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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2411.12894 | Stanley S. Coelho | Stanley S. Coelho, Lucas Queiroz, Danilo T. Alves | The time-dependent quantum harmonic oscillator: a pedagogical approach via the Lewis-Riesenfeld dynamical invariant method | 24 pages, 4 figures | null | 10.1088/1361-6404/add972 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | In quantum mechanics courses, students often solve the Schr\"odinger equation for the harmonic oscillator with time-independent parameters. However, time-dependent quantum harmonic oscillators are relevant in modeling several problems as, for instance, the description of quantum motion of particles in traps, shortcuts to adiabaticity, generation of squeezed states, as well as quantum scalar fields evolving in expanding universes. In the present paper, we discuss, with a pedagogical approach, the quantum harmonic oscillator with time-dependent frequency via the Lewis-Riesenfeld dynamical invariant method, revisiting the main steps to obtain the wave function associated with this model, and briefly discussing the relation between this oscillator and the generation of squeezed states. As examples of didactic applications of time-dependent harmonic oscillators and the Lewis-Riesenfeld method in quantum mechanics courses, we solve the following problems: the calculation of the transition probability associated with a harmonic oscillator which undergoes jumps in its frequency, and the analysis of the dynamics of a quantum particle in a Paul trap. | [
{
"version": "v1",
"created": "Tue, 19 Nov 2024 22:19:32 GMT"
},
{
"version": "v2",
"created": "Mon, 02 Dec 2024 19:59:01 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 21:45:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Coelho",
"Stanley S.",
""
],
[
"Queiroz",
"Lucas",
""
],
[
"Alves",
"Danilo T.",
""
]
] |
2411.12977 | Mircea Lic\u{a} | Mircea Lic\u{a}, Ojas Shirekar, Baptiste Colle, Chirag Raman | MindForge: Empowering Embodied Agents with Theory of Mind for Lifelong Cultural Learning | null | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Embodied agents powered by large language models (LLMs), such as Voyager, promise open-ended competence in worlds such as Minecraft. However, when powered by open-weight LLMs they still falter on elementary tasks after domain-specific fine-tuning. We propose MindForge, a generative-agent framework for cultural lifelong learning through explicit perspective taking. We introduce three key innovations: (1) a structured theory of mind representation linking percepts, beliefs, desires, and actions; (2) natural inter-agent communication; and (3) a multi-component memory system. Following the cultural learning framework, we test MindForge in both instructive and collaborative settings within Minecraft. In an instructive setting with GPT-4, MindForge agents powered by open-weight LLMs significantly outperform their Voyager counterparts in basic tasks yielding $3\times$ more tech-tree milestones and collecting $2.3\times$ more unique items than the Voyager baseline. Furthermore, in fully \textit{collaborative} settings, we find that the performance of two underachieving agents improves with more communication rounds, echoing the Condorcet Jury Theorem. MindForge agents demonstrate sophisticated behaviors, including expert-novice knowledge transfer, collaborative problem solving, and adaptation to out-of-distribution tasks through accumulated cultural experiences. | [
{
"version": "v1",
"created": "Wed, 20 Nov 2024 02:10:44 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Nov 2024 13:17:01 GMT"
},
{
"version": "v3",
"created": "Wed, 19 Feb 2025 22:59:28 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 14:09:04 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lică",
"Mircea",
""
],
[
"Shirekar",
"Ojas",
""
],
[
"Colle",
"Baptiste",
""
],
[
"Raman",
"Chirag",
""
]
] |
2411.13464 | Juan Antonio Aguilar-Saavedra | J. A. Aguilar-Saavedra | $H \to ZZ$ as a double-slit experiment | LaTeX 6 pages. Discussion on interference clarified, final version published in PLB | null | null | null | hep-ph hep-ex quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The decay $H \to ZZ \to 4\ell$, with $\ell = e,\mu$, can be used to test quantum interference in analogy to the famous double-slit experiment. The observations corresponding to `covering a slit' can be extracted from data in the $ee\mu\mu$ channel, while the observations when both `slits' are open, which include the interference, are measured in the $4e/4\mu$ channels. In this way, this process offers a unique opportunity to investigate identical-particle effects at high energies, and provide the first evidence of identical-particle behaviour for muons. Sensitivity at the $4\sigma$ level could be achieved at the high-luminosity upgrade of the Large Hadron Collider. | [
{
"version": "v1",
"created": "Wed, 20 Nov 2024 17:12:40 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:24:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Aguilar-Saavedra",
"J. A.",
""
]
] |
2411.13610 | Hao Ju | Hao Ju, Shaofei Huang, Si Liu, Zhedong Zheng | Video2BEV: Transforming Drone Videos to BEVs for Video-based Geo-localization | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing approaches to drone visual geo-localization predominantly adopt the image-based setting, where a single drone-view snapshot is matched with images from other platforms. Such task formulation, however, underutilizes the inherent video output of the drone and is sensitive to occlusions and viewpoint disparity. To address these limitations, we formulate a new video-based drone geo-localization task and propose the Video2BEV paradigm. This paradigm transforms the video into a Bird's Eye View (BEV), simplifying the subsequent \textbf{inter-platform} matching process. In particular, we employ Gaussian Splatting to reconstruct a 3D scene and obtain the BEV projection. Different from the existing transform methods, \eg, polar transform, our BEVs preserve more fine-grained details without significant distortion. To facilitate the discriminative \textbf{intra-platform} representation learning, our Video2BEV paradigm also incorporates a diffusion-based module for generating hard negative samples. To validate our approach, we introduce UniV, a new video-based geo-localization dataset that extends the image-based University-1652 dataset. UniV features flight paths at $30^\circ$ and $45^\circ$ elevation angles with increased frame rates of up to 10 frames per second (FPS). Extensive experiments on the UniV dataset show that our Video2BEV paradigm achieves competitive recall rates and outperforms conventional video-based methods. Compared to other competitive methods, our proposed approach exhibits robustness at lower elevations with more occlusions. | [
{
"version": "v1",
"created": "Wed, 20 Nov 2024 01:52:49 GMT"
},
{
"version": "v2",
"created": "Sat, 08 Mar 2025 11:49:58 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 14:51:01 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ju",
"Hao",
""
],
[
"Huang",
"Shaofei",
""
],
[
"Liu",
"Si",
""
],
[
"Zheng",
"Zhedong",
""
]
] |
2411.14013 | Mat\'ias Pizarro | Mat\'ias Pizarro, Mike Laszkiewicz, Shawkat Hesso, Dorothea Kolossa, Asja Fischer | Model Attribution and Detection of Synthetic Speech via Vocoder Fingerprints | null | null | null | null | eess.AS cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | As speech generation technology advances, so do the potential threats of misusing synthetic speech signals. This work tackles three tasks: (1) single-model attribution in an open-world setting corresponding to the task of identifying whether synthetic speech signals originate from a specific vocoder (which requires only target vocoder data), (2) model attribution in a closed-world setting that corresponds to selecting the specific model that generated a sample from a given set of models, and (3) distinguishing synthetic from real speech. We show that standardized average residuals between audio signals and their low-pass or EnCodec filtered versions serve as powerful vocoder fingerprints that can be leveraged for all tasks achieving an average AUROC of over 99% on LJSpeech and JSUT in most settings. The accompanying robustness study shows that it is also resilient to noise levels up to a certain degree. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 10:55:49 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 11:45:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pizarro",
"Matías",
""
],
[
"Laszkiewicz",
"Mike",
""
],
[
"Hesso",
"Shawkat",
""
],
[
"Kolossa",
"Dorothea",
""
],
[
"Fischer",
"Asja",
""
]
] |
2411.14312 | Leon Staresinic | Kostiantyn Drach, Leon Staresinic, Sebastian van Strien | Density of Stable Interval Translation Maps | null | null | null | null | math.DS | http://creativecommons.org/licenses/by/4.0/ | Assume that the interval $I=[0,1)$ is partitioned into finitely many intervals $I_1,\dots,I_r$ and consider a map $T\colon I\to I$ so that $T_{\vert I_s}$ is a translation for each $1 \le s \le r$. We do not assume that the images of these intervals are disjoint. Such maps are called Interval Translation Maps. Let $ITM(r)$ be the space of all such transformations, where we fix $r$ but not the intervals $I_1,\dots,I_r$, nor the translations. The set $X(T):=\bigcap_{n\ge 0} T^n[0,1)$ can be a finite union of intervals (in which case the map is called of finite type), or is a disjoint union of finitely many intervals and a Cantor set (in which case the map is called of infinite type). In this paper we show that there exists an open and dense subset $\mathcal{S}(r)$ of $ITM(r)$ consisting of stable maps, i.e. each $T\in \mathcal{S}(r)$ is of finite type, the first return map to any component of $X(T)$ corresponds to a circle rotation and $\mathcal{S}(r) \ni T \mapsto X(T)$ is continuous in the Hausdorff topology. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 17:04:29 GMT"
},
{
"version": "v2",
"created": "Fri, 22 Nov 2024 11:19:04 GMT"
},
{
"version": "v3",
"created": "Mon, 25 Nov 2024 11:45:06 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 15:45:48 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Drach",
"Kostiantyn",
""
],
[
"Staresinic",
"Leon",
""
],
[
"van Strien",
"Sebastian",
""
]
] |
2411.15317 | Daniel Luo | Daniel Luo and Alexander Wolitzky | Marginal Reputation | null | null | null | null | econ.TH | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We study reputation formation where a long-run player repeatedly observes private signals and takes actions. Short-run players observe the long-run player's past actions but not her past signals. The long-run player can thus develop a reputation for playing a distribution over actions, but not necessarily for playing a particular mapping from signals to actions. Nonetheless, we show that the long-run player can secure her Stackelberg payoff if distinct commitment types are statistically distinguishable and the Stackelberg strategy is confound-defeating. This property holds if and only if the Stackelberg strategy is the unique solution to an optimal transport problem. If the long-run player's payoff is supermodular in one-dimensional signals and actions, she secures the Stackelberg payoff if and only if the Stackelberg strategy is monotone. Applications include deterrence, delegation, signaling, and persuasion. Our results extend to the case where distinct commitment types may be indistinguishable but the Stackelberg type is salient under the prior. | [
{
"version": "v1",
"created": "Fri, 22 Nov 2024 19:16:51 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Dec 2024 21:57:28 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 03:50:42 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Luo",
"Daniel",
""
],
[
"Wolitzky",
"Alexander",
""
]
] |
2411.15733 | Tomotada Akutsu | Tomotada Akutsu and Hiroaki Yamamoto | Modeling of optical scattering from topographic surface measurements of high-quality mirrors | 13 pages, 5 figures | Applied Optics 64, 1995 (2025) | 10.1364/AO.550111 | JGW-P2416080 | physics.optics astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we revisit computational methods to obtain an angular profile of optical scattering from a smooth surface, given a two-dimensional map of topographic height errors of the surface. Quick derivations of some traditional equations and relevant references are organized to shorten the search time. A practical data-processing flow of the methods is discussed. As a case study of this flow, the core mirrors of the KAGRA interferometer are examined, and we obtain a representative scattering profile that is easily applicable to ray-tracing simulations. | [
{
"version": "v1",
"created": "Sun, 24 Nov 2024 06:30:18 GMT"
},
{
"version": "v2",
"created": "Sun, 09 Feb 2025 04:50:54 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 04:26:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Akutsu",
"Tomotada",
""
],
[
"Yamamoto",
"Hiroaki",
""
]
] |
2411.16463 | Gang Xiang | Xia Chen, Xi Zhang, Wenjie He, Yu Li, Jiating Lu, Dinghua Yang, Deren Li, Li Lei, Yong Peng and Gang Xiang | Lattice dynamics and phonon dispersion of van der Waals layered ferromagnet Fe3GaTe2 | null | Nano Letters, 25(11), 4353-4360(2025) | 10.1021/acs.nanolett.4c06408 | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/publicdomain/zero/1.0/ | Despite the tremendous progress in spintronic studies of van der Waals (vdW) room-temperature ferromagnet Fe3GaTe2, much less effort has been spent on its lattice dynamics and possible interaction with spintronic degrees of freedom. In this work, by combining Raman spectroscopy in a wide range of pressure (atmospheric pressure~19.5 GPa) and temperature (80~690 K) with first-principles calculation, we systematically studied the lattice dynamics and phonon dispersion of Fe3GaTe2. Our results show that the phonon energies of Fe3GaTe2 located at 126.0 cm-1 and 143.5 cm-1 originate from the anharmonic and harmonic vibration modes, respectively. Furthermore, the first room-temperature spin-phonon coupling in vdW ferromagnet is observed with strength of ~0.81 cm-1 at 300 K, by identifying Raman anomalies in both phonon energy and full width at half maximum (FWHM) of below Curie temperature of Fe3GaTe2. Our findings are valuable for fundamental and applied studies of vdW materials under variable conditions. | [
{
"version": "v1",
"created": "Mon, 25 Nov 2024 15:09:38 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Mar 2025 07:09:03 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Xia",
""
],
[
"Zhang",
"Xi",
""
],
[
"He",
"Wenjie",
""
],
[
"Li",
"Yu",
""
],
[
"Lu",
"Jiating",
""
],
[
"Yang",
"Dinghua",
""
],
[
"Li",
"Deren",
""
],
[
"Lei",
"Li",
""
],
[
"Peng",
"Yong",
""
],
[
"Xiang",
"Gang",
""
]
] |
2411.17304 | Tomas Kliegr | Milena Chadimov\'a, Eduard Jur\'a\v{s}ek, Tom\'a\v{s} Kliegr | Meaningless is better: hashing bias-inducing words in LLM prompts improves performance in logical reasoning and statistical learning | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a novel method, referred to as "hashing", which involves masking potentially bias-inducing words in large language models (LLMs) with hash-like meaningless identifiers to reduce cognitive biases and reliance on external knowledge. The method was tested across three sets of experiments involving a total of 490 prompts. Statistical analysis using chi-square tests showed significant improvements in all tested scenarios, which covered LLama, ChatGPT, Copilot, Gemini and Mixtral models. In the first experiment, hashing decreased the fallacy rate in a modified version of the "Linda" problem aimed at evaluating susceptibility to cognitive biases. In the second experiment, it improved LLM results on the frequent itemset extraction task. In the third experiment, we found hashing is also effective when the Linda problem is presented in a tabular format rather than text, indicating that the technique works across various input representations. Overall, the method was shown to improve bias reduction and incorporation of external knowledge. Despite bias reduction, hallucination rates were inconsistently reduced across types of LLM models. These findings suggest that masking bias-inducing terms can improve LLM performance, although its effectiveness is model- and task-dependent. | [
{
"version": "v1",
"created": "Tue, 26 Nov 2024 10:52:08 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:55:38 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chadimová",
"Milena",
""
],
[
"Jurášek",
"Eduard",
""
],
[
"Kliegr",
"Tomáš",
""
]
] |
2411.17453 | Zhen Sun | Zhen Sun, Tianshuo Cong, Yule Liu, Chenhao Lin, Xinlei He, Rongmao Chen, Xingshuo Han, Xinyi Huang | PEFTGuard: Detecting Backdoor Attacks Against Parameter-Efficient Fine-Tuning | 21 pages, 7 figures | 2025 IEEE Symposium on Security and Privacy (SP) | 10.1109/SP61157.2025.00161 | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Fine-tuning is an essential process to improve the performance of Large Language Models (LLMs) in specific domains, with Parameter-Efficient Fine-Tuning (PEFT) gaining popularity due to its capacity to reduce computational demands through the integration of low-rank adapters. These lightweight adapters, such as LoRA, can be shared and utilized on open-source platforms. However, adversaries could exploit this mechanism to inject backdoors into these adapters, resulting in malicious behaviors like incorrect or harmful outputs, which pose serious security risks to the community. Unfortunately, few current efforts concentrate on analyzing the backdoor patterns or detecting the backdoors in the adapters. To fill this gap, we first construct and release PADBench, a comprehensive benchmark that contains 13,300 benign and backdoored adapters fine-tuned with various datasets, attack strategies, PEFT methods, and LLMs. Moreover, we propose PEFTGuard, the first backdoor detection framework against PEFT-based adapters. Extensive evaluation upon PADBench shows that PEFTGuard outperforms existing detection methods, achieving nearly perfect detection accuracy (100%) in most cases. Notably, PEFTGuard exhibits zero-shot transferability on three aspects, including different attacks, PEFT methods, and adapter ranks. In addition, we consider various adaptive attacks to demonstrate the high robustness of PEFTGuard. We further explore several possible backdoor mitigation defenses, finding fine-mixing to be the most effective method. We envision that our benchmark and method can shed light on future LLM backdoor detection research. | [
{
"version": "v1",
"created": "Tue, 26 Nov 2024 14:12:09 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 18:07:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Sun",
"Zhen",
""
],
[
"Cong",
"Tianshuo",
""
],
[
"Liu",
"Yule",
""
],
[
"Lin",
"Chenhao",
""
],
[
"He",
"Xinlei",
""
],
[
"Chen",
"Rongmao",
""
],
[
"Han",
"Xingshuo",
""
],
[
"Huang",
"Xinyi",
""
]
] |
2411.17694 | Christopher Roberts | Christopher David Roberts, Frederic Vitart | Ensemble reliability and the signal-to-noise paradox in large-ensemble subseasonal forecasts | null | null | null | null | physics.ao-ph | http://creativecommons.org/licenses/by/4.0/ | Recent studies have illustrated the existence of a `signal-to-noise paradox' (SNP) in some ensemble forecasting systems that manifests as situations where the correlation between the forecast ensemble mean and the observed truth is larger than the correlation between the forecast ensemble mean and individual forecast members. A well-calibrated forecast system that simultaneously satisfies climatological and ensemble variance reliability criteria will not exhibit an SNP if sample statistics can be evaluated using a sufficiently large ensemble size ($N$) over a sufficiently large number of independent cases ($M$). However, when $M$ is finite, an apparent SNP will sometimes occur as a natural consequence of sampling uncertainty, even in a perfectly reliable ensemble with many members. In this study, we evaluate the forecast skill, reliability characteristics, and signal-to-noise properties of three large-scale atmospheric circulation indices in 100-member subseasonal reforecasts with the ECMWF IFS. Daily mean NAO forecasts generally satisfy unbiased reliability criteria within the tolerance of our estimated sampling uncertainties. Nevertheless, NAO forecasts in this dataset exhibit symptoms of the SNP at subseasonal lead times. However, we do not find robust evidence for an underestimation of the magnitude of predictable signals and do not exclude the possibility that the apparent paradox in this dataset is a consequence of observational sampling uncertainties that are insensitive to ensemble size and common to all comparisons for this set of forecast start dates and lead times. Furthermore, we demonstrate that this apparent SNP can be eliminated with an unbiased reliability calibration. However, this is achieved through overfitting such that calibrated forecasts inherit the large sampling uncertainties present in the observations and thus exhibit unphysical variations with lead time. | [
{
"version": "v1",
"created": "Tue, 26 Nov 2024 18:58:55 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 15:59:06 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 12:51:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Roberts",
"Christopher David",
""
],
[
"Vitart",
"Frederic",
""
]
] |
2411.18142 | Yao-Xiang Ding | Jingming Liu, Yumeng Li, Boyuan Xiao, Yichang Jian, Ziang Qin, Tianjia Shao, Yao-Xiang Ding, Kun Zhou | Autonomous Imagination: Closed-Loop Decomposition of Visual-to-Textual Conversion in Visual Reasoning for Multimodal Large Language Models | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Under pure textual modality, Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning tasks by decomposing them into simpler sub-problems. However, Multimodal Large Language Models (MLLMs) still struggle with some seemingly straightforward visual tasks, such as counting and solving jigsaw puzzles. We argue that these tasks challenge the ability of visual-to-textual conversion, where MLLMs convert visual information perceived from the input scene, to textual information for further reasoning and generating the answer. If the complexity of the visual input is beyond the perceptual capability of the MLLMs, without decomposing this conversion process, simply scaling inference-time reasoning cannot solve the task because it repeatedly encounters the same perceptual bottleneck. We propose an approach, autonomous imagination, to enable MLLMs to iteratively modify visual inputs (e.g. isolating objects, rearranging puzzle pieces) into intermediate visual states, decomposing visual-to-textual conversion into closed-loop visual modification steps. We show that, without any retraining, MLLMs can now solve tasks initially beyond their perceptual capability, highlighting that closed-loop visual modification can be an effective way of decomposing the visual reasoning task into solvable substeps. Project page: https://future-item.github.io/autoimagine-site/ | [
{
"version": "v1",
"created": "Wed, 27 Nov 2024 08:44:25 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 11:39:20 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 02:49:11 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Liu",
"Jingming",
""
],
[
"Li",
"Yumeng",
""
],
[
"Xiao",
"Boyuan",
""
],
[
"Jian",
"Yichang",
""
],
[
"Qin",
"Ziang",
""
],
[
"Shao",
"Tianjia",
""
],
[
"Ding",
"Yao-Xiang",
""
],
[
"Zhou",
"Kun",
""
]
] |
2411.18553 | Darius Feher | Darius Feher, Ivan Vuli\'c, Benjamin Minixhofer | Retrofitting Large Language Models with Dynamic Tokenization | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Current language models (LMs) use a fixed, static subword tokenizer. This default choice typically results in degraded efficiency and language capabilities, especially in languages other than English. To address this issue, we challenge the static design and propose retrofitting LMs with dynamic tokenization: a way to dynamically decide on token boundaries based on the input text via a subword-merging algorithm inspired by byte-pair encoding. We merge frequent subword sequences in a batch, then apply a pre-trained embedding-prediction hypernetwork to compute the token embeddings on-the-fly. For encoder-style models (e.g., XLM-R), this on average reduces token sequence lengths by >20% across 14 languages while degrading performance by less than 2%. The same method applied to pre-filling and scoring in decoder-style models (e.g., Mistral-7B) results in minimal performance degradation at up to 17% reduction in sequence length. Overall, we find that dynamic tokenization can mitigate the limitations of static tokenization by substantially improving inference speed and promoting fairness across languages, enabling more equitable and adaptable LMs. | [
{
"version": "v1",
"created": "Wed, 27 Nov 2024 17:51:58 GMT"
},
{
"version": "v2",
"created": "Sat, 14 Dec 2024 23:43:54 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 13:08:24 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Feher",
"Darius",
""
],
[
"Vulić",
"Ivan",
""
],
[
"Minixhofer",
"Benjamin",
""
]
] |
2411.19228 | Stephan Rosswog | S. Rosswog | Shock-capturing particle hydrodynamics with reproducing kernels | 28 pages, 21 figures; accepted for publication in "Computer Modeling in Engineering and Sciences" | Computer Modeling in Engineering & Sciences 2025, 143(2), 1713-1741 | 10.32604/cmes.2025.062063 | null | physics.flu-dyn astro-ph.IM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present and explore a new shock-capturing particle hydrodynamics approach. Our starting point is a commonly used discretization of smoothed particle hydrodynamics. We enhance this discretization with Roe's approximate Riemann solver, we identify its dissipative terms, and in these terms, we use slope-limited linear reconstruction. All gradients needed for our method are calculated with linearly reproducing kernels that are constructed to enforce the two lowest-order consistency relations. We scrutinize our reproducing kernel implementation carefully on a "glass-like" particle distribution, and we find that constant and linear functions are recovered to machine precision. We probe our method in a series of challenging 3D benchmark problems ranging from shocks over instabilities to Schulz-Rinne-type vorticity-creating shocks. All of our simulations show excellent agreement with analytic/reference solutions. | [
{
"version": "v1",
"created": "Thu, 28 Nov 2024 15:53:16 GMT"
},
{
"version": "v2",
"created": "Tue, 13 May 2025 14:13:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Rosswog",
"S.",
""
]
] |
2412.00347 | Truong Xuan Pham | Pham Truong Xuan | Asymptotically almost Periodic Solutions of the parabolic-parabolic Keller-Segel systems on bounded domains | 16 pages | Siberian Mathematical Journal, 2025 | 10.1134/S0037446625030218 | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we investigate the existence, uniqueness, and exponential decay of asymptotically almost periodic (AAP-) mild solutions for the parabolic-parabolic Keller-Segel systems on a bounded domain $\Omega \subset \mathbb{R}^n$ with a smooth boundary. First, we establish the well-posedness of mild solutions for the corresponding linear systems by utilizing the dispersive and smoothing estimates of the Neumann heat semigroup on the bounded domain $\Omega$. We then prove the existence and uniqueness of AAP-mild solutions for the linear systems by providing a Massera-type principle. Next, using results of the linear systems and fixed-point arguments, we derive the well-posedness of such solutions for the Keller-Segel systems. Finally, the exponential decay of these solutions is demonstrated through a Gronwall-type inequality. | [
{
"version": "v1",
"created": "Sat, 30 Nov 2024 04:17:23 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 01:21:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xuan",
"Pham Truong",
""
]
] |
2412.01538 | Riccardo Travaglino | Riccardo Travaglino, Colin Rylands, Pasquale Calabrese | Quasiparticle Picture for Entanglement Hamiltonians in Higher Dimensions | 30 pages + appendices, 5 figures | null | 10.1088/1742-5468/adb7d3 | null | cond-mat.stat-mech hep-th quant-ph | http://creativecommons.org/licenses/by/4.0/ | We employ the quasiparticle picture of entanglement evolution to obtain an effective description for the out-of-equilibrium Entanglement Hamiltonian at the hydrodynamical scale following quantum quenches in free fermionic systems in two or more spatial dimensions. Specifically, we begin by applying dimensional reduction techniques in cases where the geometry permits, building directly on established results from one-dimensional systems. Subsequently, we generalize the analysis to encompass a wider range of geometries. We obtain analytical expressions for the entanglement Hamiltonian valid at the ballistic scale, which reproduce the known quasiparticle picture predictions for the Renyi entropies and full counting statistics. We also numerically validate the results with excellent precision by considering quantum quenches from several initial configurations. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 14:28:36 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Feb 2025 14:55:56 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Travaglino",
"Riccardo",
""
],
[
"Rylands",
"Colin",
""
],
[
"Calabrese",
"Pasquale",
""
]
] |
2412.01569 | Yingli Wang | Yingli Wang and Xiaohong Duan and Ping He | Statistical Inference for Cumulative INAR($\infty$) Processes via Least-Squares | null | null | null | null | math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the cumulative Integer-Valued Autoregressive model of infinite order, denoted as INAR($\infty$), a class of processes crucial for modeling count time series and equivalent to discrete-time Hawkes processes. We propose a computationally efficient conditional least-squares (CLS) estimator to address the challenge of parameter inference in this infinite-dimensional setting. We establish the key theoretical properties of the estimator, including its consistency and asymptotic normality. A central contribution is the rigorous treatment of its large-sample distribution in a framework where the parameter dimension grows with the sample size, for which we derive the corresponding sandwich-form covariance matrix. The theoretical results are substantiated through comprehensive Monte Carlo simulations. These experiments demonstrate that the estimator's accuracy and stability systematically improve as the sample size increases, confirming its consistency. Furthermore, we show that the estimator's finite-sample distribution is well-approximated by a normal distribution, and this approximation becomes more robust with larger samples. Our work provides a complete and practical framework for statistical inference in cumulative INAR($\infty$) models. The code to reproduce the numerical experiments is publicly available at https://github.com/gagawjbytw/INAR_estimation. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 14:53:04 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Jan 2025 14:58:32 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 05:52:43 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 06:19:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wang",
"Yingli",
""
],
[
"Duan",
"Xiaohong",
""
],
[
"He",
"Ping",
""
]
] |
2412.02028 | Talant Talipov | Talant Talipov | Besicovitch-type inequality for closed geodesics on 2-dimensional spheres | Major revisions: a new section has been added containing a proof of the general case. Minor corrections have been made to the statement and proof of Theorem 3.1 | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | We prove the existence of a constant $C > 0$ such that for any Riemannian metric $g$ on a 2-dimensional sphere $S^2$, there exist two distinct closed geodesics with lengths $L_{1}$ and $L_{2}$ satisfying $L_{1} L_{2} \leq C \cdot \operatorname{Area}(S^2, g)$. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 23:17:52 GMT"
},
{
"version": "v2",
"created": "Thu, 12 Dec 2024 20:39:48 GMT"
},
{
"version": "v3",
"created": "Mon, 28 Apr 2025 09:57:34 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 19:45:28 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Talipov",
"Talant",
""
]
] |
2412.02640 | Eugenio Clerico | Eugenio Clerico | On the optimality of coin-betting for mean estimation | 14 pages, 1 figure | null | null | null | math.ST stat.ME stat.TH | http://creativecommons.org/licenses/by/4.0/ | Confidence sequences are sequences of confidence sets that adapt to incoming data while maintaining validity. Recent advances have introduced an algorithmic formulation for constructing some of the tightest confidence sequences for the mean of bounded real random variables. These approaches use a coin-betting framework, where a player sequentially bets on differences between potential mean values and observed data. This work discusses the optimality of such coin-betting formulation among algorithmic frameworks building on e-variables methods to test and estimate the mean of bounded random variables. | [
{
"version": "v1",
"created": "Tue, 03 Dec 2024 18:11:48 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 12:25:32 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Clerico",
"Eugenio",
""
]
] |
2412.03437 | Alessandro V. Cigna | Alessandro V. Cigna | The Thurston norm of graph manifolds | 22 pages, 5 figures | null | null | null | math.GT | http://creativecommons.org/licenses/by/4.0/ | The Thurston norm of a closed oriented graph manifold is a sum of absolute values of linear functionals, and either each or none of the top-dimensional faces of its unit ball are fibered. We show that, conversely, every norm that can be written as a sum of absolute values of linear functionals with rational coefficients is the nonvanishing Thurston norm of some graph manifold, with respect to a rational basis on its second real homology. Moreover, we can choose such graph manifold either to fiber over the circle or not. In particular, every symmetric polygon with rational vertices is the unit polygon of the nonvanishing Thurston norm of a graph manifold fibering over the circle. In dimension $\ge 3$ many symmetric polyhedra with rational vertices are not realizable as nonvanishing Thurston norm ball of any graph manifold. However, given such a polyhedron, we show that there is always a graph manifold whose nonvanishing Thurston norm ball induces a finer partition into cones over the faces. | [
{
"version": "v1",
"created": "Wed, 04 Dec 2024 16:27:56 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:21:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cigna",
"Alessandro V.",
""
]
] |
2412.04139 | Jungwoo Park | Jungwoo Park, Young Jin Ahn, Kee-Eung Kim, Jaewoo Kang | Monet: Mixture of Monosemantic Experts for Transformers | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Understanding the internal computations of large language models (LLMs) is crucial for aligning them with human values and preventing undesirable behaviors like toxic content generation. However, mechanistic interpretability is hindered by polysemanticity -- where individual neurons respond to multiple, unrelated concepts. While Sparse Autoencoders (SAEs) have attempted to disentangle these features through sparse dictionary learning, they have compromised LLM performance due to reliance on post-hoc reconstruction loss. To address this issue, we introduce Mixture of Monosemantic Experts for Transformers (Monet) architecture, which incorporates sparse dictionary learning directly into end-to-end Mixture-of-Experts pretraining. Our novel expert decomposition method enables scaling the expert count to 262,144 per layer while total parameters scale proportionally to the square root of the number of experts. Our analyses demonstrate mutual exclusivity of knowledge across experts and showcase the parametric knowledge encapsulated within individual experts. Moreover, Monet allows knowledge manipulation over domains, languages, and toxicity mitigation without degrading general performance. Our pursuit of transparent LLMs highlights the potential of scaling expert counts to enhance mechanistic interpretability and directly resect the internal knowledge to fundamentally adjust model behavior. The source code and pretrained checkpoints are available at https://github.com/dmis-lab/Monet. | [
{
"version": "v1",
"created": "Thu, 05 Dec 2024 13:06:03 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Dec 2024 07:49:01 GMT"
},
{
"version": "v3",
"created": "Sun, 02 Mar 2025 14:52:21 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 07:36:27 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Park",
"Jungwoo",
""
],
[
"Ahn",
"Young Jin",
""
],
[
"Kim",
"Kee-Eung",
""
],
[
"Kang",
"Jaewoo",
""
]
] |
2412.04643 | Ohad Lib | Ohad Lib, Shuheng Liu, Ronen Shekel, Qiongyi He, Marcus Huber, Yaron Bromberg, Giuseppe Vitagliano | Experimental certification of high-dimensional entanglement with randomized measurements | null | null | 10.1103/PhysRevLett.134.210202 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | High-dimensional entangled states offer higher information capacity and stronger resilience to noise compared with two-dimensional systems. However, the large number of modes and sensitivity to random rotations complicate experimental entanglement certification. Here, we experimentally certify three-dimensional entanglement in a five-dimensional two-photon state using 800 Haar-random measurements implemented via a 10-plane programmable light converter. We further demonstrate the robustness of this approach against random rotations, certifying high-dimensional entanglement despite arbitrary phase randomization of the optical modes. This method, which requires no common reference frame between parties, opens the door for high-dimensional entanglement distribution through long-range random links. | [
{
"version": "v1",
"created": "Thu, 05 Dec 2024 22:27:38 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lib",
"Ohad",
""
],
[
"Liu",
"Shuheng",
""
],
[
"Shekel",
"Ronen",
""
],
[
"He",
"Qiongyi",
""
],
[
"Huber",
"Marcus",
""
],
[
"Bromberg",
"Yaron",
""
],
[
"Vitagliano",
"Giuseppe",
""
]
] |
2412.04761 | Xiaotian Xu | Xiaotian Xu, Pei Sun, Xin Zhang, Junpeng Cao, and Tao Yang | Exact solution of the Izergin-Korepin Gaudin model with periodic and open boundaries | 25page | SciPost Phys. Core 8, 041 (2025) | 10.21468/SciPostPhysCore.8.2.041 | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We study the Izergin-Korepin Gaudin models with both periodic and open integrable boundary conditions, which describe quantum systems exhibiting novel long-range interactions. Using the Bethe ansatz approach, we derive the eigenvalues of the Gaudin operators and the corresponding Bethe ansatz equations. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 04:01:59 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:56:04 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xu",
"Xiaotian",
""
],
[
"Sun",
"Pei",
""
],
[
"Zhang",
"Xin",
""
],
[
"Cao",
"Junpeng",
""
],
[
"Yang",
"Tao",
""
]
] |
2412.04818 | Ignacio Ruiz | Ilma, M. Rafi Alam, L. Alvarez-Ruso, M. Benitez Galan, I. Ruiz Simo, S. K. Singh | Neutrino-nucleon elastic scattering in presence of non-standard interactions: cross sections and nucleon polarizations | 13 pages, 6 figures and 2 tables. This version matches the accepted one in Physical Review D | null | 10.1103/q3sd-jjm6 | null | hep-ph hep-ex hep-lat nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | New physics beyond the Standard Model (SM) may appear in the form of non-standard neutrino interactions (NSI). We have studied neutral current (anti)neutrino-nucleon scattering in presence of NSI. We obtain that in this scenario, nucleon matrix elements depend not only on the isovector axial nucleon form factor but also on the isoscalar one. For the axial form factors we consequently rely on the quark flavor decomposition performed by QCD simulations in the lattice (LQCD). We have examined cross sections and polarization observables. For the current bounds on diagonal muon flavor NSI couplings we find substantial deviations from the SM predictions in cross sections and transverse polarizations of the outgoing nucleons. In view of the progress in the precision of LQCD determinations of nucleon properties, modern measurements of neutral current (anti)neutrino-nucleon scattering will be in the position to discover or significantly constrain NSI. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 07:23:21 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:51:49 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ilma",
"",
""
],
[
"Alam",
"M. Rafi",
""
],
[
"Alvarez-Ruso",
"L.",
""
],
[
"Galan",
"M. Benitez",
""
],
[
"Simo",
"I. Ruiz",
""
],
[
"Singh",
"S. K.",
""
]
] |
2412.05023 | Chhavi Kirtani | Krishnasai Addala, Kabir Dev Paul Baghel, Navya Gupta, Rishitej Reddy Vyalla, Chhavi Kirtani, Avinash Anand, Rajiv Ratn Shah | Steps are all you need: Rethinking STEM Education with Prompt Engineering | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Few shot and Chain-of-Thought prompting have shown promise when applied to Physics Question Answering Tasks, but are limited by the lack of mathematical ability inherent to LLMs, and are prone to hallucination. By utilizing a Mixture of Experts (MoE) Model, along with analogical prompting, we are able to show improved model performance when compared to the baseline on standard LLMs. We also survey the limits of these prompting techniques and the effects they have on model performance. Additionally, we propose Analogical CoT prompting, a prompting technique designed to allow smaller, open source models to leverage Analogical prompting, something they have struggled with, possibly due to a lack of specialist training data. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 13:20:57 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Dec 2024 20:06:24 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 14:47:23 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Addala",
"Krishnasai",
""
],
[
"Baghel",
"Kabir Dev Paul",
""
],
[
"Gupta",
"Navya",
""
],
[
"Vyalla",
"Rishitej Reddy",
""
],
[
"Kirtani",
"Chhavi",
""
],
[
"Anand",
"Avinash",
""
],
[
"Shah",
"Rajiv Ratn",
""
]
] |
2412.05146 | Yeqing Qiu | Yeqing Qiu, Ye Xue, Akang Wang, Yiheng Wang, Qingjiang Shi, Zhi-Quan Luo | ROS: A GNN-based Relax-Optimize-and-Sample Framework for Max-k-Cut Problems | null | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Max-k-Cut problem is a fundamental combinatorial optimization challenge that generalizes the classic NP-complete Max-Cut problem. While relaxation techniques are commonly employed to tackle Max-k-Cut, they often lack guarantees of equivalence between the solutions of the original problem and its relaxation. To address this issue, we introduce the Relax-Optimize-and-Sample (ROS) framework. In particular, we begin by relaxing the discrete constraints to the continuous probability simplex form. Next, we pre-train and fine-tune a graph neural network model to efficiently optimize the relaxed problem. Subsequently, we propose a sampling-based construction algorithm to map the continuous solution back to a high-quality Max-k-Cut solution. By integrating geometric landscape analysis with statistical theory, we establish the consistency of function values between the continuous solution and its mapped counterpart. Extensive experimental results on random regular graphs and the Gset benchmark demonstrate that the proposed ROS framework effectively scales to large instances with up to 20000 nodes in just a few seconds, outperforming state-of-the-art algorithms. Furthermore, ROS exhibits strong generalization capabilities across both in-distribution and out-of-distribution instances, underscoring its effectiveness for large-scale optimization tasks. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 16:02:02 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 17:32:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Qiu",
"Yeqing",
""
],
[
"Xue",
"Ye",
""
],
[
"Wang",
"Akang",
""
],
[
"Wang",
"Yiheng",
""
],
[
"Shi",
"Qingjiang",
""
],
[
"Luo",
"Zhi-Quan",
""
]
] |
2412.05342 | Xiaoyu Wang | Xiaoyu Wang, Ningyuan Xi, Teng Chen, Qingqing Gu, Yue Zhao, Xiaokai Chen, Zhonglin Jiang, Yong Chen, Luo Ji | Multi-Party Supervised Fine-tuning of Language Models for Multi-Party Dialogue Generation | Accepted by IJCNN 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLM) are usually fine-tuned to participate in dyadic or two-party dialogues, which can not adapt well to multi-party dialogues (MPD), which hinders their applications in such scenarios including multi-personal meetings, discussions and daily communication. Previous LLM-based researches mainly focus on the multi-agent framework, while their base LLMs are still pairwisely fine-tuned. In this work, we design a multi-party fine-tuning framework (MuPaS) for LLMs on the multi-party dialogue datasets, and prove such a straightforward framework can let the LLM align with the multi-party conversation style efficiently and effectively. We also design two training strategies which can convert MuPaS into the MPD simulator. Substantial experiments show that MuPaS can achieve state-of-the-art multi-party response, higher accuracy of the-next-speaker prediction, higher human and automatic evaluated utterance qualities, and can even generate reasonably with out-of-distribution scene, topic and role descriptions. The MuPaS framework bridges the LLM training with more complicated multi-party applications, such as conversation generation, virtual rehearsal or meta-universe. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 09:33:47 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Dec 2024 09:47:53 GMT"
},
{
"version": "v3",
"created": "Sun, 11 May 2025 11:26:46 GMT"
},
{
"version": "v4",
"created": "Tue, 13 May 2025 03:10:40 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 09:49:06 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wang",
"Xiaoyu",
""
],
[
"Xi",
"Ningyuan",
""
],
[
"Chen",
"Teng",
""
],
[
"Gu",
"Qingqing",
""
],
[
"Zhao",
"Yue",
""
],
[
"Chen",
"Xiaokai",
""
],
[
"Jiang",
"Zhonglin",
""
],
[
"Chen",
"Yong",
""
],
[
"Ji",
"Luo",
""
]
] |
2412.05453 | Chhavi Kirtani | Krishnasai Addala, Kabir Dev Paul Baghel, Dhruv Jain, Navya Gupta, Rishitej Reddy Vyalla, Chhavi Kirtani, Avinash Anand, Rajiv Ratn Shah | Knowledge Graphs are all you need: Leveraging KGs in Physics Question Answering | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study explores the effectiveness of using knowledge graphs generated by large language models to decompose high school-level physics questions into sub-questions. We introduce a pipeline aimed at enhancing model response quality for Question Answering tasks. By employing LLMs to construct knowledge graphs that capture the internal logic of the questions, these graphs then guide the generation of subquestions. We hypothesize that this method yields sub-questions that are more logically consistent with the original questions compared to traditional decomposition techniques. Our results show that sub-questions derived from knowledge graphs exhibit significantly improved fidelity to the original question's logic. This approach not only enhances the learning experience by providing clearer and more contextually appropriate sub-questions but also highlights the potential of LLMs to transform educational methodologies. The findings indicate a promising direction for applying AI to improve the quality and effectiveness of educational content. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 22:25:23 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Dec 2024 20:40:52 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 14:43:44 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Addala",
"Krishnasai",
""
],
[
"Baghel",
"Kabir Dev Paul",
""
],
[
"Jain",
"Dhruv",
""
],
[
"Gupta",
"Navya",
""
],
[
"Vyalla",
"Rishitej Reddy",
""
],
[
"Kirtani",
"Chhavi",
""
],
[
"Anand",
"Avinash",
""
],
[
"Shah",
"Rajiv Ratn",
""
]
] |
2412.06085 | Albrecht Kurze | Albrecht Kurze, Karola K\"opferl | From Simple Sensors to Complex Context: Insights for HabiTech | CHI24 Extended Abstracts, Workshop HabiTech, May 11, 2024, Honolulu, HI, USA 2024. 4 pages, 5 figures | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We relate our previous as well as ongoing research in the domain of smart homes to the concept of HabiTech. HabiTech can benefit from existing approaches and findings in a broader context of whole buildings or communities within. Along with data comes context of data capture and data interpretation in different dimensions (spatial, temporal, social). For defining what is 'community' proximity plays a crucial role in context, both spatially as well as socially. A participatory approach for research in living in sensing environments is promising to address complexity as well as interests of different stakeholders. Often it is the complex context that makes even simple sensor data sensitive, i.e. in terms of privacy. When it comes to handle shared data then concepts from the physical world for shared spaces might be related back to the data domain. | [
{
"version": "v1",
"created": "Sun, 08 Dec 2024 22:19:57 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 13:02:03 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kurze",
"Albrecht",
""
],
[
"Köpferl",
"Karola",
""
]
] |
2412.06218 | Alexander Hulpke | Alexander Hulpke | Arithmetic in Group Extensions | null | null | null | null | math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a generalization of the concept of a pc presentation that applies to groups with a nontrivial solvable radical. Such a representation can be much more efficient in terms of memory use and even of arithmetic, than permuattion and matrix representations. We illustrate the use of such representations by constructing a maximal subgroup of the sporadic monster group and calculating its -- hitherto unknown -- character table. | [
{
"version": "v1",
"created": "Mon, 09 Dec 2024 05:27:56 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:31:14 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Hulpke",
"Alexander",
""
]
] |
2412.06845 | Pu Zhao | Pu Zhao, Xuan Shen, Zhenglun Kong, Yixin Shen, Sung-En Chang, Timothy Rupprecht, Lei Lu, Enfu Nan, Changdi Yang, Yumei He, Weiyan Shi, Xingchen Xu, Yu Huang, Wei Jiang, Wei Wang, Yue Chen, Yong He, Yanzhi Wang | 7B Fully Open Source Moxin-LLM/VLM -- From Pretraining to GRPO-based Reinforcement Learning Enhancement | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recently, Large Language Models (LLMs) have undergone a significant transformation, marked by a rapid rise in both their popularity and capabilities. Leading this evolution are proprietary LLMs like GPT-4 and GPT-o1, which have captured widespread attention in the AI community due to their remarkable performance and versatility. Simultaneously, open-source LLMs, such as LLaMA, have made great contributions to the ever-increasing popularity of LLMs due to the ease to customize and deploy the models across diverse applications. Although open-source LLMs present unprecedented opportunities for innovation and research, the commercialization of LLMs has raised concerns about transparency, reproducibility, and safety. Many open-source LLMs fail to meet fundamental transparency requirements by withholding essential components like training code and data, which may hinder further innovations on LLMs. To mitigate this issue, we introduce Moxin 7B, a fully open-source LLM developed, adhering to principles of open science, open source, open data, and open access. We release the pre-training code and configurations, training and fine-tuning datasets, and intermediate and final checkpoints, aiming to make continuous commitments to fully open-source LLMs. After pre-training the base model, we finetune the Moxin Base model with SOTA post-training framework and instruction data to obtain Moxin Instruct model. To improve the reasoning capability, we further finetune our Instruct model with chain-of-thought data distilled from DeepSeek R1, and then use Group Relative Policy Optimization (GRPO) following DeepSeek R1 to finetune our model, leading to the Moxin Reasoning model. Moreover, we develop our vision language model based on our Moxin model. Experiments show that our models achieve superior performance in various evaluations such as zero-shot evaluation, few-shot evaluation, and CoT evaluation. | [
{
"version": "v1",
"created": "Sun, 08 Dec 2024 02:01:46 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Dec 2024 19:03:58 GMT"
},
{
"version": "v3",
"created": "Thu, 10 Apr 2025 19:05:16 GMT"
},
{
"version": "v4",
"created": "Wed, 23 Apr 2025 01:38:02 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 17:10:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhao",
"Pu",
""
],
[
"Shen",
"Xuan",
""
],
[
"Kong",
"Zhenglun",
""
],
[
"Shen",
"Yixin",
""
],
[
"Chang",
"Sung-En",
""
],
[
"Rupprecht",
"Timothy",
""
],
[
"Lu",
"Lei",
""
],
[
"Nan",
"Enfu",
""
],
[
"Yang",
"Changdi",
""
],
[
"He",
"Yumei",
""
],
[
"Shi",
"Weiyan",
""
],
[
"Xu",
"Xingchen",
""
],
[
"Huang",
"Yu",
""
],
[
"Jiang",
"Wei",
""
],
[
"Wang",
"Wei",
""
],
[
"Chen",
"Yue",
""
],
[
"He",
"Yong",
""
],
[
"Wang",
"Yanzhi",
""
]
] |
2412.08184 | Jiawei Zhang | Jiawei Zhang, Weipeng Zhang, Tengji Xu, Lei Xu, Eli A. Doris, Bhavin J. Shastri, Chaoran Huang and Paul R. Prucnal | Online training and pruning of multi-wavelength photonic neural networks | 21 pages, 5 figures | null | null | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | CMOS-compatible photonic integrated circuits (PICs) are emerging as a promising platform in artificial intelligence (AI) computing. Owing to the compact footprint of microring resonators (MRRs) and the enhanced interconnect efficiency enabled by wavelength division multiplexing (WDM), MRR-based photonic neural networks (PNNs) are particularly promising for large-scale integration. However, the scalability and energy efficiency of such systems are fundamentally limited by the MRR resonance wavelength variations induced by fabrication process variations (FPVs) and environmental fluctuations. Existing solutions use post-fabrication approaches or thermo-optic tuning, incurring high control power and additional process complexity. In this work, we introduce an online training and pruning method that addresses this challenge, adapting to FPV-induced and thermally induced shifts in MRR resonance wavelength. By incorporating a power-aware pruning term into the conventional loss function, our approach simultaneously optimizes the PNN accuracy and the total power consumption for MRR tuning. In proof-of-concept on-chip experiments on the Iris dataset, our system PNNs can adaptively train to maintain a 96% classification accuracy, while achieving a 44.7% reduction in tuning power via pruning. Additionally, our approach reduces the power consumption by orders-of-magnitude on larger datasets. By addressing chip-to-chip variation and minimizing power requirements, our approach significantly improves the scalability and energy efficiency of MRR-based integrated analog photonic processors, paving the way for large-scale PICs to enable versatile applications including neural networks, photonic switching, LiDAR, and radio-frequency beamforming. | [
{
"version": "v1",
"created": "Wed, 11 Dec 2024 08:20:41 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 21:03:27 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 07:12:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhang",
"Jiawei",
""
],
[
"Zhang",
"Weipeng",
""
],
[
"Xu",
"Tengji",
""
],
[
"Xu",
"Lei",
""
],
[
"Doris",
"Eli A.",
""
],
[
"Shastri",
"Bhavin J.",
""
],
[
"Huang",
"Chaoran",
""
],
[
"Prucnal",
"Paul R.",
""
]
] |
2412.08258 | Angelo Salatino Dr | Tanay Aggarwal, Angelo Salatino, Francesco Osborne, Enrico Motta | Large Language Models for Scholarly Ontology Generation: An Extensive Analysis in the Engineering Field | Now accepted to Information Processing & Management. this is the camera ready | null | null | null | cs.DL cs.AI cs.IR | http://creativecommons.org/licenses/by/4.0/ | Ontologies of research topics are crucial for structuring scientific knowledge, enabling scientists to navigate vast amounts of research, and forming the backbone of intelligent systems such as search engines and recommendation systems. However, manual creation of these ontologies is expensive, slow, and often results in outdated and overly general representations. As a solution, researchers have been investigating ways to automate or semi-automate the process of generating these ontologies. This paper offers a comprehensive analysis of the ability of large language models (LLMs) to identify semantic relationships between different research topics, which is a critical step in the development of such ontologies. To this end, we developed a gold standard based on the IEEE Thesaurus to evaluate the task of identifying four types of relationships between pairs of topics: broader, narrower, same-as, and other. Our study evaluates the performance of seventeen LLMs, which differ in scale, accessibility (open vs. proprietary), and model type (full vs. quantised), while also assessing four zero-shot reasoning strategies. Several models have achieved outstanding results, including Mixtral-8x7B, Dolphin-Mistral-7B, and Claude 3 Sonnet, with F1-scores of 0.847, 0.920, and 0.967, respectively. Furthermore, our findings demonstrate that smaller, quantised models, when optimised through prompt engineering, can deliver performance comparable to much larger proprietary models, while requiring significantly fewer computational resources. | [
{
"version": "v1",
"created": "Wed, 11 Dec 2024 10:11:41 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:58:18 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Aggarwal",
"Tanay",
""
],
[
"Salatino",
"Angelo",
""
],
[
"Osborne",
"Francesco",
""
],
[
"Motta",
"Enrico",
""
]
] |
2412.08616 | Phan Th\`anh Nam | Charlotte Dietze and Phan Th\`anh Nam | Minimizing sequences of Sobolev inequalities revisited | 12 pages, final version to appear in IMRN | null | null | null | math.AP math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give a new proof of the compactness of minimizing sequences of the Sobolev inequalities in the critical case. Our approach relies on a simplified version of the concentration-compactness principle, which does not require any refinement of the Sobolev embedding theorem. | [
{
"version": "v1",
"created": "Wed, 11 Dec 2024 18:38:49 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:40:05 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Dietze",
"Charlotte",
""
],
[
"Nam",
"Phan Thành",
""
]
] |
2412.09607 | Carlos Esteves | Carlos Esteves, Mohammed Suhail, Ameesh Makadia | Spectral Image Tokenizer | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Image tokenizers map images to sequences of discrete tokens, and are a crucial component of autoregressive transformer-based image generation. The tokens are typically associated with spatial locations in the input image, arranged in raster scan order, which is not ideal for autoregressive modeling. In this paper, we propose to tokenize the image spectrum instead, obtained from a discrete wavelet transform (DWT), such that the sequence of tokens represents the image in a coarse-to-fine fashion. Our tokenizer brings several advantages: 1) it leverages that natural images are more compressible at high frequencies, 2) it can take and reconstruct images of different resolutions without retraining, 3) it improves the conditioning for next-token prediction -- instead of conditioning on a partial line-by-line reconstruction of the image, it takes a coarse reconstruction of the full image, 4) it enables partial decoding where the first few generated tokens can reconstruct a coarse version of the image, 5) it enables autoregressive models to be used for image upsampling. We evaluate the tokenizer reconstruction metrics as well as multiscale image generation, text-guided image upsampling and editing. | [
{
"version": "v1",
"created": "Thu, 12 Dec 2024 18:59:31 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 17:58:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Esteves",
"Carlos",
""
],
[
"Suhail",
"Mohammed",
""
],
[
"Makadia",
"Ameesh",
""
]
] |
2412.10233 | Krzysztof Ptaszynski | Krzysztof Ptaszynski, Timur Aslyamov, Massimiliano Esposito | Nonequilibrium Fluctuation-Response Relations for State Observables | 8 pages, 2 figures | null | null | null | cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | Time-integrated state observables, which quantify the fraction of time spent by the system in a specific pool of states, are important in many fields, such as chemical sensing or the theory of fluorescence spectroscopy. We derive exact identities, called Fluctuation-Response Relations (FRRs), that connect the fluctuations of such observables to their response to external perturbations in nonequilibrium steady state of Markov jump processes. Using these results, we derive novel upper and lower bounds for fluctuations. We further demonstrate their applicability for simplifying calculations of fluctuations in large Markov networks, use them to explain the physical origin of positive and negative correlations of occupation times in a double quantum dot device, and discuss their relevance for model inference. | [
{
"version": "v1",
"created": "Fri, 13 Dec 2024 16:02:38 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:50:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ptaszynski",
"Krzysztof",
""
],
[
"Aslyamov",
"Timur",
""
],
[
"Esposito",
"Massimiliano",
""
]
] |
2412.11478 | Francesco Parente | Mirna D\v{z}amonja, Francesco Parente | Properties preserved by classes of Chu transforms | 21 pages | null | null | null | math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Chu spaces and Chu transforms were first investigated in category theory by Barr and Chu in 1979. In 2000 van Benthem shifted to the model-theoretic point of view by isolating a class of infinitary two-sorted properties, the flow formulas, which are preserved by all Chu transforms. D\v{z}amonja and V\"a\"an\"anen in 2021 considered a special kind of Chu transforms, satisfying a density condition. These authors used dense Chu transforms to compare abstract logics, in particular showing that such transforms preserve compactness.
This motivates the problem of characterizing which properties are preserved by dense Chu transforms. We solve this problem by isolating the inconsistency-flow formulas, which are flow formulas with an added predicate to express inconsistency. Our main result characterizes the first-order inconsistency-flow formulas exactly as those first-order properties preserved by dense Chu transforms. Furthermore, we consider several instances of Chu transforms in concrete situations such as topological spaces, graphs and ultrafilters. In particular, we prove that Chu transforms between ultrafilters correspond to the Rudin-Keisler ordering. | [
{
"version": "v1",
"created": "Mon, 16 Dec 2024 06:46:15 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 05:18:27 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Džamonja",
"Mirna",
""
],
[
"Parente",
"Francesco",
""
]
] |
2412.11598 | Ludovic Levy Patey | Lorenzo Carlucci, Oriola Gjetaj, Quentin Le Hou\'erou and Ludovic Levy Patey | Ramsey-like theorems for the Schreier barrier | 32 pages | null | null | Roma01.Math | math.LO math.CO | http://creativecommons.org/licenses/by/4.0/ | The family of finite subsets $s$ of the natural numbers such that $|s|=1+\min s$ is known as the Schreier barrier in combinatorics and Banach Space theory, and as the family of exactly $\omega$-large sets in Logic. We formulate and prove the generalizations of Friedman's Free Set and Thin Set theorems and of Rainbow Ramsey's theorem to colorings of the Schreier barrier. We analyze the strength of these theorems from the point of view of Computability Theory and Reverse Mathematics. Surprisingly, the exactly $\omega$-large counterparts of the Thin Set and Free Set theorems can code $\emptyset^{(\omega)}$, while the exactly $\omega$-large Rainbow Ramsey theorem does not code the halting set. | [
{
"version": "v1",
"created": "Mon, 16 Dec 2024 09:36:26 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 13:25:47 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Carlucci",
"Lorenzo",
""
],
[
"Gjetaj",
"Oriola",
""
],
[
"Houérou",
"Quentin Le",
""
],
[
"Patey",
"Ludovic Levy",
""
]
] |
2412.11991 | Paul Manns | Paul Manns | Convergence of trust-region algorithms in metric spaces | null | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Trust-region algorithms can be applied to very abstract optimization problems because they do not require a specific direction of descent or gradient. This has lead to recent interest in them, in particular in the area of integer optimal control problems, where the infinite-dimensional problem formulations do not assume vector space structure.
We analyze a trust-region algorithm in the abstract setting of a metric space, a setting in which integer optimal control problems with total variation regularization can be formulated. Our analysis avoids a reset of the trust-region radius upon acceptance of the iterates when proving convergence to stationary points. This reset has been present in previous analyses of trust-region algorithms for integer optimal control problems. Our computational benchmark shows that the runtime can be considerably improved when avoiding this reset, which is now theoretically justified. | [
{
"version": "v1",
"created": "Mon, 16 Dec 2024 17:14:49 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 18:08:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Manns",
"Paul",
""
]
] |
2412.12590 | Namhyun Kim | Namhyun Kim, Juntaek Han, Jinseok Choi, Ahmed Alkhateeb, Chan-Byoung Chae, and Jeonghun Park | Integrated Sensing and Communications in Downlink FDD MIMO without CSI Feedback | submitted to possible IEEE publication | null | null | null | cs.IT eess.SP math.IT | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we propose a precoding framework for frequency division duplex (FDD) integrated sensing and communication (ISAC) systems with multiple-input multiple-output (MIMO). Specifically, we aim to maximize ergodic sum spectral efficiency (SE) while satisfying a sensing beam pattern constraint defined by the mean squared error (MSE). Our method reconstructs downlink (DL) channel state information (CSI) from uplink (UL) training signals using partial reciprocity, eliminating the need for CSI feedback. To obtain the error covariance matrix of the reconstructed DL CSI, we devise an observed Fisher information-based estimation technique. Leveraging this, to mitigate interference caused by imperfect DL CSI reconstruction and sensing operations, we propose a rate-splitting multiple access (RSMA) aided precoder optimization method. This method jointly updates the precoding vector and Lagrange multipliers by solving the nonlinear eigenvalue problem with eigenvector dependency to maximize SE. The numerical results show that the proposed design achieves precise beam pattern control, maximizes SE, and significantly improves the sensing-communication trade-off compared to the state-of-the-art methods in FDD ISAC scenarios. | [
{
"version": "v1",
"created": "Tue, 17 Dec 2024 06:46:46 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 08:34:48 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 02:48:15 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kim",
"Namhyun",
""
],
[
"Han",
"Juntaek",
""
],
[
"Choi",
"Jinseok",
""
],
[
"Alkhateeb",
"Ahmed",
""
],
[
"Chae",
"Chan-Byoung",
""
],
[
"Park",
"Jeonghun",
""
]
] |
2412.13235 | Ricardo Euler | Ricardo Euler, Pedro Maristany de las Casas, Ralf Bornd\"orfer | Logic-Constrained Shortest Paths for Flight Planning | null | null | null | null | cs.AI cs.DM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The logic-constrained shortest path problem (LCSPP) combines a one-to-one shortest path problem with satisfiability constraints imposed on the routing graph. This setting arises in flight planning, where air traffic control (ATC) authorities are enforcing a set of traffic flow restrictions (TFRs) on aircraft routes in order to increase safety and throughput. We propose a new branch and bound-based algorithm for the LCSPP. The resulting algorithm has three main degrees of freedom: the node selection rule, the branching rule and the conflict. While node selection and branching rules have been long studied in the MIP and SAT communities, most of them cannot be applied out of the box for the LCSPP. We review the existing literature and develop tailored variants of the most prominent rules. The conflict, the set of variables to which the branching rule is applied, is unique to the LCSPP. We analyze its theoretical impact on the B&B algorithm. In the second part of the paper, we show how to model the flight planning problem with TFRs as an LCSPP and solve it using the branch and bound algorithm. We demonstrate the algorithm's efficiency on a dataset consisting of a global flight graph and a set of around 20000 real TFRs obtained from our industry partner Lufthansa Systems GmbH. We make this dataset publicly available. Finally, we conduct an empirical in-depth analysis of dynamic shortest path algorithms, node selection rules, branching rules and conflicts. Carefully choosing an appropriate combination yields an improvement of an order of magnitude compared to an uninformed choice. | [
{
"version": "v1",
"created": "Tue, 17 Dec 2024 16:18:06 GMT"
},
{
"version": "v2",
"created": "Fri, 20 Dec 2024 10:38:40 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 12:20:01 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Euler",
"Ricardo",
""
],
[
"Casas",
"Pedro Maristany de las",
""
],
[
"Borndörfer",
"Ralf",
""
]
] |
2412.13423 | Sandro D. P. Vitenti | Sandro D. P. Vitenti, Fernando de Simoni, Mariana Penna-Lima, Eduardo J. Barroso | AutoKnots: Adaptive Knot Allocation for Spline Interpolation | 38 pages, 13 figures, to be submitted | Astronomy and Computing Volume 53, October 2025, 100970 | 10.1016/j.ascom.2025.100970 | null | astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In astrophysical and cosmological analyses, the increasing quality and volume of astronomical data demand efficient and precise computational tools. This work introduces a novel adaptive algorithm for automatic knots (AutoKnots) allocation in spline interpolation, designed to meet user-defined precision requirements. Unlike traditional methods that rely on manually configured knot distributions with numerous parameters, the proposed technique automatically determines the optimal number and placement of knots based on interpolation error criteria. This simplifies configuration, often requiring only a single parameter. The algorithm progressively improves the interpolation by adaptively sampling the function-to-be-approximated, $f(x)$, in regions where the interpolation error exceeds the desired threshold. All function evaluations contribute directly to the final approximation, ensuring efficiency. While each resampling step involves recomputing the interpolation table, this process is highly optimized and usually computationally negligible compared to the cost of evaluating $f(x)$. We show the algorithm's efficacy through a series of precision tests on different functions. However, the study underscores the necessity for caution when dealing with certain function types, notably those featuring plateaus. To address this challenge, a heuristic enhancement is incorporated, improving accuracy in flat regions. This algorithm has been extensively used and tested over the years. NumCosmo includes a comprehensive set of unit tests that rigorously evaluate the algorithm both directly and indirectly, underscoring its robustness and reliability. As a practical application, we compute the surface mass density $\Sigma(R)$ and the average surface mass density $\overline{\Sigma}(<R)$ for Navarro-Frenk-White and Hernquist halo density profiles, which provide analytical benchmarks. (abridged) | [
{
"version": "v1",
"created": "Wed, 18 Dec 2024 01:38:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Vitenti",
"Sandro D. P.",
""
],
[
"de Simoni",
"Fernando",
""
],
[
"Penna-Lima",
"Mariana",
""
],
[
"Barroso",
"Eduardo J.",
""
]
] |
2412.13639 | Fernando Amodeo | Fernando Amodeo, Luis Merino, Fernando Caballero | 4D Radar-Inertial Odometry based on Gaussian Modeling and Multi-Hypothesis Scan Matching | Our code and results can be publicly accessed at: https://github.com/robotics-upo/gaussian-rio-cpp | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | 4D millimeter-wave (mmWave) radars are sensors that provide robustness against adverse weather conditions (rain, snow, fog, etc.), and as such they are increasingly being used for odometry and SLAM applications. However, the noisy and sparse nature of the returned scan data proves to be a challenging obstacle for existing point cloud matching based solutions, especially those originally intended for more accurate sensors such as LiDAR. Inspired by visual odometry research around 3D Gaussian Splatting, in this paper we propose using freely positioned 3D Gaussians to create a summarized representation of a radar point cloud tolerant to sensor noise, and subsequently leverage its inherent probability distribution function for registration (similar to NDT). Moreover, we propose simultaneously optimizing multiple scan matching hypotheses in order to further increase the robustness of the system against local optima of the function. Finally, we fuse our Gaussian modeling and scan matching algorithms into an EKF radar-inertial odometry system designed after current best practices. Experiments using publicly available 4D radar datasets show that our Gaussian-based odometry is comparable to existing registration algorithms, outperforming them in several sequences. | [
{
"version": "v1",
"created": "Wed, 18 Dec 2024 09:11:24 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 10:49:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Amodeo",
"Fernando",
""
],
[
"Merino",
"Luis",
""
],
[
"Caballero",
"Fernando",
""
]
] |
2412.13727 | Dana Kovaleva | Juan Casado, Yasser Hendy, Nikola Faltova and Dana Kovaleva | Open Superclusters II. Discovery of an Inactive Supercluster Including a New Star Cluster in the Solar Neighborhood | 22 pages, 2 tables, 11 figures | null | null | null | astro-ph.SR astro-ph.GA physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The number of known star clusters, associations, and moving groups in the vicinity of the Sun has increased significantly due to the use of Gaia data. We investigated the existence and properties of open superclusters (OSC) within 500 pc of the Sun, with a particular focus on a newly identified OSC, designated HC8. Using advanced Gaia-derived astrometric data and star cluster catalogs, we identify and analyze member stars of various clusters, deriving key parameters such as ages, distances, extinctions, and kinematic properties. Our findings reveal that HC8, a previously unrecognized OSC, encompasses several young clusters, including the newly discovered cluster Duvia 1. We provide a detailed examination of HC8's star formation history and its spatial and kinematic characteristics. This study contributes to the growing body of evidence supporting the existence of highly populated primordial groups and enhances our understanding of the formation and evolution of star clusters in the Galaxy. | [
{
"version": "v1",
"created": "Wed, 18 Dec 2024 11:06:47 GMT"
},
{
"version": "v2",
"created": "Fri, 30 May 2025 17:18:58 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:43:31 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Casado",
"Juan",
""
],
[
"Hendy",
"Yasser",
""
],
[
"Faltova",
"Nikola",
""
],
[
"Kovaleva",
"Dana",
""
]
] |
2412.14851 | Adrian Petr | Guillaume Laplante-Anfossi, Adrian Petr, Vivek Shende | A universal characterization of the curved homotopy Lie and associative operads | 16 pages, added the case of symmetric operads | null | null | null | math.AT math.CT math.SG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the category of nonsymmetric dg operads valued in strict graded-mixed complexes, equipped with a distinguished arity zero weight one element which generates the weight grading, and whose differential has weight one. We show that the initial object is the curved A-infinity operad, that the forgetful functor to the category of operads under it admits a right adjoint, and that the unit of the adjunction encodes the operation of twisting a curved A-infinity algebra by a Maurer-Cartan element.
The corresponding notions for symmetric operads characterize the curved L-infinity operad and the corresponding twisting procedure. | [
{
"version": "v1",
"created": "Thu, 19 Dec 2024 13:42:59 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:58:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Laplante-Anfossi",
"Guillaume",
""
],
[
"Petr",
"Adrian",
""
],
[
"Shende",
"Vivek",
""
]
] |
2412.15383 | Lucas Napolitano | Lucas Napolitano, Adam D. Myers, Vicky Fawcett, Jessica Aguilar, Steven Ahlen, Davide Bianchi, David Brooks, Todd Claybaugh, Shaun Cole, Axel de la Macorra, Biprateep Dey, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gazta\~naga, Satya Gontcho A Gontcho, Gaston Gutierrez, Klaus Honscheid, Stephanie Juneau, Andrew Lambert, Martin Landriau, Laurent Le Guillou, Aaron Meisner, Ramon Miquel, John Moustakas, Jeffrey A. Newman, Francisco Prada, Ignasi P\'erez-R\`afols, Graziano Rossi, Eusebio Sanchez, David Schlegel, Michael Schubnell, David Sprayberry, Gregory Tarl\'e, Benjamin Alan Weaver, and Hu Zou | DESI Mg II Absorbers: Extinction Characteristics & Quasar Redshift Accuracy | 15 pages, 9 figures, 1 table, published | AJ 170 16 (2025) | 10.3847/1538-3881/adc389 | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | In this paper, we study how absorption-line systems affect the spectra and redshifts of quasars (QSOs), using catalogs of Mg II absorbers from the early data release (EDR) and first data release (DR1) of the Dark Energy Spectroscopic Instrument (DESI). We determine the reddening effect of an absorption system by fitting an un-reddened template spectrum to a sample of 50,674 QSO spectra that contain Mg II absorbers. We find that reddening caused by intervening absorbers (voff > 3500 km/s) has an average color excess of E(B-V) = 0.04 magnitudes. We find that the E(B-V) tends to be greater for absorbers at low redshifts, or those having Mg II absorption lines with higher equivalent widths, but shows no clear trend with voff for intervening systems. However, the E(B-V) of associated absorbers, those at voff < 3500 km/s, shows a strong trend with voff , increasing rapidly with decreasing voff and peaking (approximately 0.15 magnitudes) around voff = 0 km/s. We demonstrate that Mg II absorbers impact redshift estimation for QSOs by investigating the distributions of voff for associated absorbers. We find that at z > 1.5 these distributions broaden and bifurcate in a nonphysical manner. In an effort to mitigate this effect, we mask pixels associated with the Mg II absorption lines and recalculate the QSO redshifts. We find that we can recover voff populations in better agreement with those for z < 1.5 absorbers and in doing so typically shift background QSO redshifts by delta_z approximately equal to plus or minus 0.005. | [
{
"version": "v1",
"created": "Thu, 19 Dec 2024 20:21:52 GMT"
},
{
"version": "v2",
"created": "Mon, 02 Jun 2025 18:37:53 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 16:34:55 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Napolitano",
"Lucas",
""
],
[
"Myers",
"Adam D.",
""
],
[
"Fawcett",
"Vicky",
""
],
[
"Aguilar",
"Jessica",
""
],
[
"Ahlen",
"Steven",
""
],
[
"Bianchi",
"Davide",
""
],
[
"Brooks",
"David",
""
],
[
"Claybaugh",
"Todd",
""
],
[
"Cole",
"Shaun",
""
],
[
"de la Macorra",
"Axel",
""
],
[
"Dey",
"Biprateep",
""
],
[
"Font-Ribera",
"Andreu",
""
],
[
"Forero-Romero",
"Jaime E.",
""
],
[
"Gaztañaga",
"Enrique",
""
],
[
"Gontcho",
"Satya Gontcho A",
""
],
[
"Gutierrez",
"Gaston",
""
],
[
"Honscheid",
"Klaus",
""
],
[
"Juneau",
"Stephanie",
""
],
[
"Lambert",
"Andrew",
""
],
[
"Landriau",
"Martin",
""
],
[
"Guillou",
"Laurent Le",
""
],
[
"Meisner",
"Aaron",
""
],
[
"Miquel",
"Ramon",
""
],
[
"Moustakas",
"John",
""
],
[
"Newman",
"Jeffrey A.",
""
],
[
"Prada",
"Francisco",
""
],
[
"Pérez-Ràfols",
"Ignasi",
""
],
[
"Rossi",
"Graziano",
""
],
[
"Sanchez",
"Eusebio",
""
],
[
"Schlegel",
"David",
""
],
[
"Schubnell",
"Michael",
""
],
[
"Sprayberry",
"David",
""
],
[
"Tarlé",
"Gregory",
""
],
[
"Weaver",
"Benjamin Alan",
""
],
[
"Zou",
"Hu",
""
]
] |
2412.15512 | A. Lusakowski | A. {\L}usakowski, P. Bogus{\l}awski, T. Story | Bi, Cr and Ag dopants in PbTe and SnTe: impact of the host band symmetry on doping properties by ab initio calculations | null | null | null | null | cond-mat.mtrl-sci cond-mat.other | http://creativecommons.org/publicdomain/zero/1.0/ | Doping properties of Bi, Cr and Ag dopants in thermoelectric and topological materials PbTe and SnTe are analyzed based on density functional theory calculations in the local density approximations and the large supercell method. In agreement with experiment, in both PbTe and SnTe, Bi is a donor and Ag is an acceptor with a vanishing magnetic moment. In contrast, Cr is a resonant donor in PbTe, and an resonant acceptor in SnTe. We also consider the electronic structure of cation vacancies in PbTe and SnTe, since these abundant native defects induce $p$-type conductivity in both hosts. The quantitatively different impact of these dopants/defects on the host band structure of PbTe and SnTe (level energies, band splittings, band inversion, and a different level of hybridization between dopant and host states) is explained based on the group-theoretical arguments. | [
{
"version": "v1",
"created": "Fri, 20 Dec 2024 02:59:17 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 12:06:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Łusakowski",
"A.",
""
],
[
"Bogusławski",
"P.",
""
],
[
"Story",
"T.",
""
]
] |
2412.15824 | Istvan Pusztai | N. J. Guth, O. Vallhagen, P. Helander, A. Tresnjic, S. L. Newton, T. F\"ul\"op, I. Pusztai | The effect of plasmoid drifts on the pellet rocket effect in magnetic confinement fusion plasmas | null | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We detail here a semi-analytical model for the pellet rocket effect, which describes the acceleration of pellets in a fusion plasma due to asymmetries in the heat flux reaching the pellet surface and the corresponding ablation rate. This effect was shown in experiments to significantly modify the pellet trajectory, and projections for reactor scale devices indicate that it may severely limit the effectiveness of pellet injection methods. We account for asymmetries stemming both from plasma parameter gradients and an asymmetric plasmoid shielding caused by the drift of the ionized pellet cloud. For high temperature, reactor relevant scenarios, we find a wide range of initial pellet sizes and speeds where the rocket effect severely limits the penetration depth of the pellet. In these cases, the plasma parameter profile variations dominate the rocket effect. We find that for small and fast pellets, where the rocket effect is less pronounced, plasmoid shielding induced asymmetries dominate. | [
{
"version": "v1",
"created": "Fri, 20 Dec 2024 12:05:19 GMT"
},
{
"version": "v2",
"created": "Tue, 13 May 2025 11:45:13 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 13:26:28 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Guth",
"N. J.",
""
],
[
"Vallhagen",
"O.",
""
],
[
"Helander",
"P.",
""
],
[
"Tresnjic",
"A.",
""
],
[
"Newton",
"S. L.",
""
],
[
"Fülöp",
"T.",
""
],
[
"Pusztai",
"I.",
""
]
] |
2412.16730 | Anthony McFadden | Anthony P. McFadden and Jinsu Oh and Lin Zhou and Trevyn F.Q. Larson and Stephen Gill and Akash V. Dixit and Raymond Simmonds and Florent Lecocq | Interface-sensitive microwave loss in superconducting tantalum films sputtered on c-plane sapphire | null | null | null | null | quant-ph cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Quantum coherence in superconducting circuits has increased steadily over the last decades as a result of a growing understanding of the various loss mechanisms. Recently, tantalum (Ta) emerged as a promising material to address microscopic sources of loss found on niobium (Nb) or aluminum (Al) surfaces. However, the effects of film and interface microstructure on low-temperature microwave loss are still not well understood. Here we present a systematic study of the structural and electrical properties of Ta and Nb films sputtered on c-plane sapphire at varying growth temperatures. As growth temperature is increased, our results show that the onset of epitaxial growth of alpha-phase Ta correlates with lower Ta surface roughness, higher critical temperature, and higher residual resistivity ratio, but surprisingly also correlates with a significant increase in loss at microwave frequency. Notably, this high level of loss is not observed in Nb films prepared in the same way and having very similar structure. By experimentally controlling the surface on which the Ta film is nucleated, we determine that the source of loss is only present in samples having an epitaxial Ta/sapphire interface and show that it is apparently mitigated by either growing a thin, epitaxial Nb inter-layer between the Ta film and the substrate or by intentionally treating, and effectively damaging, the sapphire surface with an in-situ argon plasma before Ta growth. In addition to elucidating this interfacial microwave loss, this work provides adequate process details to aid reproducible growth of low-loss Ta films across fabrication facilities. | [
{
"version": "v1",
"created": "Sat, 21 Dec 2024 18:40:49 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 15:03:11 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"McFadden",
"Anthony P.",
""
],
[
"Oh",
"Jinsu",
""
],
[
"Zhou",
"Lin",
""
],
[
"Larson",
"Trevyn F. Q.",
""
],
[
"Gill",
"Stephen",
""
],
[
"Dixit",
"Akash V.",
""
],
[
"Simmonds",
"Raymond",
""
],
[
"Lecocq",
"Florent",
""
]
] |
2412.17394 | Hui Xiang | Hao Zheng, Zhibo Dai, Biyue Pan, Chunyang Wang, Baiyi Zhang, Hui Xiang, and Dixia Fan | AeroDiT: Diffusion Transformers for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows | null | null | null | null | physics.flu-dyn physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-time and accurate prediction of aerodynamic flow fields around airfoils is crucial for flow control and aerodynamic optimization. However, achieving this remains challenging due to the high computational costs and the non-linear nature of flow physics. Traditional Computational Fluid Dynamics (CFD) methods face limitations in balancing computational efficiency and accuracy, hindering their application in real-time scenarios. To address these challenges, this study presents AeroDiT, a novel surrogate model that integrates scalable diffusion models with transformer architectures to address these challenges. Trained on Reynolds-Averaged Navier-Stokes (RANS) simulation data for high Reynolds-number airfoil flows, AeroDiT accurately captures complex flow patterns while enabling real-time predictions. The model demonstrates impressive performance, with average relative L2 errors of 0.1, 0.025, and 0.050 for pressure p and velocity components ux, uy, confirming its reliability. The transformer-based structure allows for real-time predictions within seconds, outperforming traditional U-net diffusion models. This work underscores the potential of generative machine learning techniques to advance computational fluid dynamics, offering potential solutions to persistent challenges in simulating high-fidelity aerodynamic flows. | [
{
"version": "v1",
"created": "Mon, 23 Dec 2024 08:46:29 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 12:17:08 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zheng",
"Hao",
""
],
[
"Dai",
"Zhibo",
""
],
[
"Pan",
"Biyue",
""
],
[
"Wang",
"Chunyang",
""
],
[
"Zhang",
"Baiyi",
""
],
[
"Xiang",
"Hui",
""
],
[
"Fan",
"Dixia",
""
]
] |
2412.18570 | Katherine A. Maxwell | Katherine A. Maxwell, Alexander A. Voronov | Toward the Universal Mumford form on Sato Grassmannians | 19 pages, minor edits to Lie algebra basis vectors | null | null | MPIM-Bonn-2022 | math.AG hep-th | http://creativecommons.org/licenses/by/4.0/ | We construct a local universal Mumford form on a product of Sato Grassmannians using the flow of the Virasoro algebra. The existence of this universal Mumford form furthers the proposal that the Sato Grassmannian provides a universal moduli space with applications to string theory. Our approach using the Virasoro flow is an alternative to using the KP flow, which in particular allows for a bosonic universal Mumford form to be constructed. Applying the same method, we construct a local universal super Mumford form on a product of super Sato Grassmannians using the flow of the Neveu-Schwarz algebra. | [
{
"version": "v1",
"created": "Tue, 24 Dec 2024 17:50:52 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 06:05:04 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Maxwell",
"Katherine A.",
""
],
[
"Voronov",
"Alexander A.",
""
]
] |
2412.18585 | Katherine A. Maxwell | Katherine A. Maxwell, Alexander A. Voronov | The Neveu-Schwarz group and Schwarz's extended super Mumford form | 46 pages, minor edits to cocycle arguments | null | null | MPIM-Bonn-2022 | math.AG hep-th | http://creativecommons.org/licenses/by/4.0/ | In 1987, Albert Schwarz suggested a formula which extends the super Mumford form from the moduli space of super Riemann surfaces into the super Sato Grassmannian. His formula is a remarkably simple combination of super tau functions. We compute the Neveu-Schwarz action on super tau functions, and show that Schwarz's extended Mumford form is invariant under the the super Heisenberg-Neveu-Schwarz action, which strengthens Schwarz's proposal that a locus within the Grassmannian can serve as a universal moduli space with applications to superstring theory. Along the way, we construct the Neveu-Schwarz, super Witt, and super Heisenberg formal groups. | [
{
"version": "v1",
"created": "Tue, 24 Dec 2024 18:26:19 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 05:43:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Maxwell",
"Katherine A.",
""
],
[
"Voronov",
"Alexander A.",
""
]
] |
2412.18728 | Fabian Belmonte | Fabi\'an Belmonte and Giuseppe de Nittis | Explicit Spectral Analysis for Operators Representing the unitary group $\mathbb{U}(d)$ and its Lie algebra $\mathfrak{u}(d)$ through the Metaplectic Representation and Weyl Quantization | null | null | null | null | math.SP | http://creativecommons.org/licenses/by/4.0/ | In this article we compute and analyze the spectrum of operators defined by the metaplectic representation $\mu$ on the unitary group $\mathbb{U}(d)$ or operators defined by the corresponding induced representation $d\mu$ of the Lie algebra $\mathfrak{u}(d)$. It turns out that the point spectrum of both types of operators can be expressed in terms of the eigenvalues of the corresponding matrices. For each $A\in\mathfrak{u}(d)$, it is known that the selfadjoint operator $H_A=-i d\mu(A)$ has a quadratic Weyl symbol and we will give conditions on to guarantee that it has discrete spectrum. Under those conditions, using a known result in combinatorics, we show that the multiplicity of the eigenvalues of $H_A$ is (up to some explicit translation and scalar multiplication) a quasi polynomial of degree $d-1$. Moreover, we show that counting eigenvalues function behaves as an Ehrhart polynomial. Using the latter result, we prove a Weyl's law for the operators $H_A$. | [
{
"version": "v1",
"created": "Wed, 25 Dec 2024 01:03:54 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 00:43:21 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Belmonte",
"Fabián",
""
],
[
"de Nittis",
"Giuseppe",
""
]
] |
2412.18843 | Matteo Biagiola | Matteo Biagiola, Gianluca Ghislotti, Paolo Tonella | Improving the Readability of Automatically Generated Tests using Large Language Models | To be published in: 2025 IEEE Conference on Software Testing, Verification and Validation (ICST) | null | 10.1109/ICST62969.2025.10989020 | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Search-based test generators are effective at producing unit tests with high coverage. However, such automatically generated tests have no meaningful test and variable names, making them hard to understand and interpret by developers. On the other hand, large language models (LLMs) can generate highly readable test cases, but they are not able to match the effectiveness of search-based generators, in terms of achieved code coverage.
In this paper, we propose to combine the effectiveness of search-based generators with the readability of LLM generated tests. Our approach focuses on improving test and variable names produced by search-based tools, while keeping their semantics (i.e., their coverage) unchanged.
Our evaluation on nine industrial and open source LLMs show that our readability improvement transformations are overall semantically-preserving and stable across multiple repetitions. Moreover, a human study with ten professional developers, show that our LLM-improved tests are as readable as developer-written tests, regardless of the LLM employed. | [
{
"version": "v1",
"created": "Wed, 25 Dec 2024 09:08:53 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Biagiola",
"Matteo",
""
],
[
"Ghislotti",
"Gianluca",
""
],
[
"Tonella",
"Paolo",
""
]
] |
2412.19429 | Imteaz Rahaman Mr. | Imteaz Rahaman, Botong Li, Bobby Duersch, Hunter D. Ellis, and Kai Fu | Seed-Driven Stepwise Crystallization (SDSC) for Growing Rutile GeO2 Films via MOCVD | 17 pages, 6 figures, 1 Table | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Germanium dioxide (r-GeO2) is an emerging new ultrawide bandgap (UWBG) semiconductor with significant potential for power electronics, thanks to its large-size substrate compatibility and ambipolar doping capability. However, phase segregation during metal-organic chemical vapor deposition (MOCVD) on substrates like r-TiO2 has posed a significant barrier to achieving high-quality films. Conventional optimization of growth parameters has been found so far not very insufficient in film coverage and film quality. To address this, a seed-driven stepwise crystallization (SDSC) growth approach was employed in this study, featuring multiple sequential deposition steps on a pre-templated substrate enriched with r-GeO2 seeds. The process began with an initial 180-minute deposition to establish r-GeO2 nucleation seeds, followed by a sequence of shorter deposition steps (90, 60, 60, 60, 60, and 60 minutes). This stepwise growth strategy progressively increased the crystalline coverage to 57.4%, 77.49%, 79.73%, 93.27%, 99.17%, and ultimately 100%. Concurrently, the crystalline quality improved substantially, evidenced by a ~30% reduction in the Full Width at Half Maximum (FWHM) of X-ray diffraction rocking curves. These findings demonstrate the potential of the SDSC approach for overcoming phase segregation and achieving high-quality, large-area r-GeO2 films. | [
{
"version": "v1",
"created": "Fri, 27 Dec 2024 03:47:35 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 03:47:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Rahaman",
"Imteaz",
""
],
[
"Li",
"Botong",
""
],
[
"Duersch",
"Bobby",
""
],
[
"Ellis",
"Hunter D.",
""
],
[
"Fu",
"Kai",
""
]
] |
2412.19794 | Amit Agarwal | Amit Agarwal, Srikant Panda, Angeline Charles, Bhargava Kumar, Hitesh Patel, Priyaranjan Pattnayak, Taki Hasan Rafi, Tejaswini Kumar, Hansa Meghwani, Karan Gupta, Dong-Kyu Chae | MVTamperBench: Evaluating Robustness of Vision-Language Models | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | Multimodal Large Language Models (MLLMs), are recent advancement of Vision-Language Models (VLMs) that have driven major advances in video understanding. However, their vulnerability to adversarial tampering and manipulations remains underexplored. To address this gap, we introduce \textbf{MVTamperBench}, a benchmark that systematically evaluates MLLM robustness against five prevalent tampering techniques: rotation, masking, substitution, repetition, and dropping; based on real-world visual tampering scenarios such as surveillance interference, social media content edits, and misinformation injection. MVTamperBench comprises ~3.4K original videos, expanded into over ~17K tampered clips covering 19 distinct video manipulation tasks. This benchmark challenges models to detect manipulations in spatial and temporal coherence. We evaluate 45 recent MLLMs from 15+ model families. We reveal substantial variability in resilience across tampering types and show that larger parameter counts do not necessarily guarantee robustness. MVTamperBench sets a new benchmark for developing tamper-resilient MLLM in safety-critical applications, including detecting clickbait, preventing harmful content distribution, and enforcing policies on media platforms. We release all code, data, and benchmark to foster open research in trustworthy video understanding.
Code: https://amitbcp.github.io/MVTamperBench/ Data: https://huggingface.co/datasets/Srikant86/MVTamperBench | [
{
"version": "v1",
"created": "Fri, 27 Dec 2024 18:47:05 GMT"
},
{
"version": "v2",
"created": "Mon, 30 Dec 2024 04:20:52 GMT"
},
{
"version": "v3",
"created": "Wed, 15 Jan 2025 19:04:48 GMT"
},
{
"version": "v4",
"created": "Fri, 17 Jan 2025 18:18:21 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 16:24:02 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Agarwal",
"Amit",
""
],
[
"Panda",
"Srikant",
""
],
[
"Charles",
"Angeline",
""
],
[
"Kumar",
"Bhargava",
""
],
[
"Patel",
"Hitesh",
""
],
[
"Pattnayak",
"Priyaranjan",
""
],
[
"Rafi",
"Taki Hasan",
""
],
[
"Kumar",
"Tejaswini",
""
],
[
"Meghwani",
"Hansa",
""
],
[
"Gupta",
"Karan",
""
],
[
"Chae",
"Dong-Kyu",
""
]
] |
2412.19986 | Patricia Cahn | Ryan Blair, Patricia Cahn, Alexandra Kjuchukova, Hannah Schwartz | Rep-Tiles | 23 pages, 14 figures, 1 footnote, 1 ball number. Version 2 contains an improved exposition with some simplified proofs in Section 2. The ball number remains unchanged | null | null | null | math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An $n$-dimensional rep-tile is a compact, connected submanifold of $\mathbb{R}^n$ with non-empty interior which can be decomposed into pairwise isometric rescaled copies of itself whose interiors are disjoint. We show that every smooth compact $n$-dimensional submanifold of $\mathbb{R}^n$ with connected boundary is topologically isotopic to a polycube that tiles the $n$-cube, and hence is topologically isotopic to a rep-tile. Consequently, there is a rep-tile in the homotopy type of any finite CW complex. In addition to classifying rep-tiles in all dimensions up to isotopy, we also give new explicit constructions of rep-tiles in all dimensions, including examples in the homotopy type of any finite bouquet of spheres. | [
{
"version": "v1",
"created": "Sat, 28 Dec 2024 03:13:18 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 17:45:56 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Blair",
"Ryan",
""
],
[
"Cahn",
"Patricia",
""
],
[
"Kjuchukova",
"Alexandra",
""
],
[
"Schwartz",
"Hannah",
""
]
] |
2412.20367 | Yangfan He | Junqiao Wang, Zeng Zhang, Yangfan He, Zihao Zhang, Yuyang Song, Tianyu Shi, Yuchen Li, Hengyuan Xu, Kunyu Wu, Xin Yi, Zhongwei Wan, Xinhang Yuan, Kuan Lu, Menghao Huo, Tang Jingqun, Guangwu Qian, Keqin Li, Qiuwu Chen, Lewei He | Enhancing Code LLMs with Reinforcement Learning in Code Generation: A Survey | null | null | null | null | cs.SE cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reinforcement learning (RL) has emerged as a powerful paradigm for enhancing large language models (LLMs) in code generation and optimization. This survey systematically reviews RL-driven techniques across the code development lifecycle, from compiler-level optimizations and resource allocation strategies to end-to-end code synthesis frameworks. We first examine classical and modern RL algorithms -- spanning policy gradients, actor-critic methods, human-feedback alignment, and preference-based optimization -- and their adaptations to the unique challenges of code generation, such as sparse and delayed rewards. Next, we analyze key benchmarks, datasets, and evaluation metrics that drive progress in RL-augmented Code LLMs. Finally, we identify open problems, including the need for richer feedback sources, support for low-level and domain-specific languages, and methods to reduce computational overhead. By consolidating current insights and outlining future directions, this work aims to guide researchers and practitioners in leveraging RL to produce more robust, efficient, and human-aligned code generation systems. | [
{
"version": "v1",
"created": "Sun, 29 Dec 2024 06:15:41 GMT"
},
{
"version": "v2",
"created": "Thu, 02 Jan 2025 09:43:43 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 05:41:19 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 04:18:55 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wang",
"Junqiao",
""
],
[
"Zhang",
"Zeng",
""
],
[
"He",
"Yangfan",
""
],
[
"Zhang",
"Zihao",
""
],
[
"Song",
"Yuyang",
""
],
[
"Shi",
"Tianyu",
""
],
[
"Li",
"Yuchen",
""
],
[
"Xu",
"Hengyuan",
""
],
[
"Wu",
"Kunyu",
""
],
[
"Yi",
"Xin",
""
],
[
"Wan",
"Zhongwei",
""
],
[
"Yuan",
"Xinhang",
""
],
[
"Lu",
"Kuan",
""
],
[
"Huo",
"Menghao",
""
],
[
"Jingqun",
"Tang",
""
],
[
"Qian",
"Guangwu",
""
],
[
"Li",
"Keqin",
""
],
[
"Chen",
"Qiuwu",
""
],
[
"He",
"Lewei",
""
]
] |
2412.20600 | Madeleine Jotz | I. Ermeidis and M. Jotz | Deformations of ideals in Lie algebras | 41 pages; comments welcome! v2: typos fixes, Remark 6.22 added; submitted version | null | null | null | math.DG math.RA math.RT | http://creativecommons.org/licenses/by/4.0/ | This paper develops the deformation theory of Lie ideals. It shows that the smooth deformations of an ideal $\mathfrak i$ in a Lie algebra $\mathfrak g$ differentiate to cohomology classes in the cohomology of $\mathfrak g$ with values in its adjoint representation on $\operatorname{Hom}(\mathfrak i, \mathfrak g/\mathfrak i)$. The cohomology associated with the ideal $\mathfrak i$ in $\mathfrak g$ is compared with other Lie algebra cohomologies defined by $\mathfrak i$, such as the cohomology defined by $\mathfrak i$ as a Lie subalgebra of $\mathfrak g$ (Richardson, 1969), and the cohomology defined by the Lie algebra morphism $\mathfrak g \to \mathfrak g/\mathfrak i$.
After a choice of complement of the ideal $\mathfrak i$ in the Lie algebra $\mathfrak g$, its deformation complex is enriched to the differential graded Lie algebra that controls its deformations, in the sense that its Maurer-Cartan elements are in one-to-one correspondence with the (small) deformations of the ideal. Furthermore, the $L_{\infty}$-algebra that simultaneously controls the deformations of $\mathfrak{i}$ and of the ambient Lie bracket is identified.
Under appropriate assumptions on the low degrees of the deformation cohomology of a given Lie ideal, the (topological) rigidity and stability of ideals are studied, as well as obstructions to deformations of ideals of Lie algebras. | [
{
"version": "v1",
"created": "Sun, 29 Dec 2024 22:12:45 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:56:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ermeidis",
"I.",
""
],
[
"Jotz",
"M.",
""
]
] |
2412.20862 | Hayato Shimabukuro | Hayato Shimabukuro | Nonlinear reconstruction of 21cm global signal from 21cm power spectrum with artificial neural networks | 22 pages, 11 figures, accepted in Research in Astronomy \& Astrophysics | null | 10.1088/1674-4527/ade34f | null | astro-ph.CO | http://creativecommons.org/publicdomain/zero/1.0/ | In this paper, we propose a novel method to recover the 21cm global signal from the 21cm power spectrum using artificial neural networks (ANNs). The 21cm global signal is crucial for understanding cosmic evolution from the Dark Ages through the Epoch of Reionization (EoR). While interferometers like LOFAR, MWA, HERA, and SKA focus on detecting the 21cm power spectrum, single-dish experiments such as EDGES target the global signal. Our method utilizes ANNs to establish a connection between these two observables, providing a means to cross-validate independent 21cm line observations. This capability is significant as it allows different observational approaches to verify each other's results, ensuring greater reliability in 21cm cosmology. We demonstrate that our ANN-based approach can accurately recover the 21cm global signal across a wide redshift range (z=7.5-35) from simulated data, even when realistic thermal noise levels, such as those expected from SKA-1, are considered. This cross-validation process strengthens the robustness of 21cm signal analysis, offering a more comprehensive understanding of the early universe. | [
{
"version": "v1",
"created": "Mon, 30 Dec 2024 11:02:20 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 12:59:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Shimabukuro",
"Hayato",
""
]
] |
2501.00654 | Xindi Wu | Xindi Wu, Mengzhou Xia, Rulin Shao, Zhiwei Deng, Pang Wei Koh, Olga Russakovsky | ICONS: Influence Consensus for Vision-Language Data Selection | 31 pages, 19 figures | null | null | null | cs.CV cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Training vision-language models via instruction tuning often relies on large mixtures of data spanning diverse tasks and domains. However, these mixtures frequently include redundant information, increasing computational costs without proportional performance gains, necessitating more effective data selection strategies. Existing methods typically rely on task-agnostic heuristics to estimate data importance or focus on optimizing single tasks in isolation, limiting their effectiveness in multitask settings. In this work, we introduce ICONS, a gradient-based Influence CONsensus approach for vision-language data Selection. Our method leverages first-order training dynamics to estimate the influence of individual training examples on validation performance and aggregates these estimates across tasks via majority voting over task-specific influences. This cross-task consensus identifies data points that are consistently valuable across tasks, enabling us to prioritize examples that drive overall performance. The voting-based design further mitigates issues such as score calibration and outlier sensitivity, resulting in robust and scalable data selection for diverse multitask mixtures. With only 20% of the data from LLaVA-665K and Cambrian-7M, our selected subsets retain 98.6% and 98.8% of the performance achieved with full datasets, and can even surpass full data training at a 60% selection ratio on LLaVA-665K. Our approach also generalizes to unseen tasks and architectures, demonstrating strong transfer. We release two compact, high-utility subsets, LLaVA-ICONS-133K and Cambrian-ICONS-1.4M, preserving impactful training examples for efficient and scalable vision-language model development. | [
{
"version": "v1",
"created": "Tue, 31 Dec 2024 21:33:38 GMT"
},
{
"version": "v2",
"created": "Mon, 06 Jan 2025 18:17:30 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 18:19:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wu",
"Xindi",
""
],
[
"Xia",
"Mengzhou",
""
],
[
"Shao",
"Rulin",
""
],
[
"Deng",
"Zhiwei",
""
],
[
"Koh",
"Pang Wei",
""
],
[
"Russakovsky",
"Olga",
""
]
] |
2501.00695 | Xiaoda Qu | Xiaoda Qu, Baba C. Vemuri | Theory and Applications of Kernel Stein's Method on Riemannian Manifolds | null | null | null | null | math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | Distributional comparison is a fundamental problem in statistical data analysis with numerous applications in a variety of scientific and engineering fields. Numerous methods exist for distributional comparison but kernel Stein's method has gained significant popularity in recent times. In this paper, we first present a novel mathematically rigorous and consistent generalization of the Stein operator to Riemannian manifolds. Then we show that the kernel Stein discrepancy (KSD) defined via this operator is nearly as strong as the KSD in the Euclidean setting in terms of distinguishing the target distributions from the reference. We investigate the asymptotic properties of the minimum kernel Stein discrepancy estimator (MKSDE), apply it to goodness-of-fit testing, and compare it to the maximum likelihood estimator (MLE) experimentally. We present several examples of our theory applied to commonly encountered Riemannian manifolds in practice namely, the n-sphere, the Grassmann, Stiefel, the manifold of symmetric positive definite matrices and other Riemannian homogeneous spaces. On the aforementioned manifolds, we consider a variety of distributions with intractable normalization constants and derive closed form expressions for the KSD and MKSDE. | [
{
"version": "v1",
"created": "Wed, 01 Jan 2025 01:21:38 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 23:05:23 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Qu",
"Xiaoda",
""
],
[
"Vemuri",
"Baba C.",
""
]
] |
2501.00829 | Shuo Li | Haoxiang Tian, Xingshuo Han, Guoquan Wu, An Guo, Yuan Zhou. Jie Zhang, Shuo Li, Jun Wei, Tianwei Zhang | An LLM-Empowered Adaptive Evolutionary Algorithm For Multi-Component Deep Learning Systems | 9 | null | null | null | cs.NE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-objective evolutionary algorithms (MOEAs) are widely used for searching optimal solutions in complex multi-component applications. Traditional MOEAs for multi-component deep learning (MCDL) systems face challenges in enhancing the search efficiency while maintaining the diversity. To combat these, this paper proposes $\mu$MOEA, the first LLM-empowered adaptive evolutionary search algorithm to detect safety violations in MCDL systems. Inspired by the context-understanding ability of Large Language Models (LLMs), $\mu$MOEA promotes the LLM to comprehend the optimization problem and generate an initial population tailed to evolutionary objectives. Subsequently, it employs adaptive selection and variation to iteratively produce offspring, balancing the evolutionary efficiency and diversity. During the evolutionary process, to navigate away from the local optima, $\mu$MOEA integrates the evolutionary experience back into the LLM. This utilization harnesses the LLM's quantitative reasoning prowess to generate differential seeds, breaking away from current optimal solutions. We evaluate $\mu$MOEA in finding safety violations of MCDL systems, and compare its performance with state-of-the-art MOEA methods. Experimental results show that $\mu$MOEA can significantly improve the efficiency and diversity of the evolutionary search. | [
{
"version": "v1",
"created": "Wed, 01 Jan 2025 13:19:58 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 04:44:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tian",
"Haoxiang",
""
],
[
"Han",
"Xingshuo",
""
],
[
"Wu",
"Guoquan",
""
],
[
"Guo",
"An",
""
],
[
"Zhang",
"Yuan Zhou. Jie",
""
],
[
"Li",
"Shuo",
""
],
[
"Wei",
"Jun",
""
],
[
"Zhang",
"Tianwei",
""
]
] |
2501.01070 | Rohan Pattnaik | Rohan Pattnaik, Jeyhan S. Kartaltepe, Clive Binu | SpecPT (Spectroscopy Pre-trained Transformer) Model for Extragalactic Spectroscopy: I. Architecture and Automated Redshift Measurement | 17 pages, 13 figures. Submitted to ApJ | null | 10.3847/1538-4357/ade053 | null | astro-ph.IM astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We introduce the Spectroscopy Pre-trained Transformer (SpecPT), a transformer-based model designed to analyze spectroscopic data, with applications in spectrum reconstruction and redshift measurement. Using the Early Data Release (EDR) of the DESI survey, we evaluate SpecPT's performance on two distinct datasets: the Bright Galaxy Survey (BGS) and Emission Line Galaxy (ELG) samples. SpecPT successfully reconstructs spectra, accurately capturing emission lines, absorption features, and continuum shapes while effectively reducing noise. For redshift prediction, SpecPT achieves competitive accuracy, with Normalized Median Absolute Deviation (NMAD) values of 0.0006 and 0.0008, and catastrophic outlier fractions of 0.20% and 0.80% for BGS and ELG, respectively. Notably, SpecPT performs consistently well across the full redshift range ($0 < z < 1.6$), demonstrating its versatility and robustness. By leveraging its learned latent representations, SpecPT lays the groundwork for a foundational spectroscopic model, with potential applications in outlier detection, interstellar medium (ISM) property estimation, and transfer learning to other datasets. This work represents a first step in building a generalized framework for spectroscopic analysis, capable of scaling to the full DESI dataset and beyond. | [
{
"version": "v1",
"created": "Thu, 02 Jan 2025 05:40:39 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pattnaik",
"Rohan",
""
],
[
"Kartaltepe",
"Jeyhan S.",
""
],
[
"Binu",
"Clive",
""
]
] |
2501.01271 | Mohd Saif Ali Khan | Mohd Saif Ali Khan, Karthik RM, Samar Agnihotri | Energy-and Spectral-Efficiency Trade-off in Distributed Massive-MIMO Networks | null | null | null | null | cs.NI cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | This paper investigates a fundamental yet under-explored trade-off between energy efficiency (EE) and spectral efficiency (SE) in distributed massive MIMO (D-mMIMO) systems. Unlike conventional EE-SE trade-off studies that primarily focus on transmission power, D-mMIMO systems introduce new energy consumption factors including fronthaul signaling and distributed signal processing, which are heavily influenced by AP-UE association. This work highlights the critical need for a system-level EE-SE trade-off framework that accounts for these unique aspects of D-mMIMO. We formulate a joint optimization problem that maximizes EE while satisfying uplink sum-SE constraints, through the coordinated design of power allocation and AP-UE association strategies. By explicitly considering both transmission and infrastructure-related energy costs, our approach enables energy-aware network design without compromising throughput. Numerical simulations demonstrate the substantial impact of dynamic AP-UE association and power control on the EE-SE trade-off, providing actionable insights for an efficient deployment of large-scale distributed MIMO networks in next-generation wireless systems. | [
{
"version": "v1",
"created": "Thu, 02 Jan 2025 14:11:44 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Jan 2025 14:37:36 GMT"
},
{
"version": "v3",
"created": "Thu, 30 Jan 2025 11:01:46 GMT"
},
{
"version": "v4",
"created": "Mon, 09 Jun 2025 17:12:15 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 15:41:56 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Khan",
"Mohd Saif Ali",
""
],
[
"RM",
"Karthik",
""
],
[
"Agnihotri",
"Samar",
""
]
] |
2501.02121 | Tanner Ragan | Tanner Ragan, Mark Warnecke, Samuel T. Stout, and Perry L. Johnson | Turbulence-Resolving Integral Simulations for Wall-Bounded Flows | 25 pages, 9 figures, journal paper | null | null | null | physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The physical fidelity of turbulence models can benefit from a partial resolution of fluctuations, but doing so often comes with an increase in computational cost. To explore this trade-off in the context of wall-bounded flows, this paper introduces a framework for Turbulence-Resolving Integral Simulations (TRIS) with the goal of efficiently resolving the largest motions using a two-dimensional, three component representation of the flow defined by instantaneous wall-normal integrals of velocity and pressure. Self-sustaining turbulence with qualitatively realistic large-scale structures is demonstrated for TRIS on an open-channel (half-channel) flow configuration using moment-of-momentum integral equations derived from Navier-Stokes with relatively simple closure approximations. Evidence from Direct Numerical Simulations (DNS) suggests that TRIS can theoretically resolve 35-40% of the turbulent skin friction enhancement for friction Reynolds numbers between 180 and 5200, without a noticeable decrease or increase as a function of Reynolds number. The current implementation of TRIS can match this resolution while simulating one flow through time in ~1 minute on a single processor, even for very large Reynolds numbers. The framework facilitates a detailed apples-to-apples comparison of predicted statistics against data from DNS. Comparisons at friction Reynolds numbers of 395 and 590 show that TRIS generates a relatively accurate representation of the flow, while highlight discrepancies that demonstrate a need for improving the closure models. The present results for open-channel flow represent a proof of concept for TRIS as a new approach for wall-bounded turbulence modeling, motivating extension to more general flow configurations such as boundary layers on immersed objects. | [
{
"version": "v1",
"created": "Fri, 03 Jan 2025 22:15:23 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 21:58:37 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ragan",
"Tanner",
""
],
[
"Warnecke",
"Mark",
""
],
[
"Stout",
"Samuel T.",
""
],
[
"Johnson",
"Perry L.",
""
]
] |
2501.02436 | Hong Zhao | Yuchen Lin, Yong Zhang, Sihan Feng, Hong Zhao | Network Dynamics-Based Framework for Understanding Deep Neural Networks | 12 pages, 7 figures | null | null | null | cs.LG nlin.CD stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Advancements in artificial intelligence call for a deeper understanding of the fundamental mechanisms underlying deep learning. In this work, we propose a theoretical framework to analyze learning dynamics through the lens of dynamical systems theory. We redefine the notions of linearity and nonlinearity in neural networks by introducing two fundamental transformation units at the neuron level: order-preserving transformations and non-order-preserving transformations. Different transformation modes lead to distinct collective behaviors in weight vector organization, different modes of information extraction, and the emergence of qualitatively different learning phases. Transitions between these phases may occur during training, accounting for key phenomena such as grokking. To further characterize generalization and structural stability, we introduce the concept of attraction basins in both sample and weight spaces. The distribution of neurons with different transformation modes across layers, along with the structural characteristics of the two types of attraction basins, forms a set of core metrics for analyzing the performance of learning models. Hyperparameters such as depth, width, learning rate, and batch size act as control variables for fine-tuning these metrics. Our framework not only sheds light on the intrinsic advantages of deep learning, but also provides a novel perspective for optimizing network architectures and training strategies. | [
{
"version": "v1",
"created": "Sun, 05 Jan 2025 04:23:21 GMT"
},
{
"version": "v2",
"created": "Thu, 06 Mar 2025 15:49:50 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 14:48:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lin",
"Yuchen",
""
],
[
"Zhang",
"Yong",
""
],
[
"Feng",
"Sihan",
""
],
[
"Zhao",
"Hong",
""
]
] |
2501.02597 | Giulio Bartoli | Andrea Abrardo, Giulio Bartoli, Alberto Toccafondi | A Novel Comprehensive Multiport Network Model for Stacked Intelligent Metasurfaces (SIM) Characterization and Optimization | 15 pages, 11 figures | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reconfigurable Intelligent Surfaces (RIS) represent transformative technologies for next-generation wireless communications, offering advanced control over electromagnetic wave propagation. While RIS have been extensively studied, Stacked Intelligent Metasurfaces (SIM), which extend the RIS concept to multi-layered systems, present significant modeling and optimization challenges. This work addresses these challenges by introducing an optimization framework for SIM that, unlike previous approaches, is based on a comprehensive model without relying on specific assumptions, allowing for broader applicability of the results. We first present a model based on multi-port network theory for characterizing a general electromagnetic collaborative object (ECO) and derive a framework for ECO optimization. We then introduce the SIM as an ECO with a specific architecture, offering insights into SIM optimization for various configurations and discussing the complexities associated with each case. Finally, we demonstrate that the comprehensive model considered in this work simplifies to the model traditionally used in the literature when the assumption of unilateral propagation between the levels of the SIM is made, and mutual coupling between the SIM elements is neglected. To assess the applicability of these assumptions, a case study focused on the realization of a 2D DFT was undertaken. In this context, we highlight that these assumptions introduce a significant mismatch between the SIM model and its behavior as described by the complete model, making these approximations inadequate for optimizing the SIM. Conversely, we show that employing the complete model proposed in this paper can yield excellent performance. | [
{
"version": "v1",
"created": "Sun, 05 Jan 2025 16:39:21 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Jan 2025 10:20:12 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 16:04:42 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Abrardo",
"Andrea",
""
],
[
"Bartoli",
"Giulio",
""
],
[
"Toccafondi",
"Alberto",
""
]
] |
2501.03874 | Ning Zhang | Ning Zhang, Timothy Shea, Arto Nurmikko | Neuromorphic Optical Tracking and Imaging of Randomly Moving Targets through Strongly Scattering Media | 26 pages, 6 figures | null | null | null | cs.NE cs.CV cs.LG eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tracking and acquiring simultaneous optical images of randomly moving targets obscured by scattering media remains a challenging problem of importance to many applications that require precise object localization and identification. In this work we develop an end-to-end neuromorphic optical engineering and computational approach to demonstrate how to track and image normally invisible objects by combining an event detecting camera with a multistage neuromorphic deep learning strategy. Photons emerging from dense scattering media are detected by the event camera and converted to pixel-wise asynchronized spike trains - a first step in isolating object-specific information from the dominant uninformative background. Spiking data is fed into a deep spiking neural network (SNN) engine where object tracking and image reconstruction are performed by two separate yet interconnected modules running in parallel in discrete time steps over the event duration. Through benchtop experiments we demonstrate tracking and imaging randomly moving objects in dense turbid media as well as image reconstruction of spatially stationary but optically dynamic objects. Standardized character sets serve as representative proxies for geometrically complex objects, underscoring the method's generality. The results highlight the advantages of a fully neuromorphic approach in meeting a major imaging technology with high computational efficiency and low power consumption. | [
{
"version": "v1",
"created": "Tue, 07 Jan 2025 15:38:13 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 00:14:31 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhang",
"Ning",
""
],
[
"Shea",
"Timothy",
""
],
[
"Nurmikko",
"Arto",
""
]
] |
2501.04606 | Yangfan He | Yangfan He, Sida Li, Jianhui Wang, Kun Li, Xinyuan Song, Xinhang Yuan, Keqin Li, Kuan Lu, Menghao Huo, Jingqun Tang, Yi Xin, Jiaqi Chen, Miao Zhang, Xueqian Wang | Enhancing Low-Cost Video Editing with Lightweight Adaptors and Temporal-Aware Inversion | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in text-to-image (T2I) generation using diffusion models have enabled cost-effective video-editing applications by leveraging pre-trained models, eliminating the need for resource-intensive training. However, the frame-independence of T2I generation often results in poor temporal consistency. Existing methods address this issue through temporal layer fine-tuning or inference-based temporal propagation, but these approaches suffer from high training costs or limited temporal coherence. To address these challenges, we propose a General and Efficient Adapter (GE-Adapter) that integrates temporal-spatial and semantic consistency with Baliteral DDIM inversion. This framework introduces three key components: (1) Frame-based Temporal Consistency Blocks (FTC Blocks) to capture frame-specific features and enforce smooth inter-frame transitions via temporally-aware loss functions; (2) Channel-dependent Spatial Consistency Blocks (SCD Blocks) employing bilateral filters to enhance spatial coherence by reducing noise and artifacts; and (3) Token-based Semantic Consistency Module (TSC Module) to maintain semantic alignment using shared prompt tokens and frame-specific tokens. Our method significantly improves perceptual quality, text-image alignment, and temporal coherence, as demonstrated on the MSR-VTT dataset. Additionally, it achieves enhanced fidelity and frame-to-frame coherence, offering a practical solution for T2V editing. | [
{
"version": "v1",
"created": "Wed, 08 Jan 2025 16:41:31 GMT"
},
{
"version": "v2",
"created": "Mon, 21 Apr 2025 05:14:31 GMT"
},
{
"version": "v3",
"created": "Tue, 22 Apr 2025 18:14:08 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 04:23:47 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"He",
"Yangfan",
""
],
[
"Li",
"Sida",
""
],
[
"Wang",
"Jianhui",
""
],
[
"Li",
"Kun",
""
],
[
"Song",
"Xinyuan",
""
],
[
"Yuan",
"Xinhang",
""
],
[
"Li",
"Keqin",
""
],
[
"Lu",
"Kuan",
""
],
[
"Huo",
"Menghao",
""
],
[
"Tang",
"Jingqun",
""
],
[
"Xin",
"Yi",
""
],
[
"Chen",
"Jiaqi",
""
],
[
"Zhang",
"Miao",
""
],
[
"Wang",
"Xueqian",
""
]
] |
2501.05099 | Yingpeng Liu | Yingpeng Liu, Jiao Wu, Kesheng Xu and Muhua Zheng | Recovery of activation propagation and self-sustained oscillation abilities in stroke brain networks | 20 pages, 13 figures | null | 10.1103/PhysRevE.111.034309 | null | physics.soc-ph q-bio.NC q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Healthy brain networks usually show highly efficient information communication and self-sustained oscillation abilities. However, how the brain network structure affects these dynamics after an injury (stroke) is not very clear. The recovery of structure and dynamics of stroke brain networks over time is still not known precisely. Based on the analysis of a large number of strokes' brain network data, we show that stroke changes the network properties in connection weights, average degree, clustering, community, etc. Yet, they will recover gradually over time to some extent. We then adopt a simplified reaction-diffusion model to investigate stroke patients' activation propagation and self-sustained oscillation abilities. Our results reveal that the stroke slows the adoption time across different brain scales, indicating a weakened brain's activation propagation ability. In addition, we show that the lifetime of self-sustained oscillatory patterns at three months post-stroke patients' brains significantly departs from the healthy one. Finally, we examine the properties of core networks of self-sustained oscillatory patterns, in which the directed edges denote the main pathways of activation propagation. Our results demonstrate that the lifetime and recovery of self-sustaining patterns are related to the properties of core networks, and the properties in the post-stroke greatly vary from those in the healthy group. Most importantly, the strokes' activation propagation and self-sustained oscillation abilities significantly improve at one year post-stroke, driven by structural connection repair. This work may help us to understand the relationship between structure and function in brain disorders. | [
{
"version": "v1",
"created": "Thu, 09 Jan 2025 09:38:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Liu",
"Yingpeng",
""
],
[
"Wu",
"Jiao",
""
],
[
"Xu",
"Kesheng",
""
],
[
"Zheng",
"Muhua",
""
]
] |
2501.05534 | Henning Rose | Joschka Birk, Frank Gaede, Anna Hallin, Gregor Kasieczka, Martina Mozzanica, Henning Rose | OmniJet-${\alpha}_C$: Learning point cloud calorimeter simulations using generative transformers | null | null | null | null | hep-ph cs.LG hep-ex physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show the first use of generative transformers for generating calorimeter showers as point clouds in a high-granularity calorimeter. Using the tokenizer and generative part of the OmniJet-${\alpha}$ model, we represent the hits in the detector as sequences of integers. This model allows variable-length sequences, which means that it supports realistic shower development and does not need to be conditioned on the number of hits. Since the tokenization represents the showers as point clouds, the model learns the geometry of the showers without being restricted to any particular voxel grid. | [
{
"version": "v1",
"created": "Thu, 09 Jan 2025 19:16:41 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 15:49:41 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Birk",
"Joschka",
""
],
[
"Gaede",
"Frank",
""
],
[
"Hallin",
"Anna",
""
],
[
"Kasieczka",
"Gregor",
""
],
[
"Mozzanica",
"Martina",
""
],
[
"Rose",
"Henning",
""
]
] |
2501.05830 | Dan Rust | Gandhar Joshi, Dan Rust | Monochromatic arithmetic progressions in the Fibonacci, Thue-Morse, and Rudin-Shapiro words | Changes made to address feedback from anonymous referee. Accepted for publication in Theoretical Computer Science | null | 10.1016/j.tcs.2025.115391 | null | math.DS cs.FL math.CO | http://creativecommons.org/licenses/by/4.0/ | We investigate the lengths and starting positions of the longest monochromatic arithmetic progressions for a fixed difference in the Fibonacci word. We provide a complete classification for their lengths in terms of a simple formula. Our strongest results are proved using methods from dynamical systems, especially the dynamics of circle rotations. We also employ computer-based methods in the form of the automatic theorem-proving software Walnut. This allows us to extend recent results concerning similar questions for the Thue-Morse word and the Rudin-Shapiro word. This also allows us to obtain some results for the Fibonacci word that do not seem to be amenable to dynamical methods. | [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 10:09:33 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:28:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Joshi",
"Gandhar",
""
],
[
"Rust",
"Dan",
""
]
] |
2501.06137 | Manuel Cebrian | Manuel Cebrian, Emilia Gomez, David Fernandez Llorca | Supervision policies can shape long-term risk management in general-purpose AI models | 24 pages, 14 figures | null | null | null | cs.AI cs.CY cs.SI | http://creativecommons.org/licenses/by/4.0/ | The rapid proliferation and deployment of General-Purpose AI (GPAI) models, including large language models (LLMs), present unprecedented challenges for AI supervisory entities. We hypothesize that these entities will need to navigate an emergent ecosystem of risk and incident reporting, likely to exceed their supervision capacity. To investigate this, we develop a simulation framework parameterized by features extracted from the diverse landscape of risk, incident, or hazard reporting ecosystems, including community-driven platforms, crowdsourcing initiatives, and expert assessments. We evaluate four supervision policies: non-prioritized (first-come, first-served), random selection, priority-based (addressing the highest-priority risks first), and diversity-prioritized (balancing high-priority risks with comprehensive coverage across risk types). Our results indicate that while priority-based and diversity-prioritized policies are more effective at mitigating high-impact risks, particularly those identified by experts, they may inadvertently neglect systemic issues reported by the broader community. This oversight can create feedback loops that amplify certain types of reporting while discouraging others, leading to a skewed perception of the overall risk landscape. We validate our simulation results with several real-world datasets, including one with over a million ChatGPT interactions, of which more than 150,000 conversations were identified as risky. This validation underscores the complex trade-offs inherent in AI risk supervision and highlights how the choice of risk management policies can shape the future landscape of AI risks across diverse GPAI models used in society. | [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 17:52:34 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 23:58:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cebrian",
"Manuel",
""
],
[
"Gomez",
"Emilia",
""
],
[
"Llorca",
"David Fernandez",
""
]
] |
2501.06758 | Luca Pelizzari | Christian Bayer, Luca Pelizzari, Jia-Jie Zhu | Pricing American options under rough volatility using deep-signatures and signature-kernels | null | null | null | null | q-fin.MF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend the signature-based primal and dual solutions to the optimal stopping problem recently introduced in [Bayer et al.: Primal and dual optimal stopping with signatures, to appear in Finance & Stochastics 2025], by integrating deep-signature and signature-kernel learning methodologies. These approaches are designed for non-Markovian frameworks, in particular enabling the pricing of American options under rough volatility. We demonstrate and compare the performance within the popular rough Heston and rough Bergomi models. | [
{
"version": "v1",
"created": "Sun, 12 Jan 2025 09:40:49 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:08:41 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Bayer",
"Christian",
""
],
[
"Pelizzari",
"Luca",
""
],
[
"Zhu",
"Jia-Jie",
""
]
] |
2501.06864 | Zhen Zhong | Zhen Zhong | Fisher's Randomization Test for Causality with General Types of Treatments | null | null | null | null | stat.ME | http://creativecommons.org/licenses/by/4.0/ | We extend Fisher's randomization test (FRT) to test conditional independence between observed outcomes and treatments given covariates in both randomized experiments and observational studies, with no restriction on the variable type of treatments. Under a generalized unconfoundedness assumption, we provide causal identification for this hypothesis. Our approach requires neither the no-interference nor the positive overlap assumption, making it a widely applicable tool for detecting causal effects. A unique advantage of FRT lies in the separated roles of assignment and outcome models. The former, whether known from randomized experiments or estimated in observational studies, guarantees valid Type I error control at least asymptotically. The latter, even if misspecified, is used to construct optimal test statistics derived from Bayes factors. The synthesis of two classes of models through FRT yields a calibrated Bayesian procedure with desired frequentist properties. Recognizing that the generalized unconfoundedness assumption is untestable in observational studies, we develop a novel sensitivity analysis to assess the robustness of causal conclusions to unobserved confounding. Through a re-analysis of a panel dataset, we show how our methods can be integrated into a pipeline for observational causal inference. | [
{
"version": "v1",
"created": "Sun, 12 Jan 2025 16:25:05 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 02:18:53 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhong",
"Zhen",
""
]
] |
2501.07042 | Hiroshi Suzuki | Okuto Morikawa and Hiroshi Suzuki | Direct Monte Carlo Computation of the 't~Hooft Partition Function | 11 pages, 3 figures. The final version to appear in PTEP | null | null | RIKEN-iTHEMS-Report-25, KYUSHU-HET-308 | hep-lat | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The 't~Hooft partition function~$\mathcal{Z}_{\text{tH}}[E;B]$ of an $SU(N)$ gauge theory with the $\mathbb{Z}_N$ 1-form symmetry is defined as the Fourier transform of the partition function~$\mathcal{Z}[B]$ with respect to the spatial-temporal components of the 't~Hooft flux~$B$. Its large volume behavior detects the quantum phase of the system. When the integrand of the functional integral is real-positive, the latter partition function~$\mathcal{Z}[B]$ can be numerically computed by a Monte Carlo simulation of the $SU(N)/\mathbb{Z}_N$ gauge theory, just by counting the number of configurations of a specific 't~Hooft flux~$B$. We carry out this program for the $SU(2)$ pure Yang--Mills theory with the vanishing $\theta$-angle by employing a newly-developed hybrid Monte Carlo (HMC) algorithm (the halfway HMC) for the $SU(N)/\mathbb{Z}_N$ gauge theory. The numerical result clearly shows that all non-electric fluxes are ``light'' as expected in the ordinary confining phase with the monopole condensate. Invoking the Witten effect on~$\mathcal{Z}_{\text{tH}}[E;B]$, this also indicates the oblique confinement at~$\theta=2\pi$ with the dyon condensate. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 03:48:11 GMT"
},
{
"version": "v2",
"created": "Thu, 06 Mar 2025 12:09:54 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:44:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Morikawa",
"Okuto",
""
],
[
"Suzuki",
"Hiroshi",
""
]
] |
2501.07231 | Peter Reinholdt | Peter Reinholdt, Karl Michael Ziems, Erik Rosendahl Kjellgren, Sonia Coriani, Stephan P. A. Sauer, Jacob Kongsted | Critical Limitations in Quantum-Selected Configuration Interaction Methods | 36 pages, 6 figures | null | null | null | physics.chem-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum Selected Configuration Interaction (QSCI) methods (also known as Sample-based Quantum Diagonalization, SQD) have emerged as promising near-term approaches to solving the electronic Schr{\"o}dinger equation with quantum computers. In this work, we perform numerical analysis to show that QSCI methods face critical limitations that severely hinder their practical applicability in chemistry. Using the nitrogen molecule and the iron-sulfur cluster [2Fe-2S] as examples, we demonstrate that while QSCI can, in principle, yield high-quality configuration interaction (CI) expansions similar to classical SCI heuristics in some cases, the method struggles with inefficiencies in finding new determinants as sampling repeatedly selects already seen configurations. This inefficiency becomes especially pronounced when targeting high-accuracy results or sampling from an approximate ansatz. In cases where the sampling problem is not present, the resulting CI expansions are less compact than those generated from classical heuristics, rendering QSCI an overall more expensive method. Our findings suggest a significant drawback in QSCI methods when sampling from the ground-state distribution as the inescapable trade-off between finding sufficiently many determinants and generating compact, accurate CI expansions. This ultimately hinders utility in quantum chemistry applications, as QSCI falls behind more efficient classical counterparts. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 11:30:43 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Feb 2025 08:32:41 GMT"
},
{
"version": "v3",
"created": "Thu, 08 May 2025 11:52:39 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 11:02:24 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Reinholdt",
"Peter",
""
],
[
"Ziems",
"Karl Michael",
""
],
[
"Kjellgren",
"Erik Rosendahl",
""
],
[
"Coriani",
"Sonia",
""
],
[
"Sauer",
"Stephan P. A.",
""
],
[
"Kongsted",
"Jacob",
""
]
] |
2501.07792 | Yili Wang | Yi-Li Wang, Young-Kwon Han, Xian-Hui Ge, Sang-Jin Sin | Hall Angle of a Spatially Random Vector Model | 27 pages, 15 figures | null | null | null | hep-th cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Strange metals exhibit linear resistivity and anomalous Hall transport, yet a comprehensive theory that accounts for both phenomena is still lacking. Recent studies have shown SYK-like spatially random couplings between a Fermi surface and a bosonic field, either scalar or vector type, can yield linear-$T$ resistivity. In this paper, we continue the investigation on a vector coupling in the presence of a magnetic field. We compute the fermion and boson propagators, along with the self-energy and polarization functions, and determine their dependence on the magnetic field. Although the Hall angle does not exhibit the signature of strange-metal, the linear-in-temperature resistivity remains at low temperatures. Results indicate that random interactions can robustly support linear transport, though additional ingredients may be required to capture the full phenomenology of strange metals. | [
{
"version": "v1",
"created": "Tue, 14 Jan 2025 02:21:14 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Apr 2025 02:13:48 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 07:28:17 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 02:09:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wang",
"Yi-Li",
""
],
[
"Han",
"Young-Kwon",
""
],
[
"Ge",
"Xian-Hui",
""
],
[
"Sin",
"Sang-Jin",
""
]
] |
2501.08279 | Longtao Jiang | Longtao Jiang, Zhendong Wang, Jianmin Bao, Wengang Zhou, Dongdong Chen, Lei Shi, Dong Chen, Houqiang Li | SmartEraser: Remove Anything from Images using Masked-Region Guidance | Project at: https://longtaojiang.github.io/smarteraser.github.io/ | The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2025 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object removal has so far been dominated by the mask-and-inpaint paradigm, where the masked region is excluded from the input, leaving models relying on unmasked areas to inpaint the missing region. However, this approach lacks contextual information for the masked area, often resulting in unstable performance. In this work, we introduce SmartEraser, built with a new removing paradigm called Masked-Region Guidance. This paradigm retains the masked region in the input, using it as guidance for the removal process. It offers several distinct advantages: (a) it guides the model to accurately identify the object to be removed, preventing its regeneration in the output; (b) since the user mask often extends beyond the object itself, it aids in preserving the surrounding context in the final result. Leveraging this new paradigm, we present Syn4Removal, a large-scale object removal dataset, where instance segmentation data is used to copy and paste objects onto images as removal targets, with the original images serving as ground truths. Experimental results demonstrate that SmartEraser significantly outperforms existing methods, achieving superior performance in object removal, especially in complex scenes with intricate compositions. | [
{
"version": "v1",
"created": "Tue, 14 Jan 2025 17:55:12 GMT"
},
{
"version": "v2",
"created": "Sat, 29 Mar 2025 09:36:55 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 09:08:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Jiang",
"Longtao",
""
],
[
"Wang",
"Zhendong",
""
],
[
"Bao",
"Jianmin",
""
],
[
"Zhou",
"Wengang",
""
],
[
"Chen",
"Dongdong",
""
],
[
"Shi",
"Lei",
""
],
[
"Chen",
"Dong",
""
],
[
"Li",
"Houqiang",
""
]
] |
2501.09171 | Monnie McGee | Monnie McGee and Bivin Sadler | Generative AI Takes a Statistics Exam: A Comparison of Performance between ChatGPT3.5, ChatGPT4, and ChatGPT4o-mini | 24 pages, 2 figures, 3 tables. Submitted for publication August, 2024; revision submitted January 2025 | J. data sci. 23(2025), no. 2, 269-286 | 10.6339/25-JDS1174 | null | stat.OT cs.LG | http://creativecommons.org/licenses/by/4.0/ | Many believe that use of generative AI as a private tutor has the potential to shrink access and achievement gaps between students and schools with abundant resources versus those with fewer resources. Shrinking the gap is possible only if paid and free versions of the platforms perform with the same accuracy. In this experiment, we investigate the performance of GPT versions 3.5, 4.0, and 4o-mini on the same 16-question statistics exam given to a class of first-year graduate students. While we do not advocate using any generative AI platform to complete an exam, the use of exam questions allows us to explore aspects of ChatGPT's responses to typical questions that students might encounter in a statistics course. Results on accuracy indicate that GPT 3.5 would fail the exam, GPT4 would perform well, and GPT4o-mini would perform somewhere in between. While we acknowledge the existence of other Generative AI/LLMs, our discussion concerns only ChatGPT because it is the most widely used platform on college campuses at this time. We further investigate differences among the AI platforms in the answers for each problem using methods developed for text analytics, such as reading level evaluation and topic modeling. Results indicate that GPT3.5 and 4o-mini have characteristics that are more similar than either of them have with GPT4. | [
{
"version": "v1",
"created": "Wed, 15 Jan 2025 21:46:01 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"McGee",
"Monnie",
""
],
[
"Sadler",
"Bivin",
""
]
] |
2501.09772 | Sergei Gninenko | S.N. Gninenko, N.V. Krasnikov, V.A. Matveev | Search for $K_{S,L}$ oscillations and invisible decays into the dark sector at NA64 | 23 pages, 2 figures. Invited review partly based on arXiv:1409.2288;1503.01595;1602.03548 | Natural Science Review 1, 5 (2024) | 10.54546/NaturalSciRev.100105 | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | The decays $K_{S,L}$ have never been experimentally tested. In the Standard Model their branching ratios for the decay into two neutrinos are predicted to be extremely small, $Br(K_{S,L} \to \nu \bar{\nu}) \lesssim 10^{-16}$. We consider several natural extensions of the SM, such as two-Higgs-doublet (2HDM), 2HDM and light scalar, and dark mirror sector models, that allow to enhance the $Br(K_{S,L})$ up to a measurable level. We briefly discuss the possible search for $K_{S,L}$ decays and $K_{S,L}$ oscillations into the dark sector at the NA64 experiment at CERN with the sensitivity to $Br(K_{S,L}) \lesssim 10^{-7}-10^{-5}$. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 07:17:23 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gninenko",
"S. N.",
""
],
[
"Krasnikov",
"N. V.",
""
],
[
"Matveev",
"V. A.",
""
]
] |
2501.09810 | Om Gupta | Om Gupta, Paz Beniamini, Pawan Kumar, Steven L. Finkelstein | The Cosmic Evolution of Fast Radio Bursts Inferred from the CHIME/FRB Baseband Catalog 1 | Data & visualization scripts available: https://github.com/omguptaup/frb-cosmic-evol. 30 pages, 11 figures, 3 tables | ApJ 986 100 (2025) | 10.3847/1538-4357/add14c | null | astro-ph.HE astro-ph.CO astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Redshift and luminosity distributions are essential for understanding the cosmic evolution of extragalactic objects and phenomena, such as galaxies, gamma-ray bursts, and fast radio bursts (FRBs). For FRBs, these distributions are primarily estimated using the fluence and the Dispersion Measure (DM). Calibrating their joint distribution has been challenging due to a lack of accurate fluences in the intensity data of the CHIME/FRB survey. Using the baseband update of CHIME/FRB Catalog 1, we calibrate the 2D fluence-DM distribution for the first time. We find the energy distribution is described well by a Schechter function with power-law slope of $-1.94^{+0.14}_{-0.12}$. Testing two types of redshift evolution models suggests a likely combination of young and old formation channels. $31^{+31}_{-21}$% of FRB sources may track star formation, or correspondingly, FRB sources may have delay times of $1.94^{+1.54}_{-1.31}$ Gyr. A pure star formation tracking population is excluded by only one model at $> 2\sigma$ confidence. An updated cosmic star formation rate density evolution up to redshift 14 is constrained by compiling results from several JWST studies. The furthest FRB detection with planned radio facilities is expected to be at $z \approx 5$. A radio telescope operating at 200 MHz with a system-equivalent flux density of $\leq 0.07$ Jy (equivalent to a detection threshold of 1 mJy ms) and instantaneous sky coverage of $\gtrsim 400$ square degrees should be able to detect $630^{+730}_{-485}$ FRBs year$^{-1}$ at $z \gtrsim 6$ and $53^{+83}_{-43}$ FRBs year$^{-1}$ at $z\gtrsim 8$, which is sufficient to differentiate between reionization histories. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 19:46:13 GMT"
},
{
"version": "v2",
"created": "Fri, 25 Apr 2025 15:24:44 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 01:38:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gupta",
"Om",
""
],
[
"Beniamini",
"Paz",
""
],
[
"Kumar",
"Pawan",
""
],
[
"Finkelstein",
"Steven L.",
""
]
] |
2501.10362 | Finn Klessascheck | Finn Klessascheck, Luise Pufahl | Reviewing Uses of Regulatory Compliance Monitoring | null | null | null | null | cs.CY cs.DB | http://creativecommons.org/licenses/by/4.0/ | Organizations need to manage numerous business processes for delivering their services and products to customers. One important consideration thereby lies in the adherence to regulations such as laws, guidelines, or industry standards. In order to monitor adherence of their business processes to regulations -- in other words, their regulatory compliance -- organizations make use of various techniques that draw on process execution data of IT systems that support these processes. Previous research has investigated conformance checking, an operation of process mining, for the domains in which it is applied, its operationalization of regulations, the techniques being used, and the presentation of results produced. However, other techniques for regulatory compliance monitoring, which we summarize as compliance checking techniques, have not yet been investigated regarding these aspects in a structural manner. To this end, this work presents a systematic literature review on uses of regulatory compliance monitoring of business processes, thereby offering insights into the various techniques being used, their application and the results they generate. We highlight commonalities and differences between the approaches and find that various steps are performed manually; we also provide further impulses for research on compliance monitoring and its use in practice. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 15:54:36 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 16:55:44 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:54:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Klessascheck",
"Finn",
""
],
[
"Pufahl",
"Luise",
""
]
] |
2501.10935 | Shuai Lyu | Shuai Lyu, Zijing Tian, Zhonghong Ou, Yifan Zhu, Xiao Zhang, Qiankun Ha, Haoran Luo, Meina Song | TSVC:Tripartite Learning with Semantic Variation Consistency for Robust Image-Text Retrieval | This paper has been accepted to the Main Track of AAAI 2025. It contains 9 pages, 7 figures, and is relevant to the areas of cross-modal retrieval and machine learning. The work presents a novel approach in robust image-text retrieval using a tripartite learning framework | Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 39, No. 18, pp. 19269-19277, 2025 | 10.1609/aaai.v39i18.34121 | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Cross-modal retrieval maps data under different modality via semantic relevance. Existing approaches implicitly assume that data pairs are well-aligned and ignore the widely existing annotation noise, i.e., noisy correspondence (NC). Consequently, it inevitably causes performance degradation. Despite attempts that employ the co-teaching paradigm with identical architectures to provide distinct data perspectives, the differences between these architectures are primarily stemmed from random initialization. Thus, the model becomes increasingly homogeneous along with the training process. Consequently, the additional information brought by this paradigm is severely limited. In order to resolve this problem, we introduce a Tripartite learning with Semantic Variation Consistency (TSVC) for robust image-text retrieval. We design a tripartite cooperative learning mechanism comprising a Coordinator, a Master, and an Assistant model. The Coordinator distributes data, and the Assistant model supports the Master model's noisy label prediction with diverse data. Moreover, we introduce a soft label estimation method based on mutual information variation, which quantifies the noise in new samples and assigns corresponding soft labels. We also present a new loss function to enhance robustness and optimize training effectiveness. Extensive experiments on three widely used datasets demonstrate that, even at increasing noise ratios, TSVC exhibits significant advantages in retrieval accuracy and maintains stable training performance. | [
{
"version": "v1",
"created": "Sun, 19 Jan 2025 04:05:08 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:33:05 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lyu",
"Shuai",
""
],
[
"Tian",
"Zijing",
""
],
[
"Ou",
"Zhonghong",
""
],
[
"Zhu",
"Yifan",
""
],
[
"Zhang",
"Xiao",
""
],
[
"Ha",
"Qiankun",
""
],
[
"Luo",
"Haoran",
""
],
[
"Song",
"Meina",
""
]
] |
2501.11223 | Robert Gerstenberger | Maciej Besta, Julia Barth, Eric Schreiber, Ales Kubicek, Afonso Catarino, Robert Gerstenberger, Piotr Nyczyk, Patrick Iff, Yueling Li, Sam Houliston, Tomasz Sternal, Marcin Copik, Grzegorz Kwa\'sniewski, J\"urgen M\"uller, {\L}ukasz Flis, Hannes Eberhard, Zixuan Chen, Hubert Niewiadomski, Torsten Hoefler | Reasoning Language Models: A Blueprint | null | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reasoning language models (RLMs), also known as Large Reasoning Models (LRMs), such as OpenAI's o1 and o3, DeepSeek-R1, and Alibaba's QwQ, have redefined AI's problem-solving capabilities by extending LLMs with advanced reasoning mechanisms. Yet, their high costs, proprietary nature, and complex architectures - uniquely combining reinforcement learning (RL), search heuristics, and LLMs - present accessibility and scalability challenges. To address these, we propose a comprehensive blueprint that organizes RLM components into a modular framework, based on a survey and analysis of all RLM works. This blueprint incorporates diverse reasoning structures (chains, trees, graphs, and nested forms), reasoning strategies (e.g., Monte Carlo Tree Search, Beam Search), RL concepts (policy, value models and others), supervision schemes (Outcome-Based and Process-Based Supervision), and other related concepts (e.g., Test-Time Compute, Retrieval-Augmented Generation, agent tools). We also provide detailed mathematical formulations and algorithmic specifications to simplify RLM implementation. By showing how schemes like LLaMA-Berry, QwQ, Journey Learning, and Graph of Thoughts fit as special cases, we demonstrate the blueprint's versatility and unifying potential. To illustrate its utility, we introduce x1, a modular implementation for rapid RLM prototyping and experimentation. Using x1 and a literature review, we provide key insights, such as multi-phase training for policy and value models, and the importance of familiar training distributions. Finally, we discuss scalable RLM cloud deployments and we outline how RLMs can integrate with a broader LLM ecosystem. Our work demystifies RLM construction, democratizes advanced reasoning capabilities, and fosters innovation, aiming to mitigate the gap between "rich AI" and "poor AI" by lowering barriers to RLM design and experimentation. | [
{
"version": "v1",
"created": "Mon, 20 Jan 2025 02:16:19 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jan 2025 17:49:37 GMT"
},
{
"version": "v3",
"created": "Thu, 23 Jan 2025 14:26:08 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 13:19:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Besta",
"Maciej",
""
],
[
"Barth",
"Julia",
""
],
[
"Schreiber",
"Eric",
""
],
[
"Kubicek",
"Ales",
""
],
[
"Catarino",
"Afonso",
""
],
[
"Gerstenberger",
"Robert",
""
],
[
"Nyczyk",
"Piotr",
""
],
[
"Iff",
"Patrick",
""
],
[
"Li",
"Yueling",
""
],
[
"Houliston",
"Sam",
""
],
[
"Sternal",
"Tomasz",
""
],
[
"Copik",
"Marcin",
""
],
[
"Kwaśniewski",
"Grzegorz",
""
],
[
"Müller",
"Jürgen",
""
],
[
"Flis",
"Łukasz",
""
],
[
"Eberhard",
"Hannes",
""
],
[
"Chen",
"Zixuan",
""
],
[
"Niewiadomski",
"Hubert",
""
],
[
"Hoefler",
"Torsten",
""
]
] |
2501.11261 | Adam Wunderlich | Adam Wunderlich, Aric Sanders | The Expected Peak-to-Average Power Ratio of White Gaussian Noise in Sampled I/Q Data | 8 pages, 6 figures | IEEE Transactions on Instrumentation and Measurement, Vol. 74, 2025 | 10.1109/TIM.2025.3544288 | null | eess.SP stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the fundamental endeavors in radio frequency (RF) metrology is to measure the power of signals, where a common aim is to estimate the peak-to-average power ratio (PAPR), which quantifies the ratio of the maximum (peak) to the mean value. For a finite number of discrete-time samples of baseband in-phase and quadrature (I/Q) white Gaussian noise (WGN) that are independent and identically distributed with zero mean, we derive a closed-form, exact formula for mean PAPR that is well-approximated by the natural logarithm of the number of samples plus Euler's constant. Additionally, we give related theoretical results for the mean crest factor (CF). After comparing our main result to previously published approximate formulas, we examine how violations of the WGN assumptions in sampled I/Q data result in deviations from the expected value of PAPR. Finally, utilizing a measured RF I/Q acquisition, we illustrate how our formula for mean PAPR can be applied to spectral analysis with spectrograms to verify when measured RF emissions are WGN in a given frequency band. | [
{
"version": "v1",
"created": "Mon, 20 Jan 2025 04:01:41 GMT"
},
{
"version": "v2",
"created": "Thu, 06 Mar 2025 23:40:17 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 21:05:48 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wunderlich",
"Adam",
""
],
[
"Sanders",
"Aric",
""
]
] |
2501.12270 | Yoshiaki Yasuno | Naoki Fukutake, Shuichi Makita, Yoshiaki Yasuno | Unified image formation theory for microscopy and optical coherence tomography in 4-D space-time | null | null | 10.1364/OE.560890 | null | physics.optics | http://creativecommons.org/licenses/by-sa/4.0/ | We construct an image formation theory that covers the majority of optical microscopy techniques that use diverse coherent or incoherent light-matter interactions. The theories of individual microscopy methods could not previously be connected with other systems because of the absence of a common theoretical framework. Using the general principles of quantum physics and applying a four-dimensional representation, we unify the image formation theories of optical systems ranging from classical microscopy to cutting-edge instruments into a single framework in which light is replaced with quantum fields and the interactions are represented using double-sided Feynman diagrams. Our universal methodology requires a four-dimensional aperture that enables sufficient understanding of the associations between the different imaging types and interprets image formation appropriately for all systems, including optical coherence tomography. | [
{
"version": "v1",
"created": "Tue, 21 Jan 2025 16:41:18 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Fukutake",
"Naoki",
""
],
[
"Makita",
"Shuichi",
""
],
[
"Yasuno",
"Yoshiaki",
""
]
] |
2501.12372 | Yeounoh Chung | Yeounoh Chung, Gaurav T. Kakkar, Yu Gan, Brenton Milne, Fatma Ozcan | Is Long Context All You Need? Leveraging LLM's Extended Context for NL2SQL | 13 pages, 6 figures, VLDB 2025 | null | 10.14778/3742728.3742761 | null | cs.DB cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLMs) have demonstrated impressive capabilities across a range of natural language processing tasks. In particular, improvements in reasoning abilities and the expansion of context windows have opened new avenues for leveraging these powerful models. NL2SQL is challenging in that the natural language question is inherently ambiguous, while the SQL generation requires a precise understanding of complex data schema and semantics. One approach to this semantic ambiguous problem is to provide more and sufficient contextual information.
In this work, we explore the performance and the latency trade-offs of the extended context window (a.k.a., long context) offered by Google's state-of-the-art LLM (\textit{gemini-1.5-pro}). We study the impact of various contextual information, including column example values, question and SQL query pairs, user-provided hints, SQL documentation, and schema. To the best of our knowledge, this is the first work to study how the extended context window and extra contextual information can help NL2SQL generation with respect to both accuracy and latency cost. We show that long context LLMs are robust and do not get lost in the extended contextual information. Additionally, our long-context NL2SQL pipeline based on Google's \textit{gemini-pro-1.5} achieve strong performances on various benchmark datasets without finetuning and expensive self-consistency based techniques. | [
{
"version": "v1",
"created": "Tue, 21 Jan 2025 18:52:15 GMT"
},
{
"version": "v2",
"created": "Sat, 01 Feb 2025 02:00:46 GMT"
},
{
"version": "v3",
"created": "Thu, 13 Feb 2025 23:39:12 GMT"
},
{
"version": "v4",
"created": "Fri, 07 Mar 2025 23:17:42 GMT"
},
{
"version": "v5",
"created": "Thu, 20 Mar 2025 17:39:13 GMT"
},
{
"version": "v6",
"created": "Wed, 11 Jun 2025 17:49:37 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chung",
"Yeounoh",
""
],
[
"Kakkar",
"Gaurav T.",
""
],
[
"Gan",
"Yu",
""
],
[
"Milne",
"Brenton",
""
],
[
"Ozcan",
"Fatma",
""
]
] |
2501.12426 | Giovanni Gallavotti | Giovanni Gallavotti | Nonequilibrium and Irreversibility | Preprint of the II edition of the book with the same title. See arXiv:1311.6448. I edition: Springer, Series "Theoretical and Mathematical Physics", 2014, ISBN: 978-3-319-06757-5 (Print) 978-3-319-06758-2 (Online). arXiv admin note: substantial text overlap with arXiv:1311.6448 | null | 10.1007/978-3-031-85497-2 | FM-2025 | nlin.CD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The work concentrates on relations, which are general and model independent in chaotic system, between time averages of a few (typically {\it very few}) observables. Equilibrium thermodynamics provides a guide and here is attempted to argue that the viewpoint of Sinai-Ruelle-Bowen can be regarded as a generalization to nonequilibrum phenomena of the theory of the ensembles proposing an answer to classical question like which distributions describe the statistics of stationary states (hence extend the analysis selecting canonical, or equivalent distributions, equilibrim between the uncountably many possibilities). The special name "Chaothic Hypothesis" (CH) is given to the above attempt and its mathematical meaning is discussed. General properties are presented and applied (eg. 'Fluctuation Theorem', 'Fluctuation Patterns', 'Pairing Symmetry') and related to the basic Time Reversal symmetry: which presents irreversibility as due to chaotic motion rather than to viscous forces. The case of a simple incompressible fluid is discussed in some detail. The possibility that CH is violated in various cases is considered: and in the end it is suggested that CH is the paradigm of chaotic evolution, as the harmonic oscillators are a paradigm of ordered motions, but of course {\it tertium datur}. The exposition is informal and often restricted to heuristic analysis, with detailed references to the literature and attention to numerical simulations and importance of stressing strongly the discrete models of Physics, trying to imitate the vision of Boltzmann, is widely considered. | [
{
"version": "v1",
"created": "Tue, 21 Jan 2025 12:27:43 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gallavotti",
"Giovanni",
""
]
] |
2501.12439 | Giulia Despali | Giulia Despali, Lauro Moscardini, Dylan Nelson, Annalisa Pillepich, Volker Springel, Mark Vogelsberger | Introducing the AIDA-TNG project: galaxy formation in alternative dark matter models | 20 pages, 18 figures, accepted for publication in A&A. More information: https://gdespali.github.io/AIDA/ | A&A 697, A213 (2025) | 10.1051/0004-6361/202553836 | null | astro-ph.CO astro-ph.GA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We introduce the AIDA-TNG project, a suite of cosmological magnetohydrodynamic simulations that simultaneously model galaxy formation and different variations of the underlying dark matter model. We consider the standard cold dark matter model and five variations, including three warm dark matter scenarios and two self-interacting models with constant or velocity-dependent cross-section. In each model, we simulate two cosmological boxes of 51.7 and 110.7 Mpc on a side, with the same initial conditions as TNG50 and TNG100, and combine the variations in the physics of dark matter with the fiducial IllustrisTNG galaxy formation model. The AIDA-TNG runs are thus ideal for studying the simultaneous effect of baryons and alternative dark matter models on observable properties of galaxies and large-scale structures. We resolve haloes in the range between $10^{8}$ and $4\times10^{14}\,$M$_{\odot}$ and scales down to the nominal resolution of 570 pc in the highest resolution runs. This work presents the first results on statistical quantities such as the halo mass function and the matter power spectrum; we quantify the modification in the number of haloes and the power on scales smaller than 1 Mpc, due to the combination of baryonic and dark matter physics. Despite being calibrated on cold dark matter, we find that the TNG galaxy formation model can produce a realistic galaxy population in all scenarios. The stellar and gas mass fraction, stellar mass function, black hole mass as a function of stellar mass and star formation rate density are very similar in all dark matter models, with some deviations only in the most extreme warm dark matter model. Finally, we also quantify changes in halo structure due to warm and self-interacting dark matter, which appear in the density profiles, concentration-mass relation and galaxy sizes. | [
{
"version": "v1",
"created": "Tue, 21 Jan 2025 19:00:00 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 10:12:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Despali",
"Giulia",
""
],
[
"Moscardini",
"Lauro",
""
],
[
"Nelson",
"Dylan",
""
],
[
"Pillepich",
"Annalisa",
""
],
[
"Springel",
"Volker",
""
],
[
"Vogelsberger",
"Mark",
""
]
] |
2501.13001 | Daniele Catone | G. Ammirati, S. Turchini, F. Toschi, P. O'Keeffe, A. Paladini, F. Martelli, R. Khanfar, D. Takhellambam, S. Pescetelli, A. Agresti and D. Catone | Hole Transfer Dynamics in Thin Films of Mixed-Dimensional Quasi-2D Perovskites | null | null | 10.1002/adfm.202502825 | null | cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The understanding of charge transfer processes in mixed-dimensional quasi-2D perovskites is crucial for their application in high-performance photovoltaic devices. In this work, we investigate the link between charge transport dynamics and morphology in a thin film of quasi-2D perovskites (PEA2MAn-1PbnI3n+1), grown with a distinct dimensionality gradient, where the n=1 phase is concentrated near the substrate and phases with higher dimensionality progressively increase in concentration toward the surface. By selectively exciting the n=4 phase, we observe efficient hole transfer to the n=2 and n=3 phases occurring within few tens of picoseconds after excitation. In contrast, the n=1 phase acts as a hole-blocking layer, limiting the overall charge transport efficiency. These results emphasize the critical importance of minimizing or eliminating the n=1 layer to enhance charge carrier separation and transport, offering valuable insights into the optimization of quasi-2D perovskite-based solar cells. | [
{
"version": "v1",
"created": "Wed, 22 Jan 2025 16:40:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ammirati",
"G.",
""
],
[
"Turchini",
"S.",
""
],
[
"Toschi",
"F.",
""
],
[
"O'Keeffe",
"P.",
""
],
[
"Paladini",
"A.",
""
],
[
"Martelli",
"F.",
""
],
[
"Khanfar",
"R.",
""
],
[
"Takhellambam",
"D.",
""
],
[
"Pescetelli",
"S.",
""
],
[
"Agresti",
"A.",
""
],
[
"Catone",
"D.",
""
]
] |
2501.13184 | Cristian Barbero | Cristian Barbero, \'I\~nigo J. Sola, Benjam\'in Alonso | Ptychographic retrieval for complete ultrashort pulse amplitude swing reconstruction | 10 pages, 5 figures | Optics and Laser Technology, 188, 112939 (2025) | 10.1016/j.optlastec.2025.112939 | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ultrafast experiments require precise temporal characterization of laser pulses, where pulse reconstruction is typically achieved through iterative retrieval algorithms. In this context, the amplitude swing technique has emerged as a robust and versatile tool for measuring scalar and vector pulses, using a compact inline setup. Here, we develop a unified theory of amplitude swing and introduce a ptychographic iterative engine algorithm to retrieve scalar and vector pulses from the different amplitude swing implementations (conventional and generalizing). For scalar pulses, we show the new capability of retrieving both amplitude and phase. The algorithm is validated with simulated and experimental traces of scalar and vector pulses, which are benefited from the new algorithm. In conclusion, a powerful tool for pulse characterization is provided, extending the capabilities of amplitude swing (e.g., to double pulses) and enhancing the ultrafast optics measurement toolkit. | [
{
"version": "v1",
"created": "Wed, 22 Jan 2025 19:31:26 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Barbero",
"Cristian",
""
],
[
"Sola",
"Íñigo J.",
""
],
[
"Alonso",
"Benjamín",
""
]
] |
2501.13736 | Cheuk Ting Li | Cheuk Ting Li | Discrete Layered Entropy, Conditional Compression and a Tighter Strong Functional Representation Lemma | 33 pages, 8 figures | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a quantity called discrete layered entropy, which approximates the Shannon entropy within a logarithmic gap. Compared to the Shannon entropy, the discrete layered entropy is piecewise linear, approximates the expected length of the optimal one-to-one non-prefix code, and satisfies an elegant conditioning property. These properties make it useful for approximating the Shannon entropy in linear programming and maximum entropy problems, studying the optimal length of conditional encoding, and bounding the entropy of monotonic mixture distributions. In particular, it can give a bound $I(X;Y)+\log(I(X;Y)+3.4)+1$ for the strong functional representation lemma that significantly improves upon the best known bound. | [
{
"version": "v1",
"created": "Thu, 23 Jan 2025 15:12:11 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 20:19:03 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Li",
"Cheuk Ting",
""
]
] |
2501.14244 | Satori Tsuzuki Ph.D | Satori Tsuzuki | Multi-scale physics of cryogenic liquid helium-4: Inverse coarse-graining properties of smoothed particle hydrodynamics | null | null | 10.1088/2399-6528/ade35f | null | physics.flu-dyn cond-mat.mes-hall cond-mat.other physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Our recent numerical studies on cryogenic liquid helium-4 highlight key features of multiscale physics that can be captured using the two-fluid model. In this paper, we demonstrated that classical and quantum hydrodynamic two-fluid models are connected via scale transformations: large eddy simulation (LES) filtering links microscopic to macroscopic scales, while inverse scale transformation through SPH connects macro back to microscales. We showed that the spin angular momentum conservation term, introduced as a quantum-like correction, formally corresponds to a subgrid-scale (SGS) model derived from this transformation. Moreover, solving the classical hydrodynamic two-fluid model with SPH appears to reproduce microscopic-scale fluctuations at macroscopic scales. The amplitude of these fluctuations depends on the kernel radius. This effect may arise from truncation errors from kernel smoothing, which can qualitatively resemble such fluctuations. However, this resemblance lacks first-principle justification and should be viewed as a speculative analogy rather than a physically grounded effect. Our theoretical analysis further suggests that the Condiff viscosity model can act as an SGS model, incorporating quantum vortex interactions under point-vortex approximation into the two-fluid framework. These findings provide new insight into the microscopic structure of cryogenic helium-4 within a multiscale context. Notably, the normal fluid can be understood as a mixture of inviscid and viscous fluid particles. While molecular viscosity renders the normal fluid at microscopic scales, its small magnitude contributes little to the large-scale effective viscosity, which includes both molecular and eddy viscosities; therefore, in laminar regimes where eddy viscosity is negligible, the normal fluid may be effectively treated as inviscid at large scales if molecular viscosity is sufficiently small. | [
{
"version": "v1",
"created": "Fri, 24 Jan 2025 05:09:58 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Feb 2025 18:31:21 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Feb 2025 20:00:57 GMT"
},
{
"version": "v4",
"created": "Sat, 01 Mar 2025 18:20:43 GMT"
},
{
"version": "v5",
"created": "Mon, 10 Mar 2025 17:43:14 GMT"
},
{
"version": "v6",
"created": "Fri, 23 May 2025 19:25:06 GMT"
},
{
"version": "v7",
"created": "Thu, 05 Jun 2025 11:42:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tsuzuki",
"Satori",
""
]
] |
2501.16564 | Dinh Tuan Van | Dinh Van Tuan and Hanan Dery | Landau-level composition of bound exciton states in magnetic field | 22 pages, 8 figures. We welcome your feedback | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a theory that studies the state composition of a bound exciton in magnetic field. Using a basis set made of products of free electron and hole wavefunctions in Landau gauge, we derive a secular equation which shows the relation between Landau levels (LLs) of the electron and hole when a bound exciton is formed. Focusing on excitons in the light cone, we establish a scattering selection rule for the interaction of an electron in LL $n_\text{e}$ with a hole in LL $n_\text{h}$. We solve the resulting secular equation and identify a simple pairing law, $n_\text{e} = n_\text{h} + l$, which informs us on the construction of a bound exciton state with magnetic quantum number $l$, and on the interaction of the exciton magnetic moment with magnetic field. We obtain good agreement between theory results and recent measurements of the diamagnetic shifts of exciton states in WSe$_2$ monolayers. | [
{
"version": "v1",
"created": "Mon, 27 Jan 2025 23:15:28 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Feb 2025 16:32:37 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:33:05 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Van Tuan",
"Dinh",
""
],
[
"Dery",
"Hanan",
""
]
] |
2501.16583 | Long Peng | Long Peng, Xin Di, Zhanfeng Feng, Wenbo Li, Renjing Pei, Yang Wang, Xueyang Fu, Yang Cao, Zheng-Jun Zha | Directing Mamba to Complex Textures: An Efficient Texture-Aware State Space Model for Image Restoration | Accepted by the 34th International Joint Conference on Artificial Intelligence (IJCAI 2025) | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image restoration aims to recover details and enhance contrast in degraded images. With the growing demand for high-quality imaging (\textit{e.g.}, 4K and 8K), achieving a balance between restoration quality and computational efficiency has become increasingly critical. Existing methods, primarily based on CNNs, Transformers, or their hybrid approaches, apply uniform deep representation extraction across the image. However, these methods often struggle to effectively model long-range dependencies and largely overlook the spatial characteristics of image degradation (regions with richer textures tend to suffer more severe damage), making it hard to achieve the best trade-off between restoration quality and efficiency. To address these issues, we propose a novel texture-aware image restoration method, TAMambaIR, which simultaneously perceives image textures and achieves a trade-off between performance and efficiency. Specifically, we introduce a novel Texture-Aware State Space Model, which enhances texture awareness and improves efficiency by modulating the transition matrix of the state-space equation and focusing on regions with complex textures. Additionally, we design a {Multi-Directional Perception Block} to improve multi-directional receptive fields while maintaining low computational overhead. Extensive experiments on benchmarks for image super-resolution, deraining, and low-light image enhancement demonstrate that TAMambaIR achieves state-of-the-art performance with significantly improved efficiency, establishing it as a robust and efficient framework for image restoration. | [
{
"version": "v1",
"created": "Mon, 27 Jan 2025 23:53:49 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 04:43:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Peng",
"Long",
""
],
[
"Di",
"Xin",
""
],
[
"Feng",
"Zhanfeng",
""
],
[
"Li",
"Wenbo",
""
],
[
"Pei",
"Renjing",
""
],
[
"Wang",
"Yang",
""
],
[
"Fu",
"Xueyang",
""
],
[
"Cao",
"Yang",
""
],
[
"Zha",
"Zheng-Jun",
""
]
] |
2501.16884 | Peiling Yi | Peiling Yi, Yuhan Xia, Yunfei Long | Irony Detection, Reasoning and Understanding in Zero-shot Learning | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The generalisation of irony detection faces significant challenges, leading to substantial performance deviations when detection models are applied to diverse real-world scenarios. In this study, we find that irony-focused prompts, as generated from our IDADP framework for LLMs, can not only overcome dataset-specific limitations but also generate coherent, human-readable reasoning, transforming ironic text into its intended meaning. Based on our findings and in-depth analysis, we identify several promising directions for future research aimed at enhancing LLMs' zero-shot capabilities in irony detection, reasoning, and comprehension. These include advancing contextual awareness in irony detection, exploring hybrid symbolic-neural methods, and integrating multimodal data, among others. | [
{
"version": "v1",
"created": "Tue, 28 Jan 2025 12:13:07 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:02:02 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Yi",
"Peiling",
""
],
[
"Xia",
"Yuhan",
""
],
[
"Long",
"Yunfei",
""
]
] |
2501.17398 | Khemraj Gautam Kshetri | Khemraj Gautam Kshetri, Amneet Pal Singh Bhalla, Nitesh Nama | Consistent continuum equations and numerical benchmarks for a perturbation-based, variable-coefficient acoustofluidic solver | null | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | We present a consistent continuum framework and a variable-coefficient acoustofluidic solver to analyze acoustic streaming. A perturbation approach is used to split the compressible Navier-Stokes equations into two sub-systems: a first-order harmonic system and a time-averaged second-order mean system. Prior acoustofluidic numerical studies have typically employed simplifying assumptions regarding the second-order mass balance equation and boundary conditions. These assumptions -- frequently left unjustified -- result in an incongruent problem statement where the boundary conditions are not consistent with the governing equations. To clarify these assumptions and mitigate the associated confusion, we systematically formulate the second-order mass balance equation into two analytically equivalent but numerically distinct forms by introducing the fluid's Lagrangian and mass transport velocity. We clarify the relation between these quantities to illustrate that a zero mass transport velocity does not necessarily imply a zero Lagrangian velocity. A second-order accurate finite difference/volume scheme is used to discretize the governing equations and boundary conditions expressed in Eulerian form. We solve the first-order coupled Helmholtz system via a sparse direct solver while the second-order system is solved via a Krylov (FGMRES) solver with a projection method-based preconditioner. To ascertain the spatial accuracy and consistency of numerical implementation of the split system of equations and boundary conditions, we demonstrate verification results and convergence rates using manufactured solutions. We provide benchmark test cases to demonstrate the impact of the choice of boundary conditions on second-order flow field. Lastly, we examine a system with spatially varying density to illustrate the difference in flow fields between constant and variable density systems. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 03:28:59 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 18:34:01 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kshetri",
"Khemraj Gautam",
""
],
[
"Bhalla",
"Amneet Pal Singh",
""
],
[
"Nama",
"Nitesh",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.