamming distance $1,2,\dots,k$, where $k\ge 2$ and $N=\sum_{i=1}^k\binom{n}{i}$. More precisely, we obtain the first-order term of $p_c(Q_{k,n},\frac{N}{2})$ and some bounds on the second-order term by extending the main theorem from Balogh, Bollob{\'a}s, and Morris (2009). | [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 18:43:05 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Mar 2025 17:00:56 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 14:46:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhu",
"Fengxing",
""
]
] |
2503.12775 | Tomoki Yamagami | Akihiro Narimatsu and Tomoki Yamagami | A study of the Antlion Random Walk | 12 pages; 3 figures, including 10 subfigures in total | null | null | null | math.PR cs.LO | http://creativecommons.org/licenses/by/4.0/ | This paper treats a new type of random walk referred to as an Antlion Random Walk (ARW), which is motivated by mathematical modeling of the decision-making process using chaotic semiconductor lasers with memory parameters. We discuss the dependency of the property of the probability distribution of ARWs on the memory parameter $\alpha$ and discuss uniqueness of them in contrast to the conventional, simple RWs through similarity to the normal distribution. | [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 03:16:40 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:24:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Narimatsu",
"Akihiro",
""
],
[
"Yamagami",
"Tomoki",
""
]
] |
2503.12917 | Lin-Han Jia | Lin-Han Jia, Wen-Chao Hu, Jie-Jing Shao, Lan-Zhe Guo, Yu-Feng Li | Verification Learning: Make Unsupervised Neuro-Symbolic System Feasible | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | The current Neuro-Symbolic (NeSy) Learning paradigm suffers from an over-reliance on labeled data, so if we completely disregard labels, it leads to less symbol information, a larger solution space, and more shortcuts-issues that current Nesy systems cannot resolve. This paper introduces a novel learning paradigm, Verification Learning (VL), which addresses this challenge by transforming the label-based reasoning process in Nesy into a label-free verification process. VL achieves excellent learning results solely by relying on unlabeled data and a function that verifies whether the current predictions conform to the rules. We formalize this problem as a Constraint Optimization Problem (COP) and propose a Dynamic Combinatorial Sorting (DCS) algorithm that accelerates the solution by reducing verification attempts, effectively lowering computational costs and introduce a prior alignment method to address potential shortcuts. Our theoretical analysis points out which tasks in Nesy systems can be completed without labels and explains why rules can replace infinite labels for some tasks, while for others the rules have no effect. We validate the proposed framework through several fully unsupervised tasks including addition, sort, match, and chess, each showing significant performance and efficiency improvements. | [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 08:28:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 03:48:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jia",
"Lin-Han",
""
],
[
"Hu",
"Wen-Chao",
""
],
[
"Shao",
"Jie-Jing",
""
],
[
"Guo",
"Lan-Zhe",
""
],
[
"Li",
"Yu-Feng",
""
]
] |
2503.13323 | Pedro H. C. Sant'Anna | Andrew Baker, Brantly Callaway, Scott Cunningham, Andrew Goodman-Bacon, Pedro H. C. Sant'Anna | Difference-in-Differences Designs: A Practitioner's Guide | null | null | null | null | econ.EM stat.ME | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Difference-in-differences (DiD) is arguably the most popular quasi-experimental research design. Its canonical form, with two groups and two periods, is well-understood. However, empirical practices can be ad hoc when researchers go beyond that simple case. This article provides an organizing framework for discussing different types of DiD designs and their associated DiD estimators. It discusses covariates, weights, handling multiple periods, and staggered treatments. The organizational framework, however, applies to other extensions of DiD methods as well. | [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 16:01:22 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 01:30:29 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 03:57:27 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Baker",
"Andrew",
""
],
[
"Callaway",
"Brantly",
""
],
[
"Cunningham",
"Scott",
""
],
[
"Goodman-Bacon",
"Andrew",
""
],
[
"Sant'Anna",
"Pedro H. C.",
""
]
] |
2503.13562 | Lin-Han Jia | Lin-Han Jia, Lan-Zhe Guo, Zhi Zhou, Si-Ye Han, Zi-Wen Li, Yu-Feng Li | Achieving Unbiased Multi-Instance Learning via Balanced Fine-Grained Positive-Unlabeled Learning | null | null | null | null | stat.ML cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | In real-world applications, it is often challenging to detect anomalous samples when the anomalous information they contain is extremely limited. In such cases, both macro-level and micro-level detection using multi-instance learning (MIL) encounter significant difficulties. The former struggles because normal and anomalous samples are highly similar and hard to distinguish at the macro level, while the latter is limited by the lack of labels at the micro level. In MIL, micro-level labels are inferred from macro-level labels, which can lead to severe bias. Moreover, the more imbalanced the distribution between normal and anomalous samples, the more pronounced these limitations become. In this study, we observe that the MIL problem can be elegantly transformed into a fine-grained Positive-Unlabeled (PU) learning problem. This transformation allows us to address the imbalance issue in an unbiased manner using a micro-level balancing mechanism. To this end, we propose a novel framework-Balanced Fine-Grained Positive-Unlabeled (BFGPU)-based on rigorous theoretical foundations to address the challenges above. Extensive experiments on both public and real-world datasets demonstrate the effectiveness of BFGPU, which outperforms existing methods, even in extreme scenarios where both macro and micro-level distributions are highly imbalanced. The code is open-sourced at https://github.com/BFGPU/BFGPU. | [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 07:42:27 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:34:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jia",
"Lin-Han",
""
],
[
"Guo",
"Lan-Zhe",
""
],
[
"Zhou",
"Zhi",
""
],
[
"Han",
"Si-Ye",
""
],
[
"Li",
"Zi-Wen",
""
],
[
"Li",
"Yu-Feng",
""
]
] |
2503.14856 | Akhil Jaini | Akhil Jaini, Adam T. Deller, Yuanming Wang, Emil Lenc and Marcin Glowacki | Enhanced Astrometry of the Rapid ASKAP Continuum Survey for Precise Localisation of Fast Radio Bursts | Manuscript has been accepted for publication in the Publications of the Astronomical Society of Australia (PASA). 22 pages, 14 figures, 3 tables | Publ. Astron. Soc. Aust. 42 (2025) e060 | 10.1017/pasa.2025.28 | null | astro-ph.IM astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Fast Radio Bursts (FRBs) are short, intense radio signals from distant astrophysical sources, and their accurate localisation is crucial for probing their origins and utilising them as cosmological tools. This study focuses on improving the astrometric precision of FRBs discovered by the Australian Square Kilometre Array Pathfinder (ASKAP) by correcting systematic positional errors in the Rapid ASKAP Continuum Survey (RACS), which is used as a primary reference for ASKAP FRB localisation. We present a detailed methodology for refining astrometry in two RACS epochs (RACS-Low1 and RACS-Low3) through crossmatching with the Wide-field Infrared Survey Explorer (WISE) catalogue. The uncorrected RACS-Low1 and RACS-Low3 catalogues had significant astrometric offsets, with all-sky median values of $0.58''$ in RA and $-0.26''$ in Dec. (RACS-Low1) and $0.29''$ in RA and $1.24''$ in Dec. (RACS-Low3), with a substantial and direction-dependent scatter around these values. After correction, the median offset was completely eliminated, and the 68\% confidence interval in the all-sky residuals was reduced to $0.2''$ or better for both surveys. By validating the corrected catalogues against other, independent radio surveys, we conclude that the individual corrected RACS source positions are accurate to a 1-$\sigma$ confidence level of $0.3''$ over the bulk of the survey area, degrading slightly to $0.4''$ near the Galactic plane. This work lays the groundwork to extend our corrections to the full RACS catalogue that will enhance future radio observations, particularly for FRB studies. | [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 03:25:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jaini",
"Akhil",
""
],
[
"Deller",
"Adam T.",
""
],
[
"Wang",
"Yuanming",
""
],
[
"Lenc",
"Emil",
""
],
[
"Glowacki",
"Marcin",
""
]
] |
2503.14973 | Rishav Rishav | Rishav Rishav, Somjit Nath, Vincent Michalski, Samira Ebrahimi Kahou | Behaviour Discovery and Attribution for Explainable Reinforcement Learning | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Building trust in reinforcement learning (RL) agents requires understanding why they make certain decisions, especially in high-stakes applications like robotics, healthcare, and finance. Existing explainability methods often focus on single states or entire trajectories, either providing only local, step-wise insights or attributing decisions to coarse, episodelevel summaries. Both approaches miss the recurring strategies and temporally extended patterns that actually drive agent behavior across multiple decisions. We address this gap by proposing a fully offline, reward-free framework for behavior discovery and segmentation, enabling the attribution of actions to meaningful and interpretable behavior segments that capture recurring patterns appearing across multiple trajectories. Our method identifies coherent behavior clusters from state-action sequences and attributes individual actions to these clusters for fine-grained, behavior-centric explanations. Evaluations on four diverse offline RL environments show that our approach discovers meaningful behaviors and outperforms trajectory-level baselines in fidelity, human preference, and cluster coherence. Our code is publicly available. | [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 08:06:00 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 21:43:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Rishav",
"Rishav",
""
],
[
"Nath",
"Somjit",
""
],
[
"Michalski",
"Vincent",
""
],
[
"Kahou",
"Samira Ebrahimi",
""
]
] |
2503.16639 | Thomas Kreutz | Thomas Kreutz, Max M\"uhlh\"auser, Alejandro Sanchez Guinea | Whenever, Wherever: Towards Orchestrating Crowd Simulations with Spatio-Temporal Spawn Dynamics | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Realistic crowd simulations are essential for immersive virtual environments, relying on both individual behaviors (microscopic dynamics) and overall crowd patterns (macroscopic characteristics). While recent data-driven methods like deep reinforcement learning improve microscopic realism, they often overlook critical macroscopic features such as crowd density and flow, which are governed by spatio-temporal spawn dynamics, namely, when and where agents enter a scene. Traditional methods, like random spawn rates, stochastic processes, or fixed schedules, are not guaranteed to capture the underlying complexity or lack diversity and realism. To address this issue, we propose a novel approach called nTPP-GMM that models spatio-temporal spawn dynamics using Neural Temporal Point Processes (nTPPs) that are coupled with a spawn-conditional Gaussian Mixture Model (GMM) for agent spawn and goal positions. We evaluate our approach by orchestrating crowd simulations of three diverse real-world datasets with nTPP-GMM. Our experiments demonstrate the orchestration with nTPP-GMM leads to realistic simulations that reflect real-world crowd scenarios and allow crowd analysis. | [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 18:46:41 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 18:26:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kreutz",
"Thomas",
""
],
[
"Mühlhäuser",
"Max",
""
],
[
"Guinea",
"Alejandro Sanchez",
""
]
] |
2503.16974 | Victor Xiaoqi Wang | Julian Junyan Wang, Victor Xiaoqi Wang | Assessing Consistency and Reproducibility in the Outputs of Large Language Models: Evidence Across Diverse Finance and Accounting Tasks | 89 pages, 20 tables, 15 figures | null | null | null | q-fin.GN cs.AI cs.CE cs.CL cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study provides the first comprehensive assessment of consistency and reproducibility in Large Language Model (LLM) outputs in finance and accounting research. We evaluate how consistently LLMs produce outputs given identical inputs through extensive experimentation with 50 independent runs across five common tasks: classification, sentiment analysis, summarization, text generation, and prediction. Using three OpenAI models (GPT-3.5-turbo, GPT-4o-mini, and GPT-4o), we generate over 3.4 million outputs from diverse financial source texts and data, covering MD&As, FOMC statements, finance news articles, earnings call transcripts, and financial statements. Our findings reveal substantial but task-dependent consistency, with binary classification and sentiment analysis achieving near-perfect reproducibility, while complex tasks show greater variability. More advanced models do not consistently demonstrate better consistency and reproducibility, with task-specific patterns emerging. LLMs significantly outperform expert human annotators in consistency and maintain high agreement even where human experts significantly disagree. We further find that simple aggregation strategies across 3-5 runs dramatically improve consistency. We also find that aggregation may come with an additional benefit of improved accuracy for sentiment analysis when using newer models. Simulation analysis reveals that despite measurable inconsistency in LLM outputs, downstream statistical inferences remain remarkably robust. These findings address concerns about what we term "G-hacking," the selective reporting of favorable outcomes from multiple Generative AI runs, by demonstrating that such risks are relatively low for finance and accounting tasks. | [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 09:43:37 GMT"
},
{
"version": "v2",
"created": "Wed, 26 Mar 2025 17:48:00 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 02:40:54 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Julian Junyan",
""
],
[
"Wang",
"Victor Xiaoqi",
""
]
] |
2503.17997 | Niels Kj{\ae}rgaard | Matthew Cloutman, Matthew Chilcott, Alexander Elliott, J. Susanne Otto, Amita B. Deb, and Niels Kj{\ae}rgaard | Quantum-enabled Rydberg atomic polarimetry of radio-frequency fields | v3: intro rewritten | null | null | null | quant-ph physics.atom-ph physics.ins-det | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Rydberg atoms efficiently link photons between the radio-frequency (RF) and optical domains. They furnish a medium in which the presence of an RF field imprints on the transmission of a probe laser beam by altering the coherent coupling between atomic quantum states. The immutable atomic energy structure underpins quantum-metrological RF field measurements and has driven intensive efforts to realize inherently self-calibrated sensing devices. Here we investigate spectroscopic signatures owing to the angular momentum quantization of the atomic states utilized in an electromagnetically-induced transparency (EIT) sensing scheme for linearly polarized RF fields. Specific combinations of atomic terms are shown to give rise to universal, distinctive fingerprints in the detected optical fields upon rotating the RF field polarization. Using a dressed state picture, we identify two types of atomic angular momentum ladders that display strikingly disparate spectroscopic signatures, including the complementary absence or presence of a central spectral EIT peak. Our study adds important insights into the prospects of Rydberg atomic gases for quantum metrological electric field characterization. In particular, it calls into question prevailing interpretations of SI-traceable Rydberg atom electrometers. | [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 09:13:36 GMT"
},
{
"version": "v2",
"created": "Sun, 30 Mar 2025 09:35:37 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 03:31:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Cloutman",
"Matthew",
""
],
[
"Chilcott",
"Matthew",
""
],
[
"Elliott",
"Alexander",
""
],
[
"Otto",
"J. Susanne",
""
],
[
"Deb",
"Amita B.",
""
],
[
"Kjærgaard",
"Niels",
""
]
] |
2503.18529 | Matteo Giordano | Claudio Bonanno, Matteo Giordano | Localization in the Dirac spectrum and gauge-field topology | Minor revisions, to appear in PRD; 14 pages, 8 figures | null | null | null | hep-lat cond-mat.dis-nn | http://creativecommons.org/licenses/by/4.0/ | We study localization of the low Dirac modes in 3+1 dimensional pure $\mathrm{SU}(3)$ gauge theory at zero and nonzero imaginary $\theta$ angle, with the aim of better characterizing the relation between low-mode localization and topological features of gauge theories. We show that the mobility edge observed in the deconfined phase at $\theta=0$ is present also at nonzero $\theta$, appearing exactly at the deconfinement transition. We find that the mobility edge is affected by topology only indirectly, through its effects on the ordering of the Polyakov loop. Moreover, the change in the mobility edge is strongly correlated with the change in the Polyakov-loop expectation value, both as one moves along the critical line, and as one departs from it toward higher temperatures. This further strengthens the connection between low-mode localization and deconfinement, showing in particular the key role played by the ordering of the Polyakov loop. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 10:35:19 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 11:56:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bonanno",
"Claudio",
""
],
[
"Giordano",
"Matteo",
""
]
] |
2503.18605 | Georgios Tzounas | Liya Huang, Georgios Tzounas | Matrix Pencil-Based Analysis of Multirate Simulation Schemes | null | null | null | null | eess.SY cs.NA cs.SY math.NA | http://creativecommons.org/licenses/by/4.0/ | This paper focuses on multirate time-domain simulations of power system models. It proposes a matrix pencil-based approach to evaluate the spurious numerical deformation introduced into power system dynamics by a given multirate integration scheme. Moreover, it considers the problem of multirate partitioning and discusses a strategy for allocating state and algebraic variables to fast and slow subsystems based on modal participation factors (PFs). The suitability and features of the proposed approach are illustrated through numerical simulations that assess the accuracy effects of interfacing, as well as of various prediction and solution methods. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 12:04:13 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 07:16:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Huang",
"Liya",
""
],
[
"Tzounas",
"Georgios",
""
]
] |
2503.18682 | Samuel Rota Bul\`o | Samuel Rota Bul\`o, Nemanja Bartolovic, Lorenzo Porzi, Peter Kontschieder | Hardware-Rasterized Ray-Based Gaussian Splatting | null | null | null | null | cs.CV cs.GR | http://creativecommons.org/licenses/by-sa/4.0/ | We present a novel, hardware rasterized rendering approach for ray-based 3D Gaussian Splatting (RayGS), obtaining both fast and high-quality results for novel view synthesis. Our work contains a mathematically rigorous and geometrically intuitive derivation about how to efficiently estimate all relevant quantities for rendering RayGS models, structured with respect to standard hardware rasterization shaders. Our solution is the first enabling rendering RayGS models at sufficiently high frame rates to support quality-sensitive applications like Virtual and Mixed Reality. Our second contribution enables alias-free rendering for RayGS, by addressing MIP-related issues arising when rendering diverging scales during training and testing. We demonstrate significant performance gains, across different benchmark scenes, while retaining state-of-the-art appearance quality of RayGS. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 13:53:30 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 09:31:20 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bulò",
"Samuel Rota",
""
],
[
"Bartolovic",
"Nemanja",
""
],
[
"Porzi",
"Lorenzo",
""
],
[
"Kontschieder",
"Peter",
""
]
] |
2503.18902 | Rok Zitko | Teodor Ili\v{c}in and Rok \v{Z}itko | Variational solution of the superconducting Anderson impurity model and the band-edge singularity phenomena | 28 pages, 15 figures | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | We propose a set of variational wavefunctions for the sub-gap spin-doublet and spin-singlet eigenstates of the particle-hole symmetric superconducting Anderson impurity model. The wavefunctions include up to two Bogoliubov quasiparticles in the continuum which is necessary to correctly capture the weak-coupling asymptotics in all parameter regimes. The eigenvalue problems reduce to solving transcendental equations. We investigate how the lowest singlet state evolves with increasing charge repulsion $U$, transitioning from a proximitized state (a superposition of empty and doubly occupied impurity orbitals, corresponding to an Andreev bound state) to a local moment that is Kondo screened by Bogoliubov quasiparticles (Yu-Shiba-Rusinov state). This change occurs for $U = 2\Delta$, where $\Delta$ is the BCS gap. At this point, the band-edge effects make the eigenenergy scale in a singular way as $\Gamma^{2/3}$, where $\Gamma$ is the hybridization strength. Away from this special point, regular $\Gamma$-linear behavior is recovered, but only for $\Gamma \lesssim (U/2-\Delta)^2/\Delta$. The singular behavior thus extends over a broad range of parameters, including those relevant for some quantum devices in current use. The singular state is an equal-superposition state with maximal fluctuations between the local impurity charge configurations. Accurately capturing the band-edge singularity requires a continuum model, and it cannot be correctly described by discrete (truncated) models such as the zero-bandwidth approximation or the superconducting atomic limit. We determine the region of parameter space where the second spin-singlet state exists: in addition to the whole $U<2\Delta$ ABS region, it also includes a small part of the $U>2\Delta$ YSR region for finite values of $\Gamma$, as long as some ABS wavefunction component is admixed. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 17:15:15 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 10:27:56 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Iličin",
"Teodor",
""
],
[
"Žitko",
"Rok",
""
]
] |
2503.18913 | Sebastian Hell | Sebastian Hell, Julian Sp\"athe, Morten F{\o}rre, Robert Klas, Jan Rothhardt, Jens Limpert, Gerhard G Paulus, Robert Moshammer, Christian Ott, Stephan Fritzsche, and Matthias K\"ubel | Coincidence measurement of two-photon double ionization of argon through an autoionizing resonance | 8 pages, 5 figures | null | null | null | physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present coincidence measurements of two-photon double-ionization (TPDI) of argon driven by femtosecond pulses tunable around 26.5 eV photon energy, which are obtained from a high-harmonic generation source. The measured photoelectron spectra are interpreted with regard to three TPDI mechanisms. Theoretical predictions are obtained by an approximate model for direct TPDI and atomic structure calculations, which are implemented into a Monte Carlo simulation. The prevailing mechanism involves the excitation and prompt photoionization of an autoionizing resonance in neutral argon. We provide evidence for pronounced electron-electron interaction in this ultrafast ionization process. Furthermore, we show that the dominant TPDI mechanism can be altered by slight tuning of the photon energy. The present work paves the way for scrutinizing and controlling non-linear photoionization in the extreme ultraviolet using table-top sources. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 17:26:10 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Mar 2025 16:12:48 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 09:52:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hell",
"Sebastian",
""
],
[
"Späthe",
"Julian",
""
],
[
"Førre",
"Morten",
""
],
[
"Klas",
"Robert",
""
],
[
"Rothhardt",
"Jan",
""
],
[
"Limpert",
"Jens",
""
],
[
"Paulus",
"Gerhard G",
""
],
[
"Moshammer",
"Robert",
""
],
[
"Ott",
"Christian",
""
],
[
"Fritzsche",
"Stephan",
""
],
[
"Kübel",
"Matthias",
""
]
] |
2503.19003 | Josef Grus | Josef Grus, Claire Hanen, Zden\v{e}k Hanz\'alek | Periodic Chains Scheduling on Dedicated Resources -- A Crucial Problem in Time-Sensitive Networks | 54 pages, accepted by Computers & Operations Research | Computers & Operations Research (2025), 180, 107072 | 10.1016/j.cor.2025.107072 | null | cs.NI | http://creativecommons.org/licenses/by/4.0/ | Periodic messages transfer data from sensors to actuators in cars, planes, and complex production machines. When considering a given routing, the unicast message starts at its source and goes over several dedicated resources to reach its destination. Such unicast message can be represented as a chain of point-to-point communications. Thus, the scheduling of the periodic chains is a principal problem in time-triggered Ethernet, like IEEE 802.1Qbv Time-Sensitive Networks. This paper studies a strongly NP-hard periodic scheduling problem with harmonic periods, task chains, and dedicated resources. We analyze the problem on several levels of generality and complexity and provide the corresponding proofs. We describe a solution methodology to find a feasible schedule that minimizes the chains' degeneracy related to start-to-end latency normalized in the number of periods. We use the local search with the first fit scheduling heuristic, which we warm-start with a constraint programming model. This notably improves the schedulability of instances with up to 100% utilization and thousands (and more) of tasks, with high-quality solutions found in minutes. An efficient constraint programming matheuristic significantly reduces the degeneracy of the found schedules even further. The method is evaluated on sets of industrial-, avionic-, and automotive-inspired instances. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 13:48:52 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 10:43:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Grus",
"Josef",
""
],
[
"Hanen",
"Claire",
""
],
[
"Hanzálek",
"Zdeněk",
""
]
] |
2503.19061 | Ivan de Martino | Matteo Califano, Ivan De Martino, Daniele Vernieri | Joint estimation of the cosmological model and the mass and redshift distributions of the binary black hole population with the Einstein Telescope | 11 pages, 5 figures, 4 Tables, Accepted on PRD | Physical Review D 111, 123535 (2025) | 10.1103/7nqz-3pzn | null | astro-ph.CO gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the capability of constraining the mass and redshift distributions of binary black hole systems jointly with the underlying cosmological model using one year of observations of the Einstein Telescope. To this aim, we fixed the underlying cosmological model to a flat $\Lambda$CDM model, then we considered the mass distribution given by a smoothed power law, and the redshift distributions given by the Madau-Dickinson model. We built mock catalogs with different SNR thresholds, and finally inferred astrophysical and cosmological parameters jointly adopting a hierarchical Bayesian framework. We found that as the SNR threshold decreases, the precision on the matter density parameter $\Omega_{m,0}$ and the Hubble constant $H_0$, improves significantly due to the increased number of detectable events at high redshift. However, degeneracies between cosmological and astrophysical parameters exist and evolve with the SNR threshold. Finally, we showed that one year of observations will serve to reconstruct the mass distribution with its features. Conversely, the redshift distribution will be poorly constrained and will need more observations to improve. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 18:39:10 GMT"
},
{
"version": "v2",
"created": "Sun, 08 Jun 2025 08:31:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Califano",
"Matteo",
""
],
[
"De Martino",
"Ivan",
""
],
[
"Vernieri",
"Daniele",
""
]
] |
2503.19121 | Ilaria Marini | I. Marini, P. Popesso, K. Dolag, V. Biffi, S. Vladutescu-Zopp, T. Castro, V. Toptun, N. de Isidio, A. Dev, D. Mazengo, J. Comparat, C. Gouin, N. Malavasi, A. Merloni, T. Mroczkowski, G. Ponti, S. Shreeram, Y. Zhang | The impact of assembly history on the X-ray detectability of halos. From galaxy groups to galaxy clusters | null | A&A 698, A191 (2025) | 10.1051/0004-6361/202554677 | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Galaxy groups represent a significant fraction of the halo population, playing a crucial role in galaxy formation and evolution. However, their detection in X-rays remains challenging, raising questions about the physical mechanisms driving their detectability in current surveys. Using the Magneticum simulations, we construct a mock X-ray lightcone of the local Universe ($z<0.2$) to investigate the selection function of galaxy groups and clusters. We find that AGN activity is a key driver of baryon depletion, but late-time mergers boost X-ray brightness by replenishing the gas reservoir in the halos, highlighting the interplay between feedback processes and the environment. Our analysis shows that X-ray bright groups experience sustained late-time mass accretion, maintaining higher gas fractions and fueling the central supermassive black holes (SMBH), further increasing the X-ray emissivity in the core. In contrast, X-ray faint groups form earlier and lose most of their gas over time, resembling fossil groups. Magneticum predicts strong anti-correlations between gas fraction (or X-ray luminosity) and SMBH mass, stellar mass (both in the central galaxy and intracluster light), and group richness at fixed halo mass. We derive predictions on the hot gas fraction at fixed halos mass (e.g. a group of total mass $M_{500}=10^{13} M_{\odot}$ can have hot gas fractions in the range $f_\mathrm{gas}=0.02-0.06$ and a central SMBH with a median mass of $M_\mathrm{BH}=10^9 M_{\odot}$ and a scatter of $0.5$ dex) compatible with the most recent measurements of the baryonic fraction. These findings will aid the interpretation of future X-ray surveys, demonstrating the power of simulation-based inference. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 20:15:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:34:56 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Marini",
"I.",
""
],
[
"Popesso",
"P.",
""
],
[
"Dolag",
"K.",
""
],
[
"Biffi",
"V.",
""
],
[
"Vladutescu-Zopp",
"S.",
""
],
[
"Castro",
"T.",
""
],
[
"Toptun",
"V.",
""
],
[
"de Isidio",
"N.",
""
],
[
"Dev",
"A.",
""
],
[
"Mazengo",
"D.",
""
],
[
"Comparat",
"J.",
""
],
[
"Gouin",
"C.",
""
],
[
"Malavasi",
"N.",
""
],
[
"Merloni",
"A.",
""
],
[
"Mroczkowski",
"T.",
""
],
[
"Ponti",
"G.",
""
],
[
"Shreeram",
"S.",
""
],
[
"Zhang",
"Y.",
""
]
] |
2503.20431 | Rohit Goswami MInstP MBCS MRSC | Rohit Goswami (1,2 and 3) and Ashwini Kumar Rawat (1) and Sonaly Goswami (1) and Debabrata Goswami (1) ((1) Department of Chemistry, Indian Institute of Kanpur, India (2) Science Institute, Iceland (3) Quansight Labs, Austin, TX, USA) | Compositional Analysis of Fragrance Accords Using Femtosecond Thermal Lens Spectroscopy | 26 pages, 6 figures, 14 supplementary figures | null | 10.1002/asia.202500521 | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Femtosecond thermal lens spectroscopy (FTLS) is a powerful analytical tool, yet its application to complex, multi-component mixtures like fragrance accords remains limited. Here, we introduce and validate a unified metric, the Femtosecond Thermal Lens Integrated Magnitude (FTL-IM), to characterize such mixtures. The FTL-IM, derived from the integrated signal area, provides a direct, model-free measure of the total thermo-optical response, including critical convective effects. Applying the FTL-IM to complex six-component accords, we demonstrate its utility in predicting a mixture's thermal response from its composition through linear additivity with respect to component mole fractions. Our method quantifies the accords' behavior, revealing both the baseline contributions of components and the dominant, non-linear effects of highly-active species like Methyl Anthranilate. This consistency is validated across single-beam Z-scan, dual-beam Z-scan, and time-resolved FTLS measurements. The metric also demonstrates the necessity of single-beam measurements for interpreting dual-beam data. This work establishes a rapid, quantitative method for fragrance analysis, offering advantages for quality control by directly linking a mixture's bulk thermo-optical properties to its composition. | [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 11:06:13 GMT"
},
{
"version": "v2",
"created": "Thu, 05 Jun 2025 15:19:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Goswami",
"Rohit",
"",
"1,2 and 3"
],
[
"Rawat",
"Ashwini Kumar",
"",
"Department of Chemistry, Indian Institute of Kanpur, India"
],
[
"Goswami",
"Sonaly",
"",
"Department of Chemistry, Indian Institute of Kanpur, India"
],
[
"Goswami",
"Debabrata",
"",
"Department of Chemistry, Indian Institute of Kanpur, India"
]
] |
2503.20707 | Nadine Meyer | Gregoire F. M. Tomassi and Daniel Veldhuizen and Bruno Melo and Davide Candoli and Andreu Riera-Campeny and Oriol Romero-Isart and Nadine Meyer and Romain Quidant | Accelerated State Expansion of a Nanoparticle in a Dark Inverted Potential | null | null | null | null | quant-ph physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While the wave packet of a massive particle grows linearly under free dynamics, it grows exponentially in an inverted harmonic potential, offering a pathway to rapidly increase quantum fluctuations to macroscopic dimensions. In this work, we experimentally demonstrate this principle by expanding the center-of-mass thermal state of a 125nm silica nanoparticle to a position uncertainty of 43.4nm within 260 $\mu$s. This expansion, achieved using an inverted dark potential to minimize decoherence from photon recoil, represents a 952-fold increase, reaching a scale comparable to the nanoparticle's physical size. This work represents a key advancement toward preparing macroscopic quantum superpositions at unprecedented mass and length scales. | [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 16:40:33 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:15:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tomassi",
"Gregoire F. M.",
""
],
[
"Veldhuizen",
"Daniel",
""
],
[
"Melo",
"Bruno",
""
],
[
"Candoli",
"Davide",
""
],
[
"Riera-Campeny",
"Andreu",
""
],
[
"Romero-Isart",
"Oriol",
""
],
[
"Meyer",
"Nadine",
""
],
[
"Quidant",
"Romain",
""
]
] |
2503.20848 | Benjamin Laufer | Benjamin Laufer, Jon Kleinberg, Hoda Heidari | The Backfiring Effect of Weak AI Safety Regulation | 35 pages, 5 figures | null | null | null | cs.GT cs.AI cs.CY econ.TH | http://creativecommons.org/licenses/by/4.0/ | Recent policy proposals aim to improve the safety of general-purpose AI, but there is little understanding of the efficacy of different regulatory approaches to AI safety. We present a strategic model that explores the interactions between safety regulation, the general-purpose AI creators, and domain specialists--those who adapt the technology for specific applications. Our analysis examines how different regulatory measures, targeting different parts of the AI development chain, affect the outcome of this game. In particular, we assume AI technology is characterized by two key attributes: safety and performance. The regulator first sets a minimum safety standard that applies to one or both players, with strict penalties for non-compliance. The general-purpose creator then invests in the technology, establishing its initial safety and performance levels. Next, domain specialists refine the AI for their specific use cases, updating the safety and performance levels and taking the product to market. The resulting revenue is then distributed between the specialist and generalist through a revenue-sharing parameter. Our analysis reveals two key insights: First, weak safety regulation imposed predominantly on domain specialists can backfire. While it might seem logical to regulate AI use cases, our analysis shows that weak regulations targeting domain specialists alone can unintentionally reduce safety. This effect persists across a wide range of settings. Second, in sharp contrast to the previous finding, we observe that stronger, well-placed regulation can in fact mutually benefit all players subjected to it. When regulators impose appropriate safety standards on both general-purpose AI creators and domain specialists, the regulation functions as a commitment device, leading to safety and performance gains, surpassing what is achieved under no regulation or regulating one player alone. | [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 16:08:22 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:26:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Laufer",
"Benjamin",
""
],
[
"Kleinberg",
"Jon",
""
],
[
"Heidari",
"Hoda",
""
]
] |
2503.21192 | Xinwei Cai | Xinwei Cai, Xuejin Li, Xin Bian | Dynamics of an elliptical cylinder in confined Poiseuille flow under Navier slip boundary conditions | 17 pages, 14 figures | null | 10.1007/s10409-025-25236-x | null | physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A comprehensive understanding of surface wetting phenomena in microchannels is essential for optimizing particle transport and filtration processes. This study numerically investigates the dynamics of a freely suspended elliptical cylinder in confined Poiseuille flow, with a focus on Navier slip boundary conditions. The smoothed particle hydrodynamics method is employed, which is advantageous for its Lagrangian framework in handling dynamic fluid-solid interfaces with slip. Our results demonstrate that the slip conditions enable precise control over inertial focusing positions and particle motion modes. Compared to no-slip scenarios, unilateral wall slip induces two novel motion types: "leaning" and "rolling". When equal slip lengths are applied to both walls, even small slip values facilitate off-center inertial focusing and elevate equilibrium positions. Slip on the cylinder surface further enhances inertial lift while suppressing rotational dynamics. In particular, under strong confinement or with large particle-surface slip lengths, we identify an additional distinct motion regime termed "inclined." These findings provide new insights for active particle manipulation in microfluidic applications. | [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 06:18:35 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Cai",
"Xinwei",
""
],
[
"Li",
"Xuejin",
""
],
[
"Bian",
"Xin",
""
]
] |
2503.22313 | Zenghui Chang | Zenghui Chang, Yang Zhang, Hu Tan, Hong Cai Chen | Hybrid Time-Domain Behavior Model Based on Neural Differential Equations and RNNs | 7 pages,5 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nonlinear dynamics system identification is crucial for circuit emulation. Traditional continuous-time domain modeling approaches have limitations in fitting capability and computational efficiency when used for modeling circuit IPs and device behaviors.This paper presents a novel continuous-time domain hybrid modeling paradigm. It integrates neural network differential models with recurrent neural networks (RNNs), creating NODE-RNN and NCDE-RNN models based on neural ordinary differential equations (NODE) and neural controlled differential equations (NCDE), respectively.Theoretical analysis shows that this hybrid model has mathematical advantages in event-driven dynamic mutation response and gradient propagation stability. Validation using real data from PIN diodes in high-power microwave environments shows NCDE-RNN improves fitting accuracy by 33\% over traditional NCDE, and NODE-RNN by 24\% over CTRNN, especially in capturing nonlinear memory effects.The model has been successfully deployed in Verilog-A and validated through circuit emulation, confirming its compatibility with existing platforms and practical value.This hybrid dynamics paradigm, by restructuring the neural differential equation solution path, offers new ideas for high-precision circuit time-domain modeling and is significant for complex nonlinear circuit system modeling. | [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 10:42:52 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 13:15:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chang",
"Zenghui",
""
],
[
"Zhang",
"Yang",
""
],
[
"Tan",
"Hu",
""
],
[
"Chen",
"Hong Cai",
""
]
] |
2503.23394 | Sangyoon Bae | Sangyoon Bae, Junbeom Kwon, Shinjae Yoo, Jiook Cha | Spatiotemporal Learning of Brain Dynamics from fMRI Using Frequency-Specific Multi-Band Attention for Cognitive and Psychiatric Applications | null | null | null | null | q-bio.NC cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Understanding how the brain's complex nonlinear dynamics give rise to cognitive function remains a central challenge in neuroscience. While brain functional dynamics exhibits scale-free and multifractal properties across temporal scales, conventional neuroimaging analytics assume linearity and stationarity, failing to capture frequency-specific neural computations. Here, we introduce Multi-Band Brain Net (MBBN), the first transformer-based framework to explicitly model frequency-specific spatiotemporal brain dynamics from fMRI. MBBN integrates biologically-grounded frequency decomposition with multi-band self-attention mechanisms, enabling discovery of previously undetectable frequency-dependent network interactions. Trained on 49,673 individuals across three large-scale cohorts (UK Biobank, ABCD, ABIDE), MBBN sets a new state-of-the-art in predicting psychiatric and cognitive outcomes (depression, ADHD, ASD), showing particular strength in classification tasks with up to 52.5\% higher AUROC and provides a novel framework for predicting cognitive intelligence scores. Frequency-resolved analyses uncover disorder-specific signatures: in ADHD, high-frequency fronto-sensorimotor connectivity is attenuated and opercular somatosensory nodes emerge as dynamic hubs; in ASD, orbitofrontal-somatosensory circuits show focal high-frequency disruption together with enhanced ultra-low-frequency coupling between the temporo-parietal junction and prefrontal cortex. By integrating scale-aware neural dynamics with deep learning, MBBN delivers more accurate and interpretable biomarkers, opening avenues for precision psychiatry and developmental neuroscience. | [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 10:56:50 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:03:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bae",
"Sangyoon",
""
],
[
"Kwon",
"Junbeom",
""
],
[
"Yoo",
"Shinjae",
""
],
[
"Cha",
"Jiook",
""
]
] |
2503.23458 | Karan Mukhi | Karan Mukhi, Georg Loho, Alessandro Abate | Exact Characterization of Aggregate Flexibility via Generalized Polymatroids | null | null | null | null | eess.SY cs.CE cs.SY | http://creativecommons.org/licenses/by/4.0/ | It is well established that the aggregate flexibility inherent in populations of distributed energy resources (DERs) can be leveraged to mitigate the intermittency and uncertainty associated with renewable generation, while also providing ancillary grid services. To enable this, aggregators must effectively represent the flexibility in the populations they control to the market or system operator. A key challenge is accurately computing the aggregate flexibility of a population, which can be formally expressed as the Minkowski sum of a collection of polytopes, a problem that is generally computationally intractable. However, the flexibility polytopes of many DERs exhibit structural symmetries that can be exploited for computational efficiency. To this end, we introduce generalized polymatroids, a family of polytopes, into the flexibility aggregation literature. We demonstrate that individual flexibility sets belong to this family, enabling efficient computation of their exact Minkowski sum. For homogeneous populations of DERs we further derive simplifications that yield more succinct representations of aggregate flexibility. Additionally, we develop an efficient optimization framework over these sets and propose a vertex-based disaggregation method, to allocate aggregate flexibility among individual DERs. Finally, we validate the optimality and computational efficiency of our approach through comparisons with existing methods. | [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 14:31:14 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 11:27:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mukhi",
"Karan",
""
],
[
"Loho",
"Georg",
""
],
[
"Abate",
"Alessandro",
""
]
] |
2503.24097 | Qiyue Qian | Qiyue Qian, Hongjing Yang, Weicheng Zang, Yoon-Hyun Ryu, Shude Mao, Renkun Kuang, Jiyuan Zhang, Michael D. Albrow, Sun-Ju Chung, Andrew Gould, Cheongho Han, Kyu-Ha Hwang, Youn Kil Jung, In-Gu Shin, Yossi Shvartzvald, Jennifer C. Yee, Sang-Mok Cha, Dong-Jin Kim, Hyoun-Woo Kim, Seung-Lee Kim, Chung-Uk Lee, Dong-Joo Lee, Yongseok Lee, Byeong-Gon Park, Richard W. Pogge | Systematic Search for FFPs in KMTNet Full-Frame Images. I. Photometry Pipeline | Accepted for publication in PASP | null | null | null | astro-ph.EP astro-ph.IM astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To exhume the buried signatures of free-floating planets (FFPs) with small angular Einstein radius $\theta_{\rm E}$, we build a new full-frame difference image pipeline for the Korean Microlensing Telescope Network (KMTNet) survey based on the newly optimized pySIS package. We introduce the detailed processes of the new pipeline, including frame registration, difference image analysis, and light curve extraction. To test this pipeline, we extract 1-year light curves for 483,068 stars with $I \lesssim 17$ and conduct a model-independent search for microlensing events. The search finds 36 microlensing events, including five new events and six events discovered by other collaborations but missed by previous KMTNet searches. We find that the light curves from the new pipeline are precise enough to be sensitive to FFPs with $\theta_{\rm E} \sim 1~\mu$as. Using the new pipeline, a complete FFP search on the eight-year KMTNet images can be finished within six months and then yield the FFP mass function. The new pipeline can be used for a new KMTNet AlertFinder system, with significantly reduced false positives. | [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 13:50:38 GMT"
},
{
"version": "v2",
"created": "Mon, 12 May 2025 07:12:02 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jun 2025 08:38:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Qian",
"Qiyue",
""
],
[
"Yang",
"Hongjing",
""
],
[
"Zang",
"Weicheng",
""
],
[
"Ryu",
"Yoon-Hyun",
""
],
[
"Mao",
"Shude",
""
],
[
"Kuang",
"Renkun",
""
],
[
"Zhang",
"Jiyuan",
""
],
[
"Albrow",
"Michael D.",
""
],
[
"Chung",
"Sun-Ju",
""
],
[
"Gould",
"Andrew",
""
],
[
"Han",
"Cheongho",
""
],
[
"Hwang",
"Kyu-Ha",
""
],
[
"Jung",
"Youn Kil",
""
],
[
"Shin",
"In-Gu",
""
],
[
"Shvartzvald",
"Yossi",
""
],
[
"Yee",
"Jennifer C.",
""
],
[
"Cha",
"Sang-Mok",
""
],
[
"Kim",
"Dong-Jin",
""
],
[
"Kim",
"Hyoun-Woo",
""
],
[
"Kim",
"Seung-Lee",
""
],
[
"Lee",
"Chung-Uk",
""
],
[
"Lee",
"Dong-Joo",
""
],
[
"Lee",
"Yongseok",
""
],
[
"Park",
"Byeong-Gon",
""
],
[
"Pogge",
"Richard W.",
""
]
] |
2503.24341 | Sam Bayliss | Sarah K. Mann, Angus Cowley-Semple, Emma Bryan, Ziqiu Huang, Sandrine Heutz, Max Attwood, Sam L. Bayliss | Chemically Tuning Room Temperature Pulsed Optically Detected Magnetic Resonance | 9 pages, 5 figures | Journal of the American Chemical Society (2025) | 10.1021/jacs.5c05505 | null | quant-ph physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | Optical detection of magnetic resonance enables spin-based quantum sensing with high spatial resolution and sensitivity-even at room temperature-as exemplified by solid-state defects. Molecular systems provide a complementary, chemically tunable, platform for room-temperature optically detected magnetic resonance (ODMR)-based quantum sensing. A critical parameter governing sensing sensitivity is the optical contrast-i.e., the difference in emission between two spin states. In state-of-the-art solid-state defects such as the nitrogen-vacancy center in diamond, this contrast is approximately 30%. Here, capitalizing on chemical tunability, we show that room-temperature ODMR contrasts of 40% can be achieved in molecules. Using a nitrogen-substituted analogue of pentacene (6,13-diazapentacene), we enhance contrast compared to pentacene and, by determining the triplet kinetics through time-dependent pulsed ODMR, show how this arises from accelerated anisotropic intersystem crossing. Furthermore, we translate high-contrast room-temperature pulsed ODMR to self-assembled nanocrystals. Overall, our findings highlight the synthetic handles available to optically readable molecular spins and the opportunities to capitalize on chemical tunability for room-temperature quantum sensing. | [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 17:25:46 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mann",
"Sarah K.",
""
],
[
"Cowley-Semple",
"Angus",
""
],
[
"Bryan",
"Emma",
""
],
[
"Huang",
"Ziqiu",
""
],
[
"Heutz",
"Sandrine",
""
],
[
"Attwood",
"Max",
""
],
[
"Bayliss",
"Sam L.",
""
]
] |
2504.00402 | Jie Lu | Jiaxin Du, Mei Li, Bin Xi, Xue Zhang, Yi Sun, Chun-Gui Duan, and Jie Lu | Fast reversal of N\'{e}el vectors in antiferromagnets via domain-wall motion driven by vertically injected spin currents | 6 pages, 3 figures | Appl. Phys. Lett. 126, 242402 (2025) | 10.1063/5.0273622 | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we investigate the steady propagation of 180-degree domain walls (180DWs) of N\'{e}el vectors in thin antiferromagnetic strips under perpendicularly injected spin currents with various polarization orientations. Our results show that only spin currents polarized normally to the strip plane guarantee a quick and steady rigid flow of 180DWs, thus realize a fast reversal of N\'{e}el vectors in antiferromagnets (AFMs). Different from the common ``current-in-plane'' geometry which is feasible only for metallic AFMs, our ``current-out-of-plane'' layout under investigation can further apply to insulating AFMs (which are more common in real applications) via quantum tunneling effect. Results from this work paves the way for fine control of N\'{e}el vectors in (both metallic and insulating) AFM strips and further development of novel magnetic nanodevices based on them. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 03:52:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Du",
"Jiaxin",
""
],
[
"Li",
"Mei",
""
],
[
"Xi",
"Bin",
""
],
[
"Zhang",
"Xue",
""
],
[
"Sun",
"Yi",
""
],
[
"Duan",
"Chun-Gui",
""
],
[
"Lu",
"Jie",
""
]
] |
2504.00541 | Chiara Mariotti | C. Anastopoulos, R. Assmann, A. Ball, O. Bruning, O. Buchmueller, T. Camporesi, P. Collier, J Dainton, G. Davies, J.R. Ellis, B. Goddard, L. Gouskos, M. Klute, M. Koratzinos, G. Landsberg, K. Long, L. Malgeri, F. Maltoni, F. Moortgat, C. Mariotti, S. Myers, J.A. Osborne, M. Pierini, D.R. Tovey, D. Treille, T.S. Virdee, N. Wardle, M. Zanetti | LEP3: A High-Luminosity e+e- Higgs and ElectroweakFactory in the LHC Tunnel | 11 pages, 3 tables | null | null | null | physics.acc-ph hep-ex hep-ph | http://creativecommons.org/licenses/by/4.0/ | As stated in the 2019 European Strategy for Particle Physics (ESPP), it is of the utmost importance that the HL-LHC upgrade of the accelerator and the experiments be successfully completed in a timely manner. All necessary efforts should be devoted to achieving this goal. We also recall two of the principal recommendations of the 2019 ESPP for future accelerator initiatives, namely that 1) An electron-positron Higgs factory is the highest priority for the next collider (Rec. c). 2) Europe, together with its international partners, should investigate the technical and financial feasibility of a future hadron collider at CERN with a centre-of-mass energy of at least 100 TeV and with an electron-positron Higgs and electroweak factory as a possible first stage (Rec. e). A major objective in particle physics is always to operate an accelerator that allows a leap of an order of magnitude in the constituent centre-of-mass energy with respect to the previous one. We support FCC-ee and FCC-hh as the preferred option for CERN future, as it addresses both of the above recommendations.
The guidance for the 2025 ESPP requests, in addition to the preferred option, the inclusion of ``prioritised alternatives to be pursued if the chosen preferred option turns out not to be feasible or competitive''. Proposed alternatives to the preferred FCC option include linear, muon colliders and LHeC accelerators. In response to this request we propose reusing the existing LHC tunnel for an electron-positron collider, called LEP3, as a back-up alternative if the FCC cannot proceed. LEP3 leverages much of the R\&D conducted for FCC-ee, offers high-precision studies of Z, W, and Higgs bosons below the tt threshold, and offers potential physics performance comparable or superior to other fallback options at a lower cost while supporting continued R\&D towards a next-generation energy frontier machine. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 08:47:19 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 11:15:00 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Anastopoulos",
"C.",
""
],
[
"Assmann",
"R.",
""
],
[
"Ball",
"A.",
""
],
[
"Bruning",
"O.",
""
],
[
"Buchmueller",
"O.",
""
],
[
"Camporesi",
"T.",
""
],
[
"Collier",
"P.",
""
],
[
"Dainton",
"J",
""
],
[
"Davies",
"G.",
""
],
[
"Ellis",
"J. R.",
""
],
[
"Goddard",
"B.",
""
],
[
"Gouskos",
"L.",
""
],
[
"Klute",
"M.",
""
],
[
"Koratzinos",
"M.",
""
],
[
"Landsberg",
"G.",
""
],
[
"Long",
"K.",
""
],
[
"Malgeri",
"L.",
""
],
[
"Maltoni",
"F.",
""
],
[
"Moortgat",
"F.",
""
],
[
"Mariotti",
"C.",
""
],
[
"Myers",
"S.",
""
],
[
"Osborne",
"J. A.",
""
],
[
"Pierini",
"M.",
""
],
[
"Tovey",
"D. R.",
""
],
[
"Treille",
"D.",
""
],
[
"Virdee",
"T. S.",
""
],
[
"Wardle",
"N.",
""
],
[
"Zanetti",
"M.",
""
]
] |
2504.00931 | Indranil Biswas | Indranil Biswas, Michi-aki Inaba, Arata Komyo and Masa-Hiko Saito | Moduli spaces of framed logarithmic and parabolic connections on a Riemann surface | Final version; to appear in the journal "Moduli" | Moduli 2 (2025) e10 | 10.1112/mod.2025.5 | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct moduli spaces of framed logarithmic connections and also moduli spaces of framed parabolic connections. It is shown that these moduli spaces possess a natural algebraic symplectic structure. We also give an upper bound of the transcendence degree of the algebra of regular functions on the moduli space of parabolic connections. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 16:05:08 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Biswas",
"Indranil",
""
],
[
"Inaba",
"Michi-aki",
""
],
[
"Komyo",
"Arata",
""
],
[
"Saito",
"Masa-Hiko",
""
]
] |
2504.01595 | Cristina Ramos Almeida | C. Ramos Almeida, I. Garcia-Bernete, M. Pereira-Santaella, G. Speranza, R. Maiolino, X. Ji, A. Audibert, P. H. Cezar, J. A. Acosta-Pulido, A. Alonso-Herrero, S. Garcia-Burillo, O. Gonzalez-Martin, D. Rigopoulou, C. N. Tadhunter, A. Labiano, N. A. Levenson and F. R. Donnan | JWST MIRI reveals the diversity of nuclear mid-infrared spectra of nearby type-2 quasars | 18 pages, 10 figures. Accepted for publication in A&A | A&A 698, A194 (2025) | 10.1051/0004-6361/202453549 | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Type-2 quasars (QSO2s) are active galactic nuclei (AGN) seen through a significant amount of dust and gas that obscures the central supermassive black hole and the broad line region. Here we present new mid-infrared spectra of the central kiloparsec of five optically-selected QSO2s at redshift z~0.1 obtained with JWST/MIRI/MRS. These QSO2s belong to the QSOFEED sample and they have log Lbol=45.5-46.0 erg/s, global SFRs that place them above the main sequence, and practically identical optical spectral shape and [OIII] luminosity, but their nuclear mid-infrared spectra exhibit an unexpected diversity of both continua and features. They show: 1) 9.7 micron silicate features going from emission (strength of S9.7=0.5) to relatively strong absorption (S9.7=-1.0) and 18 and 23 micron silicates either in emission or flat. In addition, two of the QSO2s show absorption bands of CO, H2O, and aliphatic grains, indicating different levels of nuclear obscuration across the sample. 2) [NeV]/[NeII] ratios ranging from 0.1 to 2.1 and [NeIII]/[NeII] from 1.0 to 3.5, indicating different coronal line and ionizing continuum strengths. 3) Warm molecular gas masses of 1-4x10^7 Msun and warm-to-cold gas mass ratios of 1-2%, with molecular gas excitation likely due to jet-induced shocks in J1430+1339, and to UV heating and/or turbulence in J1509+0434. 4) PAH emission features with equivalent widths ranging from <0.002 to 0.075 micron, from which we measure a larger contribution from neutral molecules (PAH 11.3/6.2=1.3-3.4) and SFRs<3-7 Msun/yr. This unprecedented dataset allowed us to start exploring the role of various AGN and galaxy properties including ionizing continuum, obscuration, electron density, and jet-ISM interactions on some of the spectral differences listed above, but larger samples are now required to fully understand the diversity of QSO2s' nuclear mid-infrared spectra. | [
{
"version": "v1",
"created": "Wed, 02 Apr 2025 11:02:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Almeida",
"C. Ramos",
""
],
[
"Garcia-Bernete",
"I.",
""
],
[
"Pereira-Santaella",
"M.",
""
],
[
"Speranza",
"G.",
""
],
[
"Maiolino",
"R.",
""
],
[
"Ji",
"X.",
""
],
[
"Audibert",
"A.",
""
],
[
"Cezar",
"P. H.",
""
],
[
"Acosta-Pulido",
"J. A.",
""
],
[
"Alonso-Herrero",
"A.",
""
],
[
"Garcia-Burillo",
"S.",
""
],
[
"Gonzalez-Martin",
"O.",
""
],
[
"Rigopoulou",
"D.",
""
],
[
"Tadhunter",
"C. N.",
""
],
[
"Labiano",
"A.",
""
],
[
"Levenson",
"N. A.",
""
],
[
"Donnan",
"F. R.",
""
]
] |
2504.01763 | Rui Zhou | X. Y. Feng, Z. Zhao, J. Luo, Y. Z. Zhou, J. Yang, A. F. Fang, H. T. Yang, H.-J. Gao, R. Zhou and Guo-qing Zheng | Fully-gapped superconductivity with rotational symmetry breaking in pressurized kagome metal CsV$_3$Sb$_5$ | 25 pages, 4 figures, to appear in Nature Communications | Nat. Commun. 16, 3643 (2025) | 10.1038/s41467-025-58941-w | null | cond-mat.supr-con cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | The discovery of the kagome metal CsV$_3$Sb$_5$ has generated significant interest in its complex physical properties, particularly its superconducting behavior under different pressures, though its nature remains debated. Here, we performed low-temperature, high-pressure $^{121/123}$Sb nuclear quadrupole resonance (NQR) measurements to explore the superconducting pairing symmetry in CsV$_3$Sb$_5$. At ambient pressure, we found that the spin-lattice relaxation rate 1/$T_1$ exhibits a kink at $T \sim$ 0.4 $T_\textrm{c}$ within the superconducting state and follows a $T^3$ variation as temperature further decreases. This suggests the presence of two superconducting gaps with line nodes in the smaller one. As pressure increases beyond $P_{\rm c} \sim 1.85$ GPa, where the charge-density wave phase is completely suppressed, 1/$T_1$ shows no Hebel-Slichter peak just below $T_\textrm{c}$, and decreases rapidly, even faster than $T^5$, indicating that the gap is fully opened for pressures above $P_{\rm c}$. In this high pressure region, the angular dependence of the in-plane upper critical magnetic field $H_{\rm c2}$ breaks the $C_6$ rotational symmetry. We propose the $s+id$ pairing at $P > P_{\rm c}$ which explains both the 1/$T_1$ and $H_{\rm c2}$ behaviors. Our findings indicate that CsV$_3$Sb$_5$ is an unconventional superconductor and its superconducting state is even more exotic at high pressures. | [
{
"version": "v1",
"created": "Wed, 02 Apr 2025 14:17:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Feng",
"X. Y.",
""
],
[
"Zhao",
"Z.",
""
],
[
"Luo",
"J.",
""
],
[
"Zhou",
"Y. Z.",
""
],
[
"Yang",
"J.",
""
],
[
"Fang",
"A. F.",
""
],
[
"Yang",
"H. T.",
""
],
[
"Gao",
"H. -J.",
""
],
[
"Zhou",
"R.",
""
],
[
"Zheng",
"Guo-qing",
""
]
] |
2504.01974 | Marcelo Amanajas Pires | Erveton P. Pinto, Marcelo A. Pires, Rone N. da Silva and S\'ilvio M. Duarte Queir\'os | Cryptocurrency Time Series on the Binary Complexity-Entropy Plane: Ranking Efficiency from the Perspective of Complex Systems | 12 pages, 8 figures, 2 tables and 3 appendices | Physica A: Statistical Mechanics and its Applications, 2025 | 10.1016/j.physa.2025.130699 | null | q-fin.ST physics.data-an physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | We report the first application of a tailored Complexity-Entropy Plane designed for binary sequences and structures. We do so by considering the daily up/down price fluctuations of the largest cryptocurrencies in terms of capitalization (stable-coins excluded) that are worth $circa \,\, 90 \%$ of the total crypto market capitalization. With that, we focus on the basic elements of price motion that compare with the random walk backbone features associated with mathematical properties of the Efficient Market Hypothesis. From the location of each crypto on the Binary Complexity-Plane (BiCEP) we define an inefficiency score, $\mathcal I$, and rank them accordingly. The results based on the BiCEP analysis, which we substantiate with statistical testing, indicate that only Shiba Inu (SHIB) is significantly inefficient, whereas the largest stake of crypto trading is reckoned to operate in close-to-efficient conditions. Generically, our $\mathcal I$-based ranking hints the design and consensus architecture of a crypto is at least as relevant to efficiency as the features that are usually taken into account in the appraisal of the efficiency of financial instruments, namely canonical fiat money. Lastly, this set of results supports the validity of the binary complexity analysis. | [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 22:01:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pinto",
"Erveton P.",
""
],
[
"Pires",
"Marcelo A.",
""
],
[
"da Silva",
"Rone N.",
""
],
[
"Queirós",
"Sílvio M. Duarte",
""
]
] |
2504.02054 | Ming Lei | Ming Lei, Iwnetim I. Abate | Grotthuss-type oxygen hole polaron transport in desodiated Na$_{2}$Mn$_3$O$_7$ | null | null | 10.1103/bd8s-74zt | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Polarons are quasiparticles that arise from the coupling of electrons or holes with ionic vibrations in polarizable materials. Typically, they are either localized at a single atomic site or delocalized over multiple sites. However, after the desodiation of Na$_{2}$Mn$_{3}$O$_{7}$, we identify a rare split-hole polaron, where a single hole is shared between two adjacent oxygen atoms rather than fully localized or delocalized. We present a density functional theory (DFT) study on the migration and transport properties of these oxygen hole polarons in NaMn$_{3}$O$_{7}$ and Na$_{1.5}$Mn$_{3}$O$_{7}$. Our calculations reveal that the split polaron configuration near a sodium vacancy is the ground state, while the localized polaron acts as the transition state. Migration occurs via a stepwise charge transfer mechanism along the $b$-axis, where the split-hole polaron transitions through a localized hole state. This transport behavior closely resembles the Grotthuss mechanism, which describes proton transport in H$_{2}$O. We compute the polaron mobility as $\mu$ = 1.37 $\times$ 10$^{-5}$ cm$^2$/(V$\cdot$s) with an energy barrier of 242 meV. Using the Mulliken-Hush theory, we determine the electronic coupling parameter $V_{AB}$ = 0.87 eV. A similar migration mechanism is observed in Na$_{1.5}$Mn$_{3}$O$_{7}$, where the split polaron remains more stable than in the localized state. This study provides the first theoretical investigation of split-hole polaron migration, offering new insights into the charge transport of exotic polaronic species in materials with implications for a wide range of functional materials including battery cathodes, thermoelectrics, photocatalysts, and next-generation optoelectronic devices. | [
{
"version": "v1",
"created": "Wed, 02 Apr 2025 18:31:07 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jun 2025 21:36:28 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lei",
"Ming",
""
],
[
"Abate",
"Iwnetim I.",
""
]
] |
2504.02689 | Davy Gerard | J\'er\^ome Martin, Oscar Avalos-Ovando, Thomas Simon, Gabriel Arditi, Florian Lamaze, Julien Proust, Luiz H.G. Tizei, Zhiming Wang, Mathieu Kociak, Alexander O. Govorov, Odile St\'ephan, and Davy G\'erard | Plasmon-interband hybridization and anomalous production of hot electrons in aluminum nanoantennas | Authors' version of the published paper. 8 figures, 1 Supplementary Information | null | 10.1021/acsnano.5c05690 | null | physics.optics cond-mat.mes-hall | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Strong coupling typically occurs between two separate objects or between an object and its environment (such as an atom and a cavity). However, it can also occur between two different excitations within the same object, a situation that has been much less studied. In this study, we observe strong coupling between localized surface plasmon resonances and the interband transition in aluminum nanorods, as evidenced by optical spectroscopy and electron energy loss spectroscopy, and corroborated with numerical simulations. Strong coupling is observed between the interband transition and multiple orders of the surface plasmon mode, including dark ones. We also obtain experimental maps of the hybrid modes at the nanoscale. In each case, the associated Rabi energy, which corresponds to the energy splitting between the two polaritonic branches, is obtained. Finally, a dedicated numerical model was employed to calculate the hot electron generation rate in the nanorods. The calculations demonstrate that efficient generation of hot electrons can be achieved in the near-infrared region, when the interband transition is strongly coupled with a plasmon resonance. This high generation rate stems from the hybrid nature of the mode, as its plasmonic component provides a high absorption cross-section, while the IT part ensures efficient conversion to hot electrons. Consequently, aluminum nanorods represent an efficient source of hot electrons in the visible and near-infrared regions, with potential applications in local photochemistry, photodetection, and solar energy harvesting. | [
{
"version": "v1",
"created": "Thu, 03 Apr 2025 15:29:07 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 16:13:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Martin",
"Jérôme",
""
],
[
"Avalos-Ovando",
"Oscar",
""
],
[
"Simon",
"Thomas",
""
],
[
"Arditi",
"Gabriel",
""
],
[
"Lamaze",
"Florian",
""
],
[
"Proust",
"Julien",
""
],
[
"Tizei",
"Luiz H. G.",
""
],
[
"Wang",
"Zhiming",
""
],
[
"Kociak",
"Mathieu",
""
],
[
"Govorov",
"Alexander O.",
""
],
[
"Stéphan",
"Odile",
""
],
[
"Gérard",
"Davy",
""
]
] |
2504.02751 | Zhan Wang | Zhan Wang, Keyu Zeng, and Ziqiang Wang | Roton Superconductivity from Loop-Current Chern Metal on the Kagome Lattice | 18+4 pages, 11 figures | null | null | null | cond-mat.str-el cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Motivated by the evidence for time-reversal symmetry (TRS) breaking in nonmagnetic kagome metals AV3Sb5, a novel electronic order of persistent orbital loop-current (LC) has been proposed for the observed charge density wave (CDW) state. The LC order and its impact on the succeeding superconducting (SC) state are central to the new physics of the kagome materials. We show that the LC order fundamentally changes the pairing instability and the SC state, leading to an extraordinary topological superconductor, dubbed as a roton superconductor. In the single-orbital model on the kagome lattice, the LC-CDW state is a Chern metal near van Hove filling with a partially filled Chern band hosting 3 Chern Fermi pockets (CFPs). Cooper pairing of quasiparticles on the CFPs generates 3 SC components coupled by complex Josephson couplings induced by the TRS breaking LC. Due to the discrete quantum geometry associated with the 3-fold rotation and sublattice permutation, a small LC produces a large Josephson phase that drives the leading SC instability to the roton superconductor where the relative phases of the 3 SC components are locked at 120{\deg}, forming an emergent vortex-antivortex lattice with pair density modulations. Properties of the roton superconductor include topological chiral edge states carrying nonzero electric currents, fractional 1/3 vortex excitations and charge-6e Cooper pair triplets that are immune to the internal chiral phase fluctuations. We discuss these SC properties in connection to recent experimental evidence for TRS breaking chiral SC state in kagome superconductors, exhibiting pair density modulations, zero-field SC diode effect, and charge-6e flux quantization. These findings are also relevant for the interplay between the orbital-driven quantum anomalous Hall and SC states in other systems, e.g. the graphene and transition metal dichalcogenide based quantum materials. | [
{
"version": "v1",
"created": "Thu, 03 Apr 2025 16:43:02 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:22:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Zhan",
""
],
[
"Zeng",
"Keyu",
""
],
[
"Wang",
"Ziqiang",
""
]
] |
2504.03255 | Rui Patrick Xian | Garry A. Gabison, R. Patrick Xian | Inherent and emergent liability issues in LLM-based agentic systems: a principal-agent perspective | 22 pages (incl. appendix), accepted at REALM workshop, ACL2025 | null | null | null | cs.CY cs.CL cs.MA | http://creativecommons.org/licenses/by-sa/4.0/ | Agentic systems powered by large language models (LLMs) are becoming progressively more complex and capable. Their increasing agency and expanding deployment settings attract growing attention to effective governance policies, monitoring, and control protocols. Based on the emerging landscape of the agentic market, we analyze potential liability issues arising from the delegated use of LLM agents and their extended systems through a principal-agent perspective. Our analysis complements existing risk-based studies on artificial agency and covers the spectrum of important aspects of the principal-agent relationship and their potential consequences at deployment. Furthermore, we motivate method developments for technical governance along the directions of interpretability and behavior evaluations, reward and conflict management, and the mitigation of misalignment and misconduct through principled engineering of detection and fail-safe mechanisms. By illustrating the outstanding issues in AI liability for LLM-based agentic systems, we aim to inform the system design, auditing, and tracing to enhance transparency and liability attribution. | [
{
"version": "v1",
"created": "Fri, 04 Apr 2025 08:10:02 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 13:42:52 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Gabison",
"Garry A.",
""
],
[
"Xian",
"R. Patrick",
""
]
] |
2504.06049 | Abhishek Sarkar | Sutirtha Datta, Navnath Daundkar and Abhishek Sarkar | Directed LS category and directed parametrized topological complexity | 16 pages, We have added a new section on computations | null | null | null | math.AT | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We introduce and study a parametrized analogue of the directed topological complexity, originally developed by Goubault, Farber, and Sagnier. We establish the fibrewise basic dihomotopy invariance of directed parametrized topological complexity and explore its relationship with the parametrized topological complexity. In addition, we introduce the concept of the directed Lusternik$-$Schnirelmann (LS) category, prove its basic dihomotopy invariance, and investigate its connections with both directed topological complexity and directed parametrized topological complexity. As an application, we show that the directed LS category of the directed $n$-sphere for $n\geq 1$ is equal to two. Moreover, we compute the directed parametrized topological complexity of the Hopf fibrations and the Fadell-Neuwirth fibrations having specific directed fibration structures. | [
{
"version": "v1",
"created": "Tue, 08 Apr 2025 13:47:09 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:04:57 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Datta",
"Sutirtha",
""
],
[
"Daundkar",
"Navnath",
""
],
[
"Sarkar",
"Abhishek",
""
]
] |
2504.06937 | Qi-Yue Yu | Qi-yue Yu | Finite Field Multiple Access III: from 2-ary to p-ary | 50 pages, 5 figures | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper extends finite-field multiple-access (FFMA) techniques from binary to general $p$-ary source transmission. We introduce element-assemblage (EA) codes over GF($p^m$), which generalize element-pair (EP) codes, and define two specific types for ternary transmission: orthogonal EA codes and double codeword EA (D-CWEA) codes. We propose a unique sum-pattern mapping (USPM) constraint for the design of uniquely-decodable CWEA (UD-CWEA) codes, which include additive inverse D-CWEA (AI-D-CWEA) and basis decomposition D-CWEA (BD-D-CWEA) codes. Additionally, we introduce non-orthogonal CWEA (NO-CWEA) codes and their corresponding USPM constraint in the complex field. Furthermore, $p$-ary CWEA codes are constructed using a basis decomposition method, leveraging ternary decomposition for faster convergence and simplified encoder/decoder design. We present a performance analysis of the proposed FFMA system from two complementary perspectives: channel capacity and error performance. We demonstrate that equal power allocation achieves the theoretical channel capacity, and then investigate the finite blocklength (FBL) characteristics of FFMA systems. Moreover, we develop a rate-driven capacity alignment (CA) theorem based on the capacity-to-rate ratio (CRR) metric for error performance analysis. Finally, we compare $p$-ary transmission systems with classical binary transmission systems, revealing that low-order $p$-ary systems (e.g., $p = 3$) outperform binary systems at small loading factors, while higher-order systems (e.g., $p = 257$) excel at larger loading factors. These findings highlight the potential of $p$-ary systems, although practical implementations may benefit from decomposing $p$-ary systems into ternary systems to manage complexity. | [
{
"version": "v1",
"created": "Wed, 09 Apr 2025 14:41:36 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:04:51 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yu",
"Qi-yue",
""
]
] |
2504.06948 | Dekuan Dong | Dekuan Dong, Yingzhou Li, and Jungong Xue | A quantum algorithm for linear autonomous differential equations via Pad\'e approximation | 51 pages, 15 figures | Quantum 9, 1770 (2025) | 10.22331/q-2025-06-17-1770 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We propose a novel quantum algorithm for solving linear autonomous ordinary differential equations (ODEs) using the Pad\'e approximation. For linear autonomous ODEs, the discretized solution can be represented by a product of matrix exponentials. The proposed algorithm approximates the matrix exponential by the diagonal Pad\'e approximation, which is then encoded into a large, block-sparse linear system and solved via quantum linear system algorithms (QLSA). The detailed quantum circuit is given based on quantum oracle access to the matrix, the inhomogeneous term, and the initial state. The complexity of the proposed algorithm is analyzed. Compared to the method based on Taylor approximation, which approximates the matrix exponential using a $k$-th order Taylor series, the proposed algorithm improves the approximation order $k$ from two perspectives: 1) the explicit complexity dependency on $k$ is improved, and 2) a smaller $k$ suffices for the same precision. Numerical experiments demonstrate the advantages of the proposed algorithm comparing to other related algorithms. | [
{
"version": "v1",
"created": "Wed, 09 Apr 2025 14:54:27 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 03:24:32 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Jun 2025 10:48:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dong",
"Dekuan",
""
],
[
"Li",
"Yingzhou",
""
],
[
"Xue",
"Jungong",
""
]
] |
2504.07497 | K. Splittorff | J. Agerskov and K. Splittorff | Quantum Determinant Estimation | 13 pages, 3 figures. Added references and discussion of runtime. Version accepted for publication | null | null | null | quant-ph hep-lat | http://creativecommons.org/licenses/by/4.0/ | A quantum algorithm for computing the determinant of a unitary matrix $U\in U(N)$ is given. The algorithm requires no preparation of eigenstates of $U$ and estimates the phase of the determinant to $t$ binary digits accuracy with $\mathcal{O}(N\log^2 N+t^2)$ operations and $tN$ controlled applications of $U^{2^m}$ with $m=0,\ldots,t-1$. For an orthogonal matrix $O\in O(N)$ the algorithm can determine with certainty the sign of the determinant using $\mathcal{O}(N\log^2 N)$ operations and $N$ controlled applications of $O$. An extension of the algorithm to contractions is discussed. | [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 06:53:37 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:52:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Agerskov",
"J.",
""
],
[
"Splittorff",
"K.",
""
]
] |
2504.07738 | Andrea Loreti Dr | Andrea Loreti, Kesi Chen, Ruby George, Robert Firth, Adriano Agnello and Shinnosuke Tanaka | Automated Construction of a Knowledge Graph of Nuclear Fusion Energy for Effective Elicitation and Retrieval of Information | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this document, we discuss a multi-step approach to automated construction of a knowledge graph, for structuring and representing domain-specific knowledge from large document corpora. We apply our method to build the first knowledge graph of nuclear fusion energy, a highly specialized field characterized by vast scope and heterogeneity. This is an ideal benchmark to test the key features of our pipeline, including automatic named entity recognition and entity resolution. We show how pre-trained large language models can be used to address these challenges and we evaluate their performance against Zipf's law, which characterizes human-generated natural language. Additionally, we develop a knowledge-graph retrieval-augmented generation system that combines large language models with a multi-prompt approach. This system provides contextually relevant answers to natural-language queries, including complex multi-hop questions that require reasoning across interconnected entities. | [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 13:29:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 10:16:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Loreti",
"Andrea",
""
],
[
"Chen",
"Kesi",
""
],
[
"George",
"Ruby",
""
],
[
"Firth",
"Robert",
""
],
[
"Agnello",
"Adriano",
""
],
[
"Tanaka",
"Shinnosuke",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.