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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.10934 | Nikhil Krishnaswamy | Timothy Obiso, Kenneth Lai, Abhijnan Nath, Nikhil Krishnaswamy, James Pustejovsky | Dynamic Epistemic Friction in Dialogue | 11 pages, 2 figures, 2 tables, CoNLL 2025 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent developments in aligning Large Language Models (LLMs) with human preferences have significantly enhanced their utility in human-AI collaborative scenarios. However, such approaches often neglect the critical role of "epistemic friction," or the inherent resistance encountered when updating beliefs in response to new, conflicting, or ambiguous information. In this paper, we define dynamic epistemic friction as the resistance to epistemic integration, characterized by the misalignment between an agent's current belief state and new propositions supported by external evidence. We position this within the framework of Dynamic Epistemic Logic (Van Benthem and Pacuit, 2011), where friction emerges as nontrivial belief-revision during the interaction. We then present analyses from a situated collaborative task that demonstrate how this model of epistemic friction can effectively predict belief updates in dialogues, and we subsequently discuss how the model of belief alignment as a measure of epistemic resistance or friction can naturally be made more sophisticated to accommodate the complexities of real-world dialogue scenarios. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:41:00 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Obiso",
"Timothy",
""
],
[
"Lai",
"Kenneth",
""
],
[
"Nath",
"Abhijnan",
""
],
[
"Krishnaswamy",
"Nikhil",
""
],
[
"Pustejovsky",
"James",
""
]
] |
2506.10935 | Maxim Rakhuba | Ekaterina Grishina, Matvey Smirnov and Maxim Rakhuba | Accelerating Newton-Schulz Iteration for Orthogonalization via Chebyshev-type Polynomials | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The problem of computing optimal orthogonal approximation to a given matrix has attracted growing interest in machine learning. Notable applications include the recent Muon optimizer or Riemannian optimization on the Stiefel manifold. Among existing approaches, the Newton-Schulz iteration has emerged as a particularly effective solution, as it relies solely on matrix multiplications and thus achieves high computational efficiency on GPU hardware. Despite its efficiency, the method has inherent limitations - its coefficients are fixed and thus not optimized for a given matrix. In this paper we address this issue by proposing a Chebyshev-optimized version of Newton-Schulz (CANS). Based on the Chebyshev's alternance theorem, we theoretically derive optimal coefficients for the 3-rd order Newton-Schulz iteration and apply a Remez algorithm to compute optimal higher-degree polynomials. We leverage these polynomials to construct controlled approximate orthogonalization schemes, which is of interest in deep learning applications. Practically, we demonstrate the method's effectiveness in two key applications: orthogonalization in the Muon optimizer, and providing an efficient retraction alternative for Riemannian optimization on the Stiefel manifold. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:41:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Grishina",
"Ekaterina",
""
],
[
"Smirnov",
"Matvey",
""
],
[
"Rakhuba",
"Maxim",
""
]
] |
2506.10936 | Sergey Kurapov | Sergey Kurapov, Maxim Davidovsky | Algorithmic methods of finite discrete structures. Topological graph drawing (part III) | 120 pages, in Ukrainian language, 112 figures, a preprint of monography | null | null | null | math.CO cs.DM math.OA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The manuscript considers mathematical models for creating a topological drawing of a graph based on the methods of G. Ringel's vertex rotation theory. An algorithm is presented for generating a topological drawing of a flat part of a graph based on the selection of a basis for the cycle subspace C(G) using the Monte Carlo method. A steepest descent method for constructing a topological drawing of a flat subgraph is described in the manuscript. The topological drawing of a graph is constructed using a combination of the methods of vector intersection algebra developed by L. I. Rapport. Three stages of constructing a flat subgraph of a non-separable graph are described. The issues of constructing a Hamiltonian cycle based on constructing a flat subgraph are considered. A new method for constructing a Hamiltonian cycle of a graph based on the cycle graph of a flat subgraph is described. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:44:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kurapov",
"Sergey",
""
],
[
"Davidovsky",
"Maxim",
""
]
] |
2506.10937 | Xian Chen | Ka Hung Chan and Xinyue Huang and Nobumichi Tamura and Xian Chen | Physics-informed Machine Learning Analysis for Nanoscale Grain Mapping by Synchrotron Laue Microdiffraction | 8 pages, 5 figures | null | null | null | cond-mat.mtrl-sci hep-lat | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Understanding the grain morphology, orientation distribution, and crystal structure of nanocrystals is essential for optimizing the mechanical and physical properties of functional materials. Synchrotron X-ray Laue microdiffraction is a powerful technique for characterizing crystal structures and orientation mapping using focused X-rays. However, when grain sizes are smaller than the beam size, mixed peaks in the Laue pattern from neighboring grains limit the resolution of grain morphology mapping. We propose a physics-informed machine learning (PIML) approach that combines a CNN feature extractor with a physics-informed filtering algorithm to overcome the spatial resolution limits of X-rays, achieving nanoscale resolution for grain mapping. Our PIML method successfully resolves the grain size, orientation distribution, and morphology of Au nanocrystals through synchrotron microdiffraction scans, showing good agreement with electron backscatter diffraction results. This PIML-assisted synchrotron microdiffraction analysis can be generalized to other diffraction-based probes, enabling the characterization of nanosized structures with micron-sized probes. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:44:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chan",
"Ka Hung",
""
],
[
"Huang",
"Xinyue",
""
],
[
"Tamura",
"Nobumichi",
""
],
[
"Chen",
"Xian",
""
]
] |
2506.10938 | Daniele Guerci | Daniele Guerci, Ahmed Abouelkomsan and Liang Fu | From Fractionalization to Chiral Topological Superconductivity in Flat Chern Band | 7+9 pages, 5+6 figures | null | null | null | cond-mat.supr-con cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We show that interacting electrons in a flat Chern band can form, in addition to fractional Chern insulators, a chiral $f$-wave topological superconductor that hosts neutral Majorana fermion edge modes. Superconductivity emerges from an interaction-induced metallic state that exhibits anomalous Hall effect, as observed in rhombohedral graphene and near the $\nu=\frac{2}{3}$ fractional Chern insulator in twisted transition metal dichalcogenides. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:45:46 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Guerci",
"Daniele",
""
],
[
"Abouelkomsan",
"Ahmed",
""
],
[
"Fu",
"Liang",
""
]
] |
2506.10939 | Fr\'ed\'eric Mynard | Bryan Castro Herrej\'on and Fr\'ed\'eric Mynard | On connected subsets of a convergence space | 13 pages | null | null | null | math.GN | http://creativecommons.org/licenses/by/4.0/ | Though a convergence space is connected if and only if its topological modification is connected, connected subsets differ for the convergence and for its topological modification. We explore for what subsets connectedness for the convergence or for the topological modification turn out to be equivalent. In particular, we show that the largest set containing a given connected set for which all subsets in between are connected is the adherence of the connected set for the reciprocal modification of the convergence. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:46:01 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Herrejón",
"Bryan Castro",
""
],
[
"Mynard",
"Frédéric",
""
]
] |
2506.10940 | Tim Langen | Felix Kogel, Yuly Chamorro, Mangesh Bhattarai, Marian Rockenh\"auser, Tatsam Garg, David DeMille, Anastasia Borschevsky, Tim Langen | High-resolution spectroscopy of barium monofluoride: Odd isotopologues, hyperfine structure and isotope shifts | null | null | null | null | physics.atom-ph physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Barium monofluoride (BaF) is a promising molecular species for precision tests of fundamental symmetries and interactions. We present a combined theoretical and experimental study of BaF spectra and isotope shifts, focusing in particular on the poorly understood odd isotopologues 137BaF and 135BaF. By comparing state-of-the-art ab initio calculations with high-resolution fluorescence and absorption spectroscopy data, we provide a benchmark for electronic structure theory and disentangle the hyperfine and rovibrational spectra of the five most abundant isotopologues, from 138BaF to 134BaF. The comprehensive knowledge gained enables a King plot analysis of the isotope shifts that reveals the odd-even staggering of the barium nuclear charge radii. It also paths the way for improved laser cooling of rare BaF isotopologues and crucially supports future measurements of nuclear anapole and Schiff moments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:46:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kogel",
"Felix",
""
],
[
"Chamorro",
"Yuly",
""
],
[
"Bhattarai",
"Mangesh",
""
],
[
"Rockenhäuser",
"Marian",
""
],
[
"Garg",
"Tatsam",
""
],
[
"DeMille",
"David",
""
],
[
"Borschevsky",
"Anastasia",
""
],
[
"Langen",
"Tim",
""
]
] |
2506.10941 | Leigang Qu | Leigang Qu, Feng Cheng, Ziyan Yang, Qi Zhao, Shanchuan Lin, Yichun Shi, Yicong Li, Wenjie Wang, Tat-Seng Chua, Lu Jiang | VINCIE: Unlocking In-context Image Editing from Video | Project page: https://vincie2025.github.io/ | null | null | null | cs.CV cs.AI cs.CL cs.LG cs.MM | http://creativecommons.org/licenses/by/4.0/ | In-context image editing aims to modify images based on a contextual sequence comprising text and previously generated images. Existing methods typically depend on task-specific pipelines and expert models (e.g., segmentation and inpainting) to curate training data. In this work, we explore whether an in-context image editing model can be learned directly from videos. We introduce a scalable approach to annotate videos as interleaved multimodal sequences. To effectively learn from this data, we design a block-causal diffusion transformer trained on three proxy tasks: next-image prediction, current segmentation prediction, and next-segmentation prediction. Additionally, we propose a novel multi-turn image editing benchmark to advance research in this area. Extensive experiments demonstrate that our model exhibits strong in-context image editing capabilities and achieves state-of-the-art results on two multi-turn image editing benchmarks. Despite being trained exclusively on videos, our model also shows promising abilities in multi-concept composition, story generation, and chain-of-editing applications. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:46:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Qu",
"Leigang",
""
],
[
"Cheng",
"Feng",
""
],
[
"Yang",
"Ziyan",
""
],
[
"Zhao",
"Qi",
""
],
[
"Lin",
"Shanchuan",
""
],
[
"Shi",
"Yichun",
""
],
[
"Li",
"Yicong",
""
],
[
"Wang",
"Wenjie",
""
],
[
"Chua",
"Tat-Seng",
""
],
[
"Jiang",
"Lu",
""
]
] |
2506.10942 | Zeynep Pehlivan | Zeynep Pehlivan and Saewon Park and Alexei Sisulu Abrahams and Mika Desblancs-Patel and Benjamin David Steel and Aengus Bridgman | Building a Media Ecosystem Observatory from Scratch: Infrastructure, Methodology, and Insights | null | null | null | null | cs.DL | http://creativecommons.org/licenses/by/4.0/ | Understanding the flow of information across today's fragmented digital media landscape requires scalable, cross-platform infrastructure. In this paper, we present the Canadian Media Ecosystem Observatory, a national-scale infrastructure designed to monitor political and media discourse across platforms in near real time.
Media Ecosystem Observatory (MEO) data infrastructure features custom crawlers for major platforms, a unified indexing pipeline, and a normalization layer that harmonizes heterogeneous schemas into a common data model. Semantic embeddings are computed for each post to enable similarity search and vector-based analyses such as topic modeling and clustering. Processed and raw data are made accessible through API, dashboards and website, supporting both automated and ad hoc research workflows. We illustrate the utility of the observatory through example analyses of major Canadian political events, including Meta's 2023 news ban and the recent federal elections. As a whole, the system offers a model for digital trace infrastructure and an evolving research platform for studying the dynamics of modern media ecosystems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:46:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Pehlivan",
"Zeynep",
""
],
[
"Park",
"Saewon",
""
],
[
"Abrahams",
"Alexei Sisulu",
""
],
[
"Desblancs-Patel",
"Mika",
""
],
[
"Steel",
"Benjamin David",
""
],
[
"Bridgman",
"Aengus",
""
]
] |
2506.10943 | Jyothish Pari | Adam Zweiger, Jyothish Pari, Han Guo, Ekin Aky\"urek, Yoon Kim, Pulkit Agrawal | Self-Adapting Language Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are powerful but static; they lack mechanisms to adapt their weights in response to new tasks, knowledge, or examples. We introduce Self-Adapting LLMs (SEAL), a framework that enables LLMs to self-adapt by generating their own finetuning data and update directives. Given a new input, the model produces a self-edit-a generation that may restructure the information in different ways, specify optimization hyperparameters, or invoke tools for data augmentation and gradient-based updates. Through supervised finetuning (SFT), these self-edits result in persistent weight updates, enabling lasting adaptation. To train the model to produce effective self-edits, we use a reinforcement learning loop with the downstream performance of the updated model as the reward signal. Unlike prior approaches that rely on separate adaptation modules or auxiliary networks, SEAL directly uses the model's own generation to control its adaptation process. Experiments on knowledge incorporation and few-shot generalization show that SEAL is a promising step toward language models capable of self-directed adaptation. Our website and code is available at https://jyopari.github.io/posts/seal. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:48:13 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zweiger",
"Adam",
""
],
[
"Pari",
"Jyothish",
""
],
[
"Guo",
"Han",
""
],
[
"Akyürek",
"Ekin",
""
],
[
"Kim",
"Yoon",
""
],
[
"Agrawal",
"Pulkit",
""
]
] |
2506.10944 | Matteo Carli | Jingxuan Ding and Laura Zichi and Matteo Carli and Menghang Wang and Albert Musaelian and Yu Xie and Boris Kozinsky | Coupled reaction and diffusion governing interface evolution in solid-state batteries | null | null | null | null | cond-mat.mtrl-sci cs.LG physics.chem-ph physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding and controlling the atomistic-level reactions governing the formation of the solid-electrolyte interphase (SEI) is crucial for the viability of next-generation solid state batteries. However, challenges persist due to difficulties in experimentally characterizing buried interfaces and limits in simulation speed and accuracy. We conduct large-scale explicit reactive simulations with quantum accuracy for a symmetric battery cell, {\symcell}, enabled by active learning and deep equivariant neural network interatomic potentials. To automatically characterize the coupled reactions and interdiffusion at the interface, we formulate and use unsupervised classification techniques based on clustering in the space of local atomic environments. Our analysis reveals the formation of a previously unreported crystalline disordered phase, Li$_2$S$_{0.72}$P$_{0.14}$Cl$_{0.14}$, in the SEI, that evaded previous predictions based purely on thermodynamics, underscoring the importance of explicit modeling of full reaction and transport kinetics. Our simulations agree with and explain experimental observations of the SEI formations and elucidate the Li creep mechanisms, critical to dendrite initiation, characterized by significant Li motion along the interface. Our approach is to crease a digital twin from first principles, without adjustable parameters fitted to experiment. As such, it offers capabilities to gain insights into atomistic dynamics governing complex heterogeneous processes in solid-state synthesis and electrochemistry. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:49:05 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ding",
"Jingxuan",
""
],
[
"Zichi",
"Laura",
""
],
[
"Carli",
"Matteo",
""
],
[
"Wang",
"Menghang",
""
],
[
"Musaelian",
"Albert",
""
],
[
"Xie",
"Yu",
""
],
[
"Kozinsky",
"Boris",
""
]
] |
2506.10945 | Jacky Jiang | Jacky Jiang, Natalie Klco, Olivia Di Matteo | Non-Abelian dynamics on a cube: improving quantum compilation through qudit-based simulations | 22 pages, 22 figures, code available | null | null | null | quant-ph hep-lat nucl-th | http://creativecommons.org/licenses/by/4.0/ | Recent developments in mapping lattice gauge theories relevant to the Standard Model onto digital quantum computers identify scalable paths with well-defined quantum compilation challenges toward the continuum. As an entry point to these challenges, we address the simulation of SU(2) lattice gauge theory. Using qudit registers to encode the digitized gauge field, we provide quantum resource estimates, in terms of elementary qudit gates, for arbitrarily high local gauge field truncations. We then demonstrate an end-to-end simulation of real-time, qutrit-digitized SU(2) dynamics on a cube. Through optimizing the simulation, we improved circuit decompositions for uniformly-controlled qudit rotations, an algorithmic primitive for general applications of quantum computing. The decompositions also apply to mixed-dimensional qudit systems, which we found advantageous for compiling lattice gauge theory simulations. Furthermore, we parallelize the evolution of opposite faces in anticipation of similar opportunities arising in three-dimensional lattice volumes. This work details an ambitious executable for future qudit hardware and attests to the value of codesign strategies between lattice gauge theory simulation and quantum compilation. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:49:08 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jiang",
"Jacky",
""
],
[
"Klco",
"Natalie",
""
],
[
"Di Matteo",
"Olivia",
""
]
] |
2506.10946 | Duo Zhou | Evelyn Ma, Duo Zhou, Peizhi Niu, Huiting Zhou, Huan Zhang, Olgica Milenkovic, S. Rasoul Etesami | GUARD: Guided Unlearning and Retention via Data Attribution for Large Language Models | null | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Unlearning in large language models (LLMs) is becoming increasingly important due to regulatory compliance, copyright protection, and privacy concerns. However, a key challenge in LLM unlearning is unintended forgetting, where the removal of specific data inadvertently impairs the utility of the model and its retention of valuable, desired information. While prior work has primarily focused on architectural innovations, the influence of data-level factors on unlearning performance remains underexplored. As a result, existing methods often suffer from degraded retention when forgetting high-impact data. To address this, we propose GUARD-a novel framework for Guided Unlearning And Retention via Data attribution. At its core, GUARD introduces a lightweight proxy data attribution metric tailored for LLM unlearning, which quantifies the "alignment" between the forget and retain sets while remaining computationally efficient. Building on this, we design a novel unlearning objective that assigns adaptive, nonuniform unlearning weights to samples, inversely proportional to their proxy attribution scores. Through such a reallocation of unlearning power, GUARD mitigates unintended losses in retention. We provide rigorous theoretical guarantees that GUARD significantly enhances retention while maintaining forgetting metrics comparable to prior methods. Extensive experiments on the TOFU benchmark across multiple LLM architectures demonstrate that GUARD substantially improves utility preservation while ensuring effective unlearning. Notably, GUARD reduces utility sacrifice on the Retain Set by up to 194.92% in terms of Truth Ratio when forgetting 10% of the training data. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:49:09 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ma",
"Evelyn",
""
],
[
"Zhou",
"Duo",
""
],
[
"Niu",
"Peizhi",
""
],
[
"Zhou",
"Huiting",
""
],
[
"Zhang",
"Huan",
""
],
[
"Milenkovic",
"Olgica",
""
],
[
"Etesami",
"S. Rasoul",
""
]
] |
2506.10947 | Rulin Shao | Rulin Shao, Shuyue Stella Li, Rui Xin, Scott Geng, Yiping Wang, Sewoong Oh, Simon Shaolei Du, Nathan Lambert, Sewon Min, Ranjay Krishna, Yulia Tsvetkov, Hannaneh Hajishirzi, Pang Wei Koh, Luke Zettlemoyer | Spurious Rewards: Rethinking Training Signals in RLVR | null | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We show that reinforcement learning with verifiable rewards (RLVR) can elicit strong mathematical reasoning in certain models even with spurious rewards that have little, no, or even negative correlation with the correct answer. For example, RLVR improves MATH-500 performance for Qwen2.5-Math-7B in absolute points by 21.4% (random reward), 13.8% (format reward), 24.1% (incorrect label), 26.0% (1-shot RL), and 27.1% (majority voting) -- nearly matching the 29.1% gained with ground truth rewards. However, the spurious rewards that work for Qwen often fail to yield gains with other model families like Llama3 or OLMo2. In particular, we find code reasoning -- thinking in code without actual code execution -- to be a distinctive Qwen2.5-Math behavior that becomes significantly more frequent after RLVR, from 65% to over 90%, even with spurious rewards. Overall, we hypothesize that, given the lack of useful reward signal, RLVR must somehow be surfacing useful reasoning representations learned during pretraining, although the exact mechanism remains a topic for future work. We suggest that future RLVR research should possibly be validated on diverse models rather than a single de facto choice, as we show that it is easy to get significant performance gains on Qwen models even with completely spurious reward signals. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:49:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Shao",
"Rulin",
""
],
[
"Li",
"Shuyue Stella",
""
],
[
"Xin",
"Rui",
""
],
[
"Geng",
"Scott",
""
],
[
"Wang",
"Yiping",
""
],
[
"Oh",
"Sewoong",
""
],
[
"Du",
"Simon Shaolei",
""
],
[
"Lambert",
"Nathan",
""
],
[
"Min",
"Sewon",
""
],
[
"Krishna",
"Ranjay",
""
],
[
"Tsvetkov",
"Yulia",
""
],
[
"Hajishirzi",
"Hannaneh",
""
],
[
"Koh",
"Pang Wei",
""
],
[
"Zettlemoyer",
"Luke",
""
]
] |
2506.10948 | Lior Wolf | Boaz Lavon and Shahar Katz and Lior Wolf | Execution Guided Line-by-Line Code Generation | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel approach to neural code generation that incorporates real-time execution signals into the language model generation process. While large language models (LLMs) have demonstrated impressive code generation capabilities, they typically do not utilize execution feedback during inference, a critical signal that human programmers regularly leverage. Our method, Execution-Guided Classifier-Free Guidance (EG-CFG), dynamically incorporates execution signals as the model generates code, providing line-by-line feedback that guides the generation process toward executable solutions. EG-CFG employs a multi-stage process: first, we conduct beam search to sample candidate program completions for each line; second, we extract execution signals by executing these candidates against test cases; and finally, we incorporate these signals into the prompt during generation. By maintaining consistent signals across tokens within the same line and refreshing signals at line boundaries, our approach provides coherent guidance while preserving syntactic structure. Moreover, the method naturally supports native parallelism at the task level in which multiple agents operate in parallel, exploring diverse reasoning paths and collectively generating a broad set of candidate solutions. Our experiments across diverse coding tasks demonstrate that EG-CFG significantly improves code generation performance compared to standard approaches, achieving state-of-the-art results across various levels of complexity, from foundational problems to challenging competitive programming tasks. Our code is available at: https://github.com/boazlavon/eg_cfg | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:50:05 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lavon",
"Boaz",
""
],
[
"Katz",
"Shahar",
""
],
[
"Wolf",
"Lior",
""
]
] |
2506.10949 | Yueh-Han Chen | Chen Yueh-Han, Nitish Joshi, Yulin Chen, Maksym Andriushchenko, Rico Angell, He He | Monitoring Decomposition Attacks in LLMs with Lightweight Sequential Monitors | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Current LLM safety defenses fail under decomposition attacks, where a malicious goal is decomposed into benign subtasks that circumvent refusals. The challenge lies in the existing shallow safety alignment techniques: they only detect harm in the immediate prompt and do not reason about long-range intent, leaving them blind to malicious intent that emerges over a sequence of seemingly benign instructions. We therefore propose adding an external monitor that observes the conversation at a higher granularity. To facilitate our study of monitoring decomposition attacks, we curate the largest and most diverse dataset to date, including question-answering, text-to-image, and agentic tasks. We verify our datasets by testing them on frontier LLMs and show an 87% attack success rate on average on GPT-4o. This confirms that decomposition attack is broadly effective. Additionally, we find that random tasks can be injected into the decomposed subtasks to further obfuscate malicious intents. To defend in real time, we propose a lightweight sequential monitoring framework that cumulatively evaluates each subtask. We show that a carefully prompt engineered lightweight monitor achieves a 93% defense success rate, beating reasoning models like o3 mini as a monitor. Moreover, it remains robust against random task injection and cuts cost by 90% and latency by 50%. Our findings suggest that lightweight sequential monitors are highly effective in mitigating decomposition attacks and are viable in deployment. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:50:58 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yueh-Han",
"Chen",
""
],
[
"Joshi",
"Nitish",
""
],
[
"Chen",
"Yulin",
""
],
[
"Andriushchenko",
"Maksym",
""
],
[
"Angell",
"Rico",
""
],
[
"He",
"He",
""
]
] |
2506.10950 | Siddharth Dwivedi | Siddharth Dwivedi | A closer look at the algebraic-operator correspondence between position and momentum space in Quantum Mechanics | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | In this work we take a closer look at the algebraic-operator correspondence between the momentum ($k$) space and the position ($x$) space which gives us the form of the momentum operator in position space and vice versa in Quantum Mechanics. Taking the definition of the Fourier Transform as the starting point, we present a two parameter generalisation of the form of the momentum operator in position space followed by a generalisation of the momentum eigenfunctions and a few other results. It is also seen that the results obtained here have a well defined limit to the standard results of Quantum Mechanics. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:52:15 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Dwivedi",
"Siddharth",
""
]
] |
2506.10951 | Fr\'ed\'eric Mynard | Fr\'ed\'eric Mynard | On (pre-) approach spaces within convergence approach spaces | 21 pages | null | null | null | math.GN | http://creativecommons.org/licenses/by/4.0/ | The purpose of this note is to illustrate a parallel between (pre)topologies when seen among convergence spaces and (pre)approach spaces when seen among convergence approach spaces, that appears to be a more complete parallel than in the traditional presentation of these approach structures. This sheds some light on the reflector from the category of convergence approach spaces to that of approach spaces and even on the structure of approach spaces as such. This point of view allows for a characterization of approach spaces among convergence approach spaces represented as pointfree convergence frames as in [18]. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:53:20 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Mynard",
"Frédéric",
""
]
] |
2506.10952 | Howe Tissue | Mozhi Zhang, Howe Tissue, Lu Wang, Xipeng Qiu | Domain2Vec: Vectorizing Datasets to Find the Optimal Data Mixture without Training | Accepted to ICML2025 | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce~\textsc{Domain2Vec}, a novel approach that decomposes any dataset into a linear combination of several \emph{meta-domains}, a new concept designed to capture the key underlying features of datasets. \textsc{Domain2Vec} maintains a vocabulary of meta-domains and uses a classifier to decompose any given dataset into a domain vector that corresponds to a distribution over this vocabulary. These domain vectors enable the identification of the optimal data mixture for language model (LM) pretraining in a training-free manner under the \emph{\textbf{D}istribution \textbf{A}lignment \textbf{A}ssumption} (DA$^{2}$), which suggests that when the data distributions of the training set and the validation set are better aligned, a lower validation loss is achieved. Moreover, \textsc{Domain2vec} can be seamlessly integrated into previous works to model the relationship between domain vectors and LM performance, greatly enhancing the efficiency and scalability of previous methods. Extensive experiments demonstrate that \textsc{Domain2Vec} helps find the data mixture that enhances downstream task performance with minimal computational overhead. Specifically, \textsc{Domain2Vec} achieves the same validation loss on Pile-CC using only $51.5\%$ of the computation required when training on the original mixture of The Pile dataset. Under equivalent compute budget, \textsc{Domain2Vec} improves downstream performance by an average of $2.83\%$. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:53:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Mozhi",
""
],
[
"Tissue",
"Howe",
""
],
[
"Wang",
"Lu",
""
],
[
"Qiu",
"Xipeng",
""
]
] |
2506.10953 | Xing Han L\`u | Xing Han L\`u, Gaurav Kamath, Marius Mosbach, Siva Reddy | Build the web for agents, not agents for the web | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in Large Language Models (LLMs) and multimodal counterparts have spurred significant interest in developing web agents -- AI systems capable of autonomously navigating and completing tasks within web environments. While holding tremendous promise for automating complex web interactions, current approaches face substantial challenges due to the fundamental mismatch between human-designed interfaces and LLM capabilities. Current methods struggle with the inherent complexity of web inputs, whether processing massive DOM trees, relying on screenshots augmented with additional information, or bypassing the user interface entirely through API interactions. This position paper advocates for a paradigm shift in web agent research: rather than forcing web agents to adapt to interfaces designed for humans, we should develop a new interaction paradigm specifically optimized for agentic capabilities. To this end, we introduce the concept of an Agentic Web Interface (AWI), an interface specifically designed for agents to navigate a website. We establish six guiding principles for AWI design, emphasizing safety, efficiency, and standardization, to account for the interests of all primary stakeholders. This reframing aims to overcome fundamental limitations of existing interfaces, paving the way for more efficient, reliable, and transparent web agent design, which will be a collaborative effort involving the broader ML community. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:53:58 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lù",
"Xing Han",
""
],
[
"Kamath",
"Gaurav",
""
],
[
"Mosbach",
"Marius",
""
],
[
"Reddy",
"Siva",
""
]
] |
2506.10954 | Lianghong Guo Xys | Lianghong Guo, Yanlin Wang, Caihua Li, Pengyu Yang, Jiachi Chen, Wei Tao, Yingtian Zou, Duyu Tang, and Zibin Zheng | SWE-Factory: Your Automated Factory for Issue Resolution Training Data and Evaluation Benchmarks | null | null | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Constructing large-scale datasets for the GitHub issue resolution task is crucial for both training and evaluating the software engineering capabilities of Large Language Models (LLMs). However, the traditional process for creating such benchmarks is notoriously challenging and labor-intensive, particularly in the stages of setting up evaluation environments, grading test outcomes, and validating task instances. In this paper, we propose SWE-Factory, an automated pipeline designed to address these challenges. To tackle these issues, our pipeline integrates three core automated components. First, we introduce SWE-Builder, a multi-agent system that automates evaluation environment construction, which employs four specialized agents that work in a collaborative, iterative loop and leverages an environment memory pool to enhance efficiency. Second, we introduce a standardized, exit-code-based grading method that eliminates the need for manually writing custom parsers. Finally, we automate the fail2pass validation process using these reliable exit code signals. Experiments on 671 issues across four programming languages show that our pipeline can effectively construct valid task instances; for example, with GPT-4.1-mini, our SWE-Builder constructs 269 valid instances at $0.045 per instance, while with Gemini-2.5-flash, it achieves comparable performance at the lowest cost of $0.024 per instance. We also demonstrate that our exit-code-based grading achieves 100% accuracy compared to manual inspection, and our automated fail2pass validation reaches a precision of 0.92 and a recall of 1.00. We hope our automated pipeline will accelerate the collection of large-scale, high-quality GitHub issue resolution datasets for both training and evaluation. Our code and datasets are released at https://github.com/DeepSoftwareAnalytics/swe-factory. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:54:17 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Guo",
"Lianghong",
""
],
[
"Wang",
"Yanlin",
""
],
[
"Li",
"Caihua",
""
],
[
"Yang",
"Pengyu",
""
],
[
"Chen",
"Jiachi",
""
],
[
"Tao",
"Wei",
""
],
[
"Zou",
"Yingtian",
""
],
[
"Tang",
"Duyu",
""
],
[
"Zheng",
"Zibin",
""
]
] |
2506.10955 | Aayush Karan | Aayush Karan, Kulin Shah, Sitan Chen | ReGuidance: A Simple Diffusion Wrapper for Boosting Sample Quality on Hard Inverse Problems | 38 pages, 14 figures | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | There has been a flurry of activity around using pretrained diffusion models as informed data priors for solving inverse problems, and more generally around steering these models using reward models. Training-free methods like diffusion posterior sampling (DPS) and its many variants have offered flexible heuristic algorithms for these tasks, but when the reward is not informative enough, e.g., in hard inverse problems with low signal-to-noise ratio, these techniques veer off the data manifold, failing to produce realistic outputs. In this work, we devise a simple wrapper, ReGuidance, for boosting both the sample realism and reward achieved by these methods. Given a candidate solution $\hat{x}$ produced by an algorithm of the user's choice, we propose inverting the solution by running the unconditional probability flow ODE in reverse starting from $\hat{x}$, and then using the resulting latent as an initialization for DPS. We evaluate our wrapper on hard inverse problems like large box in-painting and super-resolution with high upscaling. Whereas state-of-the-art baselines visibly fail, we find that applying our wrapper on top of these baselines significantly boosts sample quality and measurement consistency. We complement these findings with theory proving that on certain multimodal data distributions, ReGuidance simultaneously boosts the reward and brings the candidate solution closer to the data manifold. To our knowledge, this constitutes the first rigorous algorithmic guarantee for DPS. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:55:17 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Karan",
"Aayush",
""
],
[
"Shah",
"Kulin",
""
],
[
"Chen",
"Sitan",
""
]
] |
2506.10956 | Volker Deringer | John L. A. Gardner and Daniel F. Thomas du Toit and Chiheb Ben Mahmoud and Zo\'e Faure Beaulieu and Veronika Juraskova and Laura-Bianca Pa\c{s}ca and Louise A. M. Rosset and Fernanda Duarte and Fausto Martelli and Chris J. Pickard and Volker L. Deringer | Distillation of atomistic foundation models across architectures and chemical domains | null | null | null | null | physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine-learned interatomic potentials have transformed computational research in the physical sciences. Recent atomistic `foundation' models have changed the field yet again: trained on many different chemical elements and domains, these potentials are widely applicable, but comparably slow and resource-intensive to run. Here we show how distillation via synthetic data can be used to cheaply transfer knowledge from atomistic foundation models to a range of different architectures, unlocking much smaller, more efficient potentials. We demonstrate speed-ups of $> 10\times$ by distilling from one graph-network architecture into another, and $> 100\times$ by leveraging the atomic cluster expansion framework. We showcase applicability across chemical and materials domains: from liquid water to hydrogen under extreme conditions; from porous silica and a hybrid halide perovskite solar-cell material to modelling organic reactions. Our work shows how distillation can support the routine and computationally efficient use of current and future atomistic foundation models in real-world scientific research. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:55:48 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Gardner",
"John L. A.",
""
],
[
"Toit",
"Daniel F. Thomas du",
""
],
[
"Mahmoud",
"Chiheb Ben",
""
],
[
"Beaulieu",
"Zoé Faure",
""
],
[
"Juraskova",
"Veronika",
""
],
[
"Paşca",
"Laura-Bianca",
""
],
[
"Rosset",
"Louise A. M.",
""
],
[
"Duarte",
"Fernanda",
""
],
[
"Martelli",
"Fausto",
""
],
[
"Pickard",
"Chris J.",
""
],
[
"Deringer",
"Volker L.",
""
]
] |
2506.10957 | Guo Chuan Thiang | Matthias Ludewig and Guo Chuan Thiang | Large-scale quantization of trace I: Finite propagation operators | 60 pages, 3 figures | null | null | null | math.KT cond-mat.mes-hall math-ph math.MP math.OA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inspired by parallel developments in coarse geometry in mathematics and exact macroscopic quantization in physics, we present a family of general trace formulae which are universally quantized and depend only on large-scale geometric features of the input data. They generalize, to arbitrary dimensions, formulae found by Roe in his partitioned manifold index theorem, as well as the Kubo and Kitaev formulae for 2D Hall conductance used in physics. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:56:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ludewig",
"Matthias",
""
],
[
"Thiang",
"Guo Chuan",
""
]
] |
2506.10958 | Randy Palamar | Randy Palamar, Mohammad Rahim Sobhani, Darren Dahunsi, Negar Majidi, Afshin Kashani Ilkhechi, Joy Wang, Jeremy Brown, and Roger Zemp | Bias-Switchable Row-Column Array Imaging using Fast Orthogonal Row-Column Electronic Scanning (FORCES) Compared with Conventional Row-Column Array Imaging | null | null | null | null | eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Row-Column Arrays (RCAs) offer an attractive alternative to fully wired 2D-arrays for 3D-ultrasound, due to their greatly simplified wiring. However, conventional RCAs face challenges related to their long elements. These include an inability to image beyond the shadow of the aperture and an inability to focus in both transmit and receive for desired scan planes. To address these limitations, we recently developed bias-switchable RCAs, also known as Top Orthogonal to Bottom Electrode (TOBE) arrays. These arrays provide novel opportunities to read out from every element of the array and achieve high-quality images. While TOBE arrays and their associated imaging schemes have shown promise, they have not yet been directly compared experimentally to conventional RCA imaging techniques. This study aims to provide such a comparison, demonstrating superior B-scan and volumetric images from two electrostrictive relaxor TOBE arrays, using a method called Fast Orthogonal Row-Column Electronic scanning (FORCES), compared to conventional RCA imaging schemes, including Tilted Plane Wave (TPW) compounding and Virtual Line Source (VLS) imaging. The study quantifies resolution and Generalized Contrast to Noise Ratio (gCNR) in phantoms, and also demonstrates volumetric acquisitions in phantom and animal models. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:56:24 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Palamar",
"Randy",
""
],
[
"Sobhani",
"Mohammad Rahim",
""
],
[
"Dahunsi",
"Darren",
""
],
[
"Majidi",
"Negar",
""
],
[
"Ilkhechi",
"Afshin Kashani",
""
],
[
"Wang",
"Joy",
""
],
[
"Brown",
"Jeremy",
""
],
[
"Zemp",
"Roger",
""
]
] |
2506.10959 | Zhaiming Shen | Zhaiming Shen, Alexander Hsu, Rongjie Lai, Wenjing Liao | Understanding In-Context Learning on Structured Manifolds: Bridging Attention to Kernel Methods | null | null | null | null | cs.LG cs.AI math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While in-context learning (ICL) has achieved remarkable success in natural language and vision domains, its theoretical understanding--particularly in the context of structured geometric data--remains unexplored. In this work, we initiate a theoretical study of ICL for regression of H\"older functions on manifolds. By establishing a novel connection between the attention mechanism and classical kernel methods, we derive generalization error bounds in terms of the prompt length and the number of training tasks. When a sufficient number of training tasks are observed, transformers give rise to the minimax regression rate of H\"older functions on manifolds, which scales exponentially with the intrinsic dimension of the manifold, rather than the ambient space dimension. Our result also characterizes how the generalization error scales with the number of training tasks, shedding light on the complexity of transformers as in-context algorithm learners. Our findings provide foundational insights into the role of geometry in ICL and novels tools to study ICL of nonlinear models. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:56:26 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Shen",
"Zhaiming",
""
],
[
"Hsu",
"Alexander",
""
],
[
"Lai",
"Rongjie",
""
],
[
"Liao",
"Wenjing",
""
]
] |
2506.10960 | Ningyu Zhang | Kangwei Liu, Siyuan Cheng, Bozhong Tian, Xiaozhuan Liang, Yuyang Yin, Meng Han, Ningyu Zhang, Bryan Hooi, Xi Chen, Shumin Deng | ChineseHarm-Bench: A Chinese Harmful Content Detection Benchmark | Work in progress | null | null | null | cs.CL cs.AI cs.CR cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) have been increasingly applied to automated harmful content detection tasks, assisting moderators in identifying policy violations and improving the overall efficiency and accuracy of content review. However, existing resources for harmful content detection are predominantly focused on English, with Chinese datasets remaining scarce and often limited in scope. We present a comprehensive, professionally annotated benchmark for Chinese content harm detection, which covers six representative categories and is constructed entirely from real-world data. Our annotation process further yields a knowledge rule base that provides explicit expert knowledge to assist LLMs in Chinese harmful content detection. In addition, we propose a knowledge-augmented baseline that integrates both human-annotated knowledge rules and implicit knowledge from large language models, enabling smaller models to achieve performance comparable to state-of-the-art LLMs. Code and data are available at https://github.com/zjunlp/ChineseHarm-bench. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:57:05 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liu",
"Kangwei",
""
],
[
"Cheng",
"Siyuan",
""
],
[
"Tian",
"Bozhong",
""
],
[
"Liang",
"Xiaozhuan",
""
],
[
"Yin",
"Yuyang",
""
],
[
"Han",
"Meng",
""
],
[
"Zhang",
"Ningyu",
""
],
[
"Hooi",
"Bryan",
""
],
[
"Chen",
"Xi",
""
],
[
"Deng",
"Shumin",
""
]
] |
2506.10961 | Alice Curtin | Alice P. Curtin, Shion Andrew, Sunil Simha, Alice Cai, Kenzie Nimmo, Shami Chatterjee, Amanda M. Cook, Fengqiu Adam Dong, Yuxin Dong, Tarraneh Eftekhari, Wen-fai Fong, Emmanuel Fonseca, Jason W. T. Hessels, Ronniy C. Joseph, Victoria Kaspi, Calvin Leung, Robert Main, Kiyoshi W. Masui, Ryan Mckinven, Daniele Michilli, Mason Ng, Ayush Pandhi, Aaron B. Pearlman, Ziggy Pleunis, Mawson W. Sammons, Ketan R. Sand, Paul Scholz, Vishwangi Shah, Kaitlyn Shin, and Aaron Tohuvavohu | Discovery and Localization of the Swift-Observed FRB 20241228A in a Star-forming Host Galaxy | Submitted to ApJ | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | On 2024 December 28, CHIME/FRB detected the thus-far non-repeating FRB 20241228A with a real-time signal-to-noise ratio of $>50$. Approximately 112~s later, the X-ray Telescope onboard the Neil Gehrels Swift Observatory was on source, the fastest follow-up to-date of a non-repeating FRB (Tohuvavohu et al. in prep.). Using CHIME/FRB and two of the three CHIME/FRB Outriggers, we obtained a Very Long Baseline Interferometry localization for FRB 20241228A with a 1$\sigma$ confidence ellipse of 11$^{\prime\prime}$ by 0.2$^{\prime\prime}$. This represents the first published localization using both the CHIME-KKO and CHIME-GBO Outriggers. We associate FRB 20241228A with a star-forming galaxy at a redshift of $z = 0.1614\pm0.0002$. The persistent X-ray luminosity limit at this source's location and distance is $<1.2 \times 10^{43}$ erg s$^{-1}$ in the $0.3-10$ keV band, the most stringent limit of any non-repeating FRB to-date (Tohuvavohu et al. in prep.). The stellar mass ($\sim 2.6 \times 10^{10}\,M_{\odot}$) and star formation rate ($\sim 2.9\,M_{\odot}$~yr$^{-1}$) of the host galaxy of FRB 20241228A are consistent with the broader FRB host galaxy population. We measure significant scattering ($\sim$1ms) and scintillation ($\sim$20 kHz at 600 MHz) along the line of sight to this source, and suggest the scintillation screen is Galactic while the scattering screen is extragalactic. FRB 20241228A represents an exciting example of a new era in which we can harness VLBI-localizations and rapid high-energy follow-up to probe FRB progenitors. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:57:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Curtin",
"Alice P.",
""
],
[
"Andrew",
"Shion",
""
],
[
"Simha",
"Sunil",
""
],
[
"Cai",
"Alice",
""
],
[
"Nimmo",
"Kenzie",
""
],
[
"Chatterjee",
"Shami",
""
],
[
"Cook",
"Amanda M.",
""
],
[
"Dong",
"Fengqiu Adam",
""
],
[
"Dong",
"Yuxin",
""
],
[
"Eftekhari",
"Tarraneh",
""
],
[
"Fong",
"Wen-fai",
""
],
[
"Fonseca",
"Emmanuel",
""
],
[
"Hessels",
"Jason W. T.",
""
],
[
"Joseph",
"Ronniy C.",
""
],
[
"Kaspi",
"Victoria",
""
],
[
"Leung",
"Calvin",
""
],
[
"Main",
"Robert",
""
],
[
"Masui",
"Kiyoshi W.",
""
],
[
"Mckinven",
"Ryan",
""
],
[
"Michilli",
"Daniele",
""
],
[
"Ng",
"Mason",
""
],
[
"Pandhi",
"Ayush",
""
],
[
"Pearlman",
"Aaron B.",
""
],
[
"Pleunis",
"Ziggy",
""
],
[
"Sammons",
"Mawson W.",
""
],
[
"Sand",
"Ketan R.",
""
],
[
"Scholz",
"Paul",
""
],
[
"Shah",
"Vishwangi",
""
],
[
"Shin",
"Kaitlyn",
""
],
[
"Tohuvavohu",
"Aaron",
""
]
] |
2506.10962 | Yuanhui Huang | Yuanhui Huang, Weiliang Chen, Wenzhao Zheng, Yueqi Duan, Jie Zhou, Jiwen Lu | SpectralAR: Spectral Autoregressive Visual Generation | Project Page: https://huang-yh.github.io/spectralar/ | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Autoregressive visual generation has garnered increasing attention due to its scalability and compatibility with other modalities compared with diffusion models. Most existing methods construct visual sequences as spatial patches for autoregressive generation. However, image patches are inherently parallel, contradicting the causal nature of autoregressive modeling. To address this, we propose a Spectral AutoRegressive (SpectralAR) visual generation framework, which realizes causality for visual sequences from the spectral perspective. Specifically, we first transform an image into ordered spectral tokens with Nested Spectral Tokenization, representing lower to higher frequency components. We then perform autoregressive generation in a coarse-to-fine manner with the sequences of spectral tokens. By considering different levels of detail in images, our SpectralAR achieves both sequence causality and token efficiency without bells and whistles. We conduct extensive experiments on ImageNet-1K for image reconstruction and autoregressive generation, and SpectralAR achieves 3.02 gFID with only 64 tokens and 310M parameters. Project page: https://huang-yh.github.io/spectralar/. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:57:44 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Yuanhui",
""
],
[
"Chen",
"Weiliang",
""
],
[
"Zheng",
"Wenzhao",
""
],
[
"Duan",
"Yueqi",
""
],
[
"Zhou",
"Jie",
""
],
[
"Lu",
"Jiwen",
""
]
] |
2506.10963 | Yuxuan Luo | Yuxuan Luo, Yuhui Yuan, Junwen Chen, Haonan Cai, Ziyi Yue, Yuwei Yang, Fatima Zohra Daha, Ji Li, Zhouhui Lian | MMMG: A Massive, Multidisciplinary, Multi-Tier Generation Benchmark for Text-to-Image Reasoning | null | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | In this paper, we introduce knowledge image generation as a new task, alongside the Massive Multi-Discipline Multi-Tier Knowledge-Image Generation Benchmark (MMMG) to probe the reasoning capability of image generation models. Knowledge images have been central to human civilization and to the mechanisms of human learning--a fact underscored by dual-coding theory and the picture-superiority effect. Generating such images is challenging, demanding multimodal reasoning that fuses world knowledge with pixel-level grounding into clear explanatory visuals. To enable comprehensive evaluation, MMMG offers 4,456 expert-validated (knowledge) image-prompt pairs spanning 10 disciplines, 6 educational levels, and diverse knowledge formats such as charts, diagrams, and mind maps. To eliminate confounding complexity during evaluation, we adopt a unified Knowledge Graph (KG) representation. Each KG explicitly delineates a target image's core entities and their dependencies. We further introduce MMMG-Score to evaluate generated knowledge images. This metric combines factual fidelity, measured by graph-edit distance between KGs, with visual clarity assessment. Comprehensive evaluations of 16 state-of-the-art text-to-image generation models expose serious reasoning deficits--low entity fidelity, weak relations, and clutter--with GPT-4o achieving an MMMG-Score of only 50.20, underscoring the benchmark's difficulty. To spur further progress, we release FLUX-Reason (MMMG-Score of 34.45), an effective and open baseline that combines a reasoning LLM with diffusion models and is trained on 16,000 curated knowledge image-prompt pairs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:58:09 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Luo",
"Yuxuan",
""
],
[
"Yuan",
"Yuhui",
""
],
[
"Chen",
"Junwen",
""
],
[
"Cai",
"Haonan",
""
],
[
"Yue",
"Ziyi",
""
],
[
"Yang",
"Yuwei",
""
],
[
"Daha",
"Fatima Zohra",
""
],
[
"Li",
"Ji",
""
],
[
"Lian",
"Zhouhui",
""
]
] |
2506.10964 | Rico Henri Herzog | Rico H Herzog, Till Degkwitz, Trivik Verma | The Urban Model Platform: A Public Backbone for Modeling and Simulation in Urban Digital Twins | null | null | null | null | cs.CY cs.HC | http://creativecommons.org/licenses/by/4.0/ | Urban digital twins are increasingly perceived as a way to pool the growing digital resources of cities for the purpose of a more sustainable and integrated urban planning. Models and simulations are central to this undertaking: They enable "what if?" scenarios, create insights and describe relationships between the vast data that is being collected. However, the process of integrating and subsequently using models in urban digital twins is an inherently complex undertaking. It raises questions about how to represent urban complexity, how to deal with uncertain assUrban Model Platformtions and modeling paradigms, and how to capture underlying power relations. Existent approaches in the domain largely focus on monolithic and centralized solutions in the tradition of neoliberal city-making, oftentimes prohibiting pluralistic and open interoperable models. Using a participatory design for participatory systems approach together with the City of Hamburg, Germany, we find that an open Urban Model Platform can function both as a public technological backbone for modeling and simulation in urban digital twins and as a socio-technical framework for a collaborative and pluralistic representation of urban processes. Such a platform builds on open standards, allows for a decentralized integration of models, enables communication between models and supports a multi-model approach to representing urban systems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:58:10 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Herzog",
"Rico H",
""
],
[
"Degkwitz",
"Till",
""
],
[
"Verma",
"Trivik",
""
]
] |
2506.10965 | Seth Musser | Yi Huang, Seth Musser, Jihang Zhu, Yang-Zhi Chou, Sankar Das Sarma | Apparent inconsistency between Streda formula and Hall conductivity in reentrant integer quantum anomalous Hall effect in twisted MoTe$_2$ | 8 pages, 3 figures | null | null | null | cond-mat.str-el cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Recent experiments in twisted bilayer MoTe$_2$ (tMoTe$_2$) have uncovered a rich landscape of correlated phases. In this work, we investigate the reentrant integer quantum anomalous Hall (RIQAH) states reported in F. Xu, et. al., arXiv:2504.06972 which displays a notable mismatch between the Hall conductivity measured via transport and that inferred from the Streda formula. We argue that this discrepancy can be explained if the RIQAH state is a quantum Hall bubble phase, analogous to similar well-established phenomena in 2D GaAs quantum wells. While this explains the RIQAH state at a filling $\nu = -0.63$, the other RIQAH state at $\nu = -0.7$ has a smaller slope necessitating a different interpretation. We propose that this discrepancy arises due to a nearby resistive peak masking the true slope. Furthermore, we identify this resistive peak as a signature of a magnetic phase transition near $\nu = -0.75$, possibly driven by a Van Hove singularity. The anomalous Hall response and Landau fan evolution across this transition suggest a change in Fermi surface topology and a phase with zero Chern number, potentially corresponding to an inter-valley coherent state. These observations offer new insights into the nature of the RIQAH states and raise the possibility that the nearby superconducting phase may have an inter-valley coherent normal state. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Yi",
""
],
[
"Musser",
"Seth",
""
],
[
"Zhu",
"Jihang",
""
],
[
"Chou",
"Yang-Zhi",
""
],
[
"Sarma",
"Sankar Das",
""
]
] |
2506.10966 | Ning Gao | Ning Gao, Yilun Chen, Shuai Yang, Xinyi Chen, Yang Tian, Hao Li, Haifeng Huang, Hanqing Wang, Tai Wang, Jiangmiao Pang | GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Robotic manipulation in real-world settings remains challenging, especially regarding robust generalization. Existing simulation platforms lack sufficient support for exploring how policies adapt to varied instructions and scenarios. Thus, they lag behind the growing interest in instruction-following foundation models like LLMs, whose adaptability is crucial yet remains underexplored in fair comparisons. To bridge this gap, we introduce GenManip, a realistic tabletop simulation platform tailored for policy generalization studies. It features an automatic pipeline via LLM-driven task-oriented scene graph to synthesize large-scale, diverse tasks using 10K annotated 3D object assets. To systematically assess generalization, we present GenManip-Bench, a benchmark of 200 scenarios refined via human-in-the-loop corrections. We evaluate two policy types: (1) modular manipulation systems integrating foundation models for perception, reasoning, and planning, and (2) end-to-end policies trained through scalable data collection. Results show that while data scaling benefits end-to-end methods, modular systems enhanced with foundation models generalize more effectively across diverse scenarios. We anticipate this platform to facilitate critical insights for advancing policy generalization in realistic conditions. Project Page: https://genmanip.axi404.top/. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:04 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Gao",
"Ning",
""
],
[
"Chen",
"Yilun",
""
],
[
"Yang",
"Shuai",
""
],
[
"Chen",
"Xinyi",
""
],
[
"Tian",
"Yang",
""
],
[
"Li",
"Hao",
""
],
[
"Huang",
"Haifeng",
""
],
[
"Wang",
"Hanqing",
""
],
[
"Wang",
"Tai",
""
],
[
"Pang",
"Jiangmiao",
""
]
] |
2506.10967 | Qizhe Zhang | Qizhe Zhang, Mengzhen Liu, Lichen Li, Ming Lu, Yuan Zhang, Junwen Pan, Qi She, Shanghang Zhang | Beyond Attention or Similarity: Maximizing Conditional Diversity for Token Pruning in MLLMs | 22 pages, 5 figures, code: https://github.com/Theia-4869/CDPruner, project page: https://theia-4869.github.io/CDPruner | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In multimodal large language models (MLLMs), the length of input visual tokens is often significantly greater than that of their textual counterparts, leading to a high inference cost. Many works aim to address this issue by removing redundant visual tokens. However, current approaches either rely on attention-based pruning, which retains numerous duplicate tokens, or use similarity-based pruning, overlooking the instruction relevance, consequently causing suboptimal performance. In this paper, we go beyond attention or similarity by proposing a novel visual token pruning method named CDPruner, which maximizes the conditional diversity of retained tokens. We first define the conditional similarity between visual tokens conditioned on the instruction, and then reformulate the token pruning problem with determinantal point process (DPP) to maximize the conditional diversity of the selected subset. The proposed CDPruner is training-free and model-agnostic, allowing easy application to various MLLMs. Extensive experiments across diverse MLLMs show that CDPruner establishes new state-of-the-art on various vision-language benchmarks. By maximizing conditional diversity through DPP, the selected subset better represents the input images while closely adhering to user instructions, thereby preserving strong performance even with high reduction ratios. When applied to LLaVA, CDPruner reduces FLOPs by 95\% and CUDA latency by 78\%, while maintaining 94\% of the original accuracy. Our code is available at https://github.com/Theia-4869/CDPruner. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:09 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Qizhe",
""
],
[
"Liu",
"Mengzhen",
""
],
[
"Li",
"Lichen",
""
],
[
"Lu",
"Ming",
""
],
[
"Zhang",
"Yuan",
""
],
[
"Pan",
"Junwen",
""
],
[
"She",
"Qi",
""
],
[
"Zhang",
"Shanghang",
""
]
] |
2506.10968 | Justin Kerr | Justin Kerr, Kush Hari, Ethan Weber, Chung Min Kim, Brent Yi, Tyler Bonnen, Ken Goldberg, Angjoo Kanazawa | Eye, Robot: Learning to Look to Act with a BC-RL Perception-Action Loop | Project page: https://www.eyerobot.net/ | null | null | null | cs.RO cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | Humans do not passively observe the visual world -- we actively look in order to act. Motivated by this principle, we introduce EyeRobot, a robotic system with gaze behavior that emerges from the need to complete real-world tasks. We develop a mechanical eyeball that can freely rotate to observe its surroundings and train a gaze policy to control it using reinforcement learning. We accomplish this by first collecting teleoperated demonstrations paired with a 360 camera. This data is imported into a simulation environment that supports rendering arbitrary eyeball viewpoints, allowing episode rollouts of eye gaze on top of robot demonstrations. We then introduce a BC-RL loop to train the hand and eye jointly: the hand (BC) agent is trained from rendered eye observations, and the eye (RL) agent is rewarded when the hand produces correct action predictions. In this way, hand-eye coordination emerges as the eye looks towards regions which allow the hand to complete the task. EyeRobot implements a foveal-inspired policy architecture allowing high resolution with a small compute budget, which we find also leads to the emergence of more stable fixation as well as improved ability to track objects and ignore distractors. We evaluate EyeRobot on five panoramic workspace manipulation tasks requiring manipulation in an arc surrounding the robot arm. Our experiments suggest EyeRobot exhibits hand-eye coordination behaviors which effectively facilitate manipulation over large workspaces with a single camera. See project site for videos: https://www.eyerobot.net/ | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:11 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kerr",
"Justin",
""
],
[
"Hari",
"Kush",
""
],
[
"Weber",
"Ethan",
""
],
[
"Kim",
"Chung Min",
""
],
[
"Yi",
"Brent",
""
],
[
"Bonnen",
"Tyler",
""
],
[
"Goldberg",
"Ken",
""
],
[
"Kanazawa",
"Angjoo",
""
]
] |
2506.10969 | Caleb Lammers | Caleb Lammers, Joshua N. Winn | The Gap-Giant Association: Are Planets Hiding in the Gaps? | 11 pages, 4 figures, 1 table. Submitted to ApJ | null | null | null | astro-ph.EP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A handful of stars are known to host both an inner system of multiple transiting planets and an outer giant planet. These systems all feature a prominent gap between the orbits of two of the transiting planets, distinguishing them from typical multiplanet systems with more uniform orbital spacings. The reason for the association between inner gaps and outer giants is unknown. In this paper, we assess whether undiscovered planets might occupy these gaps in systems with outer giants. For each of the four relevant systems - Kepler-48, Kepler-65, Kepler-90, and Kepler-139 - we found that a $\sim 2 - 20 M_\oplus$ planet could reside in the gap without inducing dynamical instability. However, in each case, the gravitational influence of the outer giant planet is insufficient to tilt the orbit of the hypothetical planet by enough to prevent transits, ruling out a proposed theory for the observed gap-giant association. The gaps might instead contain smaller, undetectable planets ($ \lesssim 0.5 - 1\,R_\oplus$), or be entirely devoid of planets. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:18 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lammers",
"Caleb",
""
],
[
"Winn",
"Joshua N.",
""
]
] |
2506.10970 | Xian Chen | Chenbo Zhang and Xian Chen | FerroAI: A Deep Learning Model for Predicting Phase Diagrams of Ferroelectric Materials | 10 pages, 8 figures, 2 tables | null | null | null | cond-mat.mtrl-sci cond-mat.dis-nn | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Composition-temperature phase diagrams are crucial for designing ferroelectric materials, however predicting them accurately remains challenging due to limited phase transformation data and the constraints of conventional methods. Here, we utilize natural language processing (NLP) to text-mine 41,597 research articles, compiling a dataset of 2,838 phase transformations across 846 ferroelectric materials. Leveraging this dataset, we develop FerroAI, a deep learning model for phase diagram prediction. FerroAI successfully predicts phase boundaries and transformations among different crystal symmetries in Ce/Zr co-doped BaTiO$3$ (BT)-$x$Ba${0.7}$Ca$_{0.3}$TiO$_3$ (BCT). It also identifies a morphotropic phase boundary in Zr/Hf co-doped BT-$x$BCT at $x = 0.3$, guiding the discovery of a new ferroelectric material with an experimentally measured dielectric constant of 9535. These results establish FerroAI as a powerful tool for phase diagram construction, guiding the design of high-performance ferroelectric materials. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Chenbo",
""
],
[
"Chen",
"Xian",
""
]
] |
2506.10971 | Ye He | He Ye, Rojas Kevin, Tao Molei | What Exactly Does Guidance Do in Masked Discrete Diffusion Models | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | We study masked discrete diffusion models with classifier-free guidance (CFG). Assuming no score error nor discretization error, we derive an explicit solution to the guided reverse dynamics, so that how guidance influences the sampling behavior can be precisely characterized. When the full data distribution is a mixture over classes and the goal is to sample from a specific class, guidance amplifies class-specific regions while suppresses regions shared with other classes. This effect depends on the guidance strength $w$ and induces distinct covariance structures in the sampled distribution. Notably, we observe quantitatively different behaviors in $1$D and $2$D. We also show that for large $w$, the decay rate of the total variation ($\mathrm{TV}$) along the reverse dynamics is double-exponential in $w$ for both $1$D and $2$D. These findings highlight the role of guidance, not just in shaping the output distribution, but also in controlling the dynamics of the sampling trajectory. Our theoretical analysis is supported by experiments that illustrate the geometric effects of guidance and its impact on convergence. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ye",
"He",
""
],
[
"Kevin",
"Rojas",
""
],
[
"Molei",
"Tao",
""
]
] |
2506.10972 | Qiufeng Wang | Houyi Li, Wenzhen Zheng, Qiufeng Wang, Zhenyu Ding, Haoying Wang, Zili Wang, Shijie Xuyang, Ning Ding, Shuigeng Zhou, Xiangyu Zhang, Daxin Jiang | Farseer: A Refined Scaling Law in Large Language Models | 34 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training Large Language Models (LLMs) is prohibitively expensive, creating a critical scaling gap where insights from small-scale experiments often fail to transfer to resource-intensive production systems, thereby hindering efficient innovation. To bridge this, we introduce Farseer, a novel and refined scaling law offering enhanced predictive accuracy across scales. By systematically constructing a model loss surface $L(N,D)$, Farseer achieves a significantly better fit to empirical data than prior laws (e.g., Chinchilla's law). Our methodology yields accurate, robust, and highly generalizable predictions, demonstrating excellent extrapolation capabilities, improving upon Chinchilla's law by reducing extrapolation error by 433\%. This allows for the reliable evaluation of competing training strategies across all $(N,D)$ settings, enabling conclusions from small-scale ablation studies to be confidently extrapolated to predict large-scale performance. Furthermore, Farseer provides new insights into optimal compute allocation, better reflecting the nuanced demands of modern LLM training. To validate our approach, we trained an extensive suite of approximately 1,000 LLMs across diverse scales and configurations, consuming roughly 3 million NVIDIA H100 GPU hours. We are comprehensively open-sourcing all models, data, results, and logs at https://github.com/Farseer-Scaling-Law/Farseer to foster further research. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Li",
"Houyi",
""
],
[
"Zheng",
"Wenzhen",
""
],
[
"Wang",
"Qiufeng",
""
],
[
"Ding",
"Zhenyu",
""
],
[
"Wang",
"Haoying",
""
],
[
"Wang",
"Zili",
""
],
[
"Xuyang",
"Shijie",
""
],
[
"Ding",
"Ning",
""
],
[
"Zhou",
"Shuigeng",
""
],
[
"Zhang",
"Xiangyu",
""
],
[
"Jiang",
"Daxin",
""
]
] |
2506.10973 | Miguel Liu-Schiaffini | Julius Berner, Miguel Liu-Schiaffini, Jean Kossaifi, Valentin Duruisseaux, Boris Bonev, Kamyar Azizzadenesheli, Anima Anandkumar | Principled Approaches for Extending Neural Architectures to Function Spaces for Operator Learning | null | null | null | null | cs.LG cs.AI cs.NA math.FA math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A wide range of scientific problems, such as those described by continuous-time dynamical systems and partial differential equations (PDEs), are naturally formulated on function spaces. While function spaces are typically infinite-dimensional, deep learning has predominantly advanced through applications in computer vision and natural language processing that focus on mappings between finite-dimensional spaces. Such fundamental disparities in the nature of the data have limited neural networks from achieving a comparable level of success in scientific applications as seen in other fields. Neural operators are a principled way to generalize neural networks to mappings between function spaces, offering a pathway to replicate deep learning's transformative impact on scientific problems. For instance, neural operators can learn solution operators for entire classes of PDEs, e.g., physical systems with different boundary conditions, coefficient functions, and geometries. A key factor in deep learning's success has been the careful engineering of neural architectures through extensive empirical testing. Translating these neural architectures into neural operators allows operator learning to enjoy these same empirical optimizations. However, prior neural operator architectures have often been introduced as standalone models, not directly derived as extensions of existing neural network architectures. In this paper, we identify and distill the key principles for constructing practical implementations of mappings between infinite-dimensional function spaces. Using these principles, we propose a recipe for converting several popular neural architectures into neural operators with minimal modifications. This paper aims to guide practitioners through this process and details the steps to make neural operators work in practice. Our code can be found at https://github.com/neuraloperator/NNs-to-NOs | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:31 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Berner",
"Julius",
""
],
[
"Liu-Schiaffini",
"Miguel",
""
],
[
"Kossaifi",
"Jean",
""
],
[
"Duruisseaux",
"Valentin",
""
],
[
"Bonev",
"Boris",
""
],
[
"Azizzadenesheli",
"Kamyar",
""
],
[
"Anandkumar",
"Anima",
""
]
] |
2506.10974 | Ningyu Zhang | Yixin Ou, Yujie Luo, Jingsheng Zheng, Lanning Wei, Shuofei Qiao, Jintian Zhang, Da Zheng, Huajun Chen, Ningyu Zhang | AutoMind: Adaptive Knowledgeable Agent for Automated Data Science | Ongoing work. Code is at https://github.com/innovatingAI/AutoMind | null | null | null | cs.CL cs.AI cs.HC cs.LG cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Model (LLM) agents have shown great potential in addressing real-world data science problems. LLM-driven data science agents promise to automate the entire machine learning pipeline, yet their real-world effectiveness remains limited. Existing frameworks depend on rigid, pre-defined workflows and inflexible coding strategies; consequently, they excel only on relatively simple, classical problems and fail to capture the empirical expertise that human practitioners bring to complex, innovative tasks. In this work, we introduce AutoMind, an adaptive, knowledgeable LLM-agent framework that overcomes these deficiencies through three key advances: (1) a curated expert knowledge base that grounds the agent in domain expert knowledge, (2) an agentic knowledgeable tree search algorithm that strategically explores possible solutions, and (3) a self-adaptive coding strategy that dynamically tailors code generation to task complexity. Evaluations on two automated data science benchmarks demonstrate that AutoMind delivers superior performance versus state-of-the-art baselines. Additional analyses confirm favorable effectiveness, efficiency, and qualitative solution quality, highlighting AutoMind as an efficient and robust step toward fully automated data science. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ou",
"Yixin",
""
],
[
"Luo",
"Yujie",
""
],
[
"Zheng",
"Jingsheng",
""
],
[
"Wei",
"Lanning",
""
],
[
"Qiao",
"Shuofei",
""
],
[
"Zhang",
"Jintian",
""
],
[
"Zheng",
"Da",
""
],
[
"Chen",
"Huajun",
""
],
[
"Zhang",
"Ningyu",
""
]
] |
2506.10975 | Weiliang Chen | Weiliang Chen, Wenzhao Zheng, Yu Zheng, Lei Chen, Jie Zhou, Jiwen Lu, Yueqi Duan | GenWorld: Towards Detecting AI-generated Real-world Simulation Videos | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The flourishing of video generation technologies has endangered the credibility of real-world information and intensified the demand for AI-generated video detectors. Despite some progress, the lack of high-quality real-world datasets hinders the development of trustworthy detectors. In this paper, we propose GenWorld, a large-scale, high-quality, and real-world simulation dataset for AI-generated video detection. GenWorld features the following characteristics: (1) Real-world Simulation: GenWorld focuses on videos that replicate real-world scenarios, which have a significant impact due to their realism and potential influence; (2) High Quality: GenWorld employs multiple state-of-the-art video generation models to provide realistic and high-quality forged videos; (3) Cross-prompt Diversity: GenWorld includes videos generated from diverse generators and various prompt modalities (e.g., text, image, video), offering the potential to learn more generalizable forensic features. We analyze existing methods and find they fail to detect high-quality videos generated by world models (i.e., Cosmos), revealing potential drawbacks of ignoring real-world clues. To address this, we propose a simple yet effective model, SpannDetector, to leverage multi-view consistency as a strong criterion for real-world AI-generated video detection. Experiments show that our method achieves superior results, highlighting a promising direction for explainable AI-generated video detection based on physical plausibility. We believe that GenWorld will advance the field of AI-generated video detection. Project Page: https://chen-wl20.github.io/GenWorld | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:33 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chen",
"Weiliang",
""
],
[
"Zheng",
"Wenzhao",
""
],
[
"Zheng",
"Yu",
""
],
[
"Chen",
"Lei",
""
],
[
"Zhou",
"Jie",
""
],
[
"Lu",
"Jiwen",
""
],
[
"Duan",
"Yueqi",
""
]
] |
2506.10976 | Luka Rute\v{s}i\'c | Nata\v{s}a Krklec Jerinki\'c, Luka Rute\v{s}i\'c | ASMOP: Additional sampling stochastic trust region method for multi-objective problems | null | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider an unconstrained multi-criteria optimization problem with finite sum objective functions. The proposed algorithm belongs to a non-monotone trust-region framework where additional sampling approach is used to govern the sample size and the acceptance of a candidate point. Depending on the problem, the method can result in a mini-batch or an increasing sample size approach. Therefore, this work can be viewed as an extension of additional sampling trust region method for scalar finite sum function minimization presented in the literature. We show stochastic convergence of the proposed scheme for twice continuously-differentiable, but possibly non-convex objective functions, under assumptions standard for this framework. The experiments on logistic regression and least squares problems show the efficiency of the proposed scheme and its competitiveness with the relevant state-of-the-art methods for the considered problems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jerinkić",
"Nataša Krklec",
""
],
[
"Rutešić",
"Luka",
""
]
] |
2506.10977 | Sicheng Zuo | Sicheng Zuo, Wenzhao Zheng, Xiaoyong Han, Longchao Yang, Yong Pan, Jiwen Lu | QuadricFormer: Scene as Superquadrics for 3D Semantic Occupancy Prediction | Project page: https://zuosc19.github.io/QuadricFormer/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D occupancy prediction is crucial for robust autonomous driving systems as it enables comprehensive perception of environmental structures and semantics. Most existing methods employ dense voxel-based scene representations, ignoring the sparsity of driving scenes and resulting in inefficiency. Recent works explore object-centric representations based on sparse Gaussians, but their ellipsoidal shape prior limits the modeling of diverse structures. In real-world driving scenes, objects exhibit rich geometries (e.g., cuboids, cylinders, and irregular shapes), necessitating excessive ellipsoidal Gaussians densely packed for accurate modeling, which leads to inefficient representations. To address this, we propose to use geometrically expressive superquadrics as scene primitives, enabling efficient representation of complex structures with fewer primitives through their inherent shape diversity. We develop a probabilistic superquadric mixture model, which interprets each superquadric as an occupancy probability distribution with a corresponding geometry prior, and calculates semantics through probabilistic mixture. Building on this, we present QuadricFormer, a superquadric-based model for efficient 3D occupancy prediction, and introduce a pruning-and-splitting module to further enhance modeling efficiency by concentrating superquadrics in occupied regions. Extensive experiments on the nuScenes dataset demonstrate that QuadricFormer achieves state-of-the-art performance while maintaining superior efficiency. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zuo",
"Sicheng",
""
],
[
"Zheng",
"Wenzhao",
""
],
[
"Han",
"Xiaoyong",
""
],
[
"Yang",
"Longchao",
""
],
[
"Pan",
"Yong",
""
],
[
"Lu",
"Jiwen",
""
]
] |
2506.10978 | Donghoon Ahn | Donghoon Ahn, Jiwon Kang, Sanghyun Lee, Minjae Kim, Jaewon Min, Wooseok Jang, Saungwu Lee, Sayak Paul, Susung Hong, Seungryong Kim | Fine-Grained Perturbation Guidance via Attention Head Selection | Project page: https://cvlab-kaist.github.io/HeadHunter/ | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recent guidance methods in diffusion models steer reverse sampling by perturbing the model to construct an implicit weak model and guide generation away from it. Among these approaches, attention perturbation has demonstrated strong empirical performance in unconditional scenarios where classifier-free guidance is not applicable. However, existing attention perturbation methods lack principled approaches for determining where perturbations should be applied, particularly in Diffusion Transformer (DiT) architectures where quality-relevant computations are distributed across layers. In this paper, we investigate the granularity of attention perturbations, ranging from the layer level down to individual attention heads, and discover that specific heads govern distinct visual concepts such as structure, style, and texture quality. Building on this insight, we propose "HeadHunter", a systematic framework for iteratively selecting attention heads that align with user-centric objectives, enabling fine-grained control over generation quality and visual attributes. In addition, we introduce SoftPAG, which linearly interpolates each selected head's attention map toward an identity matrix, providing a continuous knob to tune perturbation strength and suppress artifacts. Our approach not only mitigates the oversmoothing issues of existing layer-level perturbation but also enables targeted manipulation of specific visual styles through compositional head selection. We validate our method on modern large-scale DiT-based text-to-image models including Stable Diffusion 3 and FLUX.1, demonstrating superior performance in both general quality enhancement and style-specific guidance. Our work provides the first head-level analysis of attention perturbation in diffusion models, uncovering interpretable specialization within attention layers and enabling practical design of effective perturbation strategies. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ahn",
"Donghoon",
""
],
[
"Kang",
"Jiwon",
""
],
[
"Lee",
"Sanghyun",
""
],
[
"Kim",
"Minjae",
""
],
[
"Min",
"Jaewon",
""
],
[
"Jang",
"Wooseok",
""
],
[
"Lee",
"Saungwu",
""
],
[
"Paul",
"Sayak",
""
],
[
"Hong",
"Susung",
""
],
[
"Kim",
"Seungryong",
""
]
] |
2506.10979 | Sohee Yang | Sohee Yang, Sang-Woo Lee, Nora Kassner, Daniela Gottesman, Sebastian Riedel, Mor Geva | How Well Can Reasoning Models Identify and Recover from Unhelpful Thoughts? | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recent reasoning models show the ability to reflect, backtrack, and self-validate their reasoning, which is crucial in spotting mistakes and arriving at accurate solutions. A natural question that arises is how effectively models can perform such self-reevaluation. We tackle this question by investigating how well reasoning models identify and recover from four types of unhelpful thoughts: uninformative rambling thoughts, thoughts irrelevant to the question, thoughts misdirecting the question as a slightly different question, and thoughts that lead to incorrect answers. We show that models are effective at identifying most unhelpful thoughts but struggle to recover from the same thoughts when these are injected into their thinking process, causing significant performance drops. Models tend to naively continue the line of reasoning of the injected irrelevant thoughts, which showcases that their self-reevaluation abilities are far from a general "meta-cognitive" awareness. Moreover, we observe non/inverse-scaling trends, where larger models struggle more than smaller ones to recover from short irrelevant thoughts, even when instructed to reevaluate their reasoning. We demonstrate the implications of these findings with a jailbreak experiment using irrelevant thought injection, showing that the smallest models are the least distracted by harmful-response-triggering thoughts. Overall, our findings call for improvement in self-reevaluation of reasoning models to develop better reasoning and safer systems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yang",
"Sohee",
""
],
[
"Lee",
"Sang-Woo",
""
],
[
"Kassner",
"Nora",
""
],
[
"Gottesman",
"Daniela",
""
],
[
"Riedel",
"Sebastian",
""
],
[
"Geva",
"Mor",
""
]
] |
2506.10980 | Junqi You | Junqi You, Chieh Hubert Lin, Weijie Lyu, Zhengbo Zhang, Ming-Hsuan Yang | InstaInpaint: Instant 3D-Scene Inpainting with Masked Large Reconstruction Model | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent advances in 3D scene reconstruction enable real-time viewing in virtual and augmented reality. To support interactive operations for better immersiveness, such as moving or editing objects, 3D scene inpainting methods are proposed to repair or complete the altered geometry. However, current approaches rely on lengthy and computationally intensive optimization, making them impractical for real-time or online applications. We propose InstaInpaint, a reference-based feed-forward framework that produces 3D-scene inpainting from a 2D inpainting proposal within 0.4 seconds. We develop a self-supervised masked-finetuning strategy to enable training of our custom large reconstruction model (LRM) on the large-scale dataset. Through extensive experiments, we analyze and identify several key designs that improve generalization, textural consistency, and geometric correctness. InstaInpaint achieves a 1000x speed-up from prior methods while maintaining a state-of-the-art performance across two standard benchmarks. Moreover, we show that InstaInpaint generalizes well to flexible downstream applications such as object insertion and multi-region inpainting. More video results are available at our project page: https://dhmbb2.github.io/InstaInpaint_page/. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"You",
"Junqi",
""
],
[
"Lin",
"Chieh Hubert",
""
],
[
"Lyu",
"Weijie",
""
],
[
"Zhang",
"Zhengbo",
""
],
[
"Yang",
"Ming-Hsuan",
""
]
] |
2506.10981 | Weiliang Chen | Weiliang Chen, Jiayi Bi, Yuanhui Huang, Wenzhao Zheng, Yueqi Duan | SceneCompleter: Dense 3D Scene Completion for Generative Novel View Synthesis | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Generative models have gained significant attention in novel view synthesis (NVS) by alleviating the reliance on dense multi-view captures. However, existing methods typically fall into a conventional paradigm, where generative models first complete missing areas in 2D, followed by 3D recovery techniques to reconstruct the scene, which often results in overly smooth surfaces and distorted geometry, as generative models struggle to infer 3D structure solely from RGB data. In this paper, we propose SceneCompleter, a novel framework that achieves 3D-consistent generative novel view synthesis through dense 3D scene completion. SceneCompleter achieves both visual coherence and 3D-consistent generative scene completion through two key components: (1) a geometry-appearance dual-stream diffusion model that jointly synthesizes novel views in RGBD space; (2) a scene embedder that encodes a more holistic scene understanding from the reference image. By effectively fusing structural and textural information, our method demonstrates superior coherence and plausibility in generative novel view synthesis across diverse datasets. Project Page: https://chen-wl20.github.io/SceneCompleter | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chen",
"Weiliang",
""
],
[
"Bi",
"Jiayi",
""
],
[
"Huang",
"Yuanhui",
""
],
[
"Zheng",
"Wenzhao",
""
],
[
"Duan",
"Yueqi",
""
]
] |
2506.10982 | Sebastian Sanokowski | Sebastian Sanokowski, Lukas Gruber, Christoph Bartmann, Sepp Hochreiter, Sebastian Lehner | Rethinking Losses for Diffusion Bridge Samplers | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion bridges are a promising class of deep-learning methods for sampling from unnormalized distributions. Recent works show that the Log Variance (LV) loss consistently outperforms the reverse Kullback-Leibler (rKL) loss when using the reparametrization trick to compute rKL-gradients. While the on-policy LV loss yields identical gradients to the rKL loss when combined with the log-derivative trick for diffusion samplers with non-learnable forward processes, this equivalence does not hold for diffusion bridges or when diffusion coefficients are learned. Based on this insight we argue that for diffusion bridges the LV loss does not represent an optimization objective that can be motivated like the rKL loss via the data processing inequality. Our analysis shows that employing the rKL loss with the log-derivative trick (rKL-LD) does not only avoid these conceptual problems but also consistently outperforms the LV loss. Experimental results with different types of diffusion bridges on challenging benchmarks show that samplers trained with the rKL-LD loss achieve better performance. From a practical perspective we find that rKL-LD requires significantly less hyperparameter optimization and yields more stable training behavior. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:59:58 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sanokowski",
"Sebastian",
""
],
[
"Gruber",
"Lukas",
""
],
[
"Bartmann",
"Christoph",
""
],
[
"Hochreiter",
"Sepp",
""
],
[
"Lehner",
"Sebastian",
""
]
] |
1109.1809 | Eric Thrane | J. Abadie, B. P. Abbott, R. Abbott, M. Abernathy, T. Accadia, F. Acernese, C. Adams, R. Adhikari, P. Ajith, B. Allen, G. S. Allen, E. Amador Ceron, R. S. Amin, S. B. Anderson, W. G. Anderson, F. Antonucci, M. A. Arain, M. C. Araya, M. Aronsson, K. G. Arun, Y. Aso, S. M. Aston, P. Astone, D. Atkinson, P. Aufmuth, C. Aulbert, S. Babak, P. Baker, G. Ballardin, S. Ballmer, D. Barker, S. Barnum, F. Barone, B. Barr, P. Barriga, L. Barsotti, M. Barsuglia, M. A. Barton, I. Bartos, R. Bassiri, M. Bastarrika, J. Bauchrowitz, Th. S. Bauer, B. Behnke, M.G. Beker, A. Belletoile, M. Benacquista, A. Bertolini, J. Betzwieser, N. Beveridge, P. T. Beyersdorf, S. Bigotta, I. A. Bilenko, G. Billingsley, J. Birch, S. Birindelli, R. Biswas, M. Bitossi, M. A. Bizouard, E. Black, J. K. Blackburn, L. Blackburn, D. Blair, B. Bland, M. Blom, C. Boccara, O. Bock, T. P. Bodiya, R. Bondarescu, F. Bondu, L. Bonelli, R. Bonnand, R. Bork, M. Born, S. Bose, L. Bosi, B. Bouhou, M. Boyle, S. Braccini, C. Bradaschia, P. R. Brady, V. B. Braginsky, J. E. Brau, J. Breyer, D. O. Bridges, A. Brillet, M. Brinkmann, V. Brisson, M. Britzger, A. F. Brooks, D. A. Brown, R. Budzy\'nski, T. Bulik, H. J. Bulten, A. Buonanno, J. Burguet--Castell, O. Burmeister, D. Buskulic, C. Buy, R. L. Byer, L. Cadonati, G. Cagnoli, J. Cain, E. Calloni, J. B. Camp, E. Campagna, P. Campsie, J. Cannizzo, K. Cannon, B. Canuel, J. Cao, C. Capano, F. Carbognani, S. Caride, S. Caudill, M. Cavagli\`a, F. Cavalier, R. Cavalieri, G. Cella, C. Cepeda, E. Cesarini, T. Chalermsongsak, E. Chalkley, P. Charlton, E. Chassande-Mottin, S. Chelkowski, Y. Chen, A. Chincarini, N. Christensen, S. S. Y. Chua, C. T. Y. Chung, D. Clark, J. Clark, J. H. Clayton, F. Cleva, E. Coccia, C. N. Colacino, J. Colas, A. Colla, M. Colombini, R. Conte, D. Cook, T. R. Corbitt, N. Cornish, A. Corsi, C. A. Costa, J.-P. Coulon, D. M. Coward, D. C. Coyne, J. D. E. Creighton, T. D. Creighton, A. M. Cruise, R. M. Culter, A. Cumming, L. Cunningham, E. Cuoco, K. Dahl, S. L. Danilishin, R. Dannenberg, S. D'Antonio, K. Danzmann, K. Das, V. Dattilo, B. Daudert, M. Davier, G. Davies, A. Davis, E. J. Daw, R. Day, T. Dayanga, R. De Rosa, D. DeBra, J. Degallaix, M. del Prete, V. Dergachev, R. DeRosa, R. DeSalvo, P. Devanka, S. Dhurandhar, L. Di Fiore, A. Di Lieto, I. Di Palma, M. Di Paolo Emilio, A. Di Virgilio, M. D\'iaz, A. Dietz, F. Donovan, K. L. Dooley, E. E. Doomes, S. Dorsher, E. S. D. Douglas, M. Drago, R. W. P. Drever, J. C. Driggers, J. Dueck, J.-C. Dumas, T. Eberle, M. Edgar, M. Edwards, A. Effler, P. Ehrens, R. Engel, T. Etzel, M. Evans, T. Evans, V. Fafone, S. Fairhurst, Y. Fan, B. F. Farr, D. Fazi, H. Fehrmann, D. Feldbaum, I. Ferrante, F. Fidecaro, L. S. Finn, I. Fiori, R. Flaminio, M. Flanigan, K. Flasch, S. Foley, C. Forrest, E. Forsi, N. Fotopoulos, J.-D. Fournier, J. Franc, S. Frasca, F. Frasconi, M. Frede, M. Frei, Z. Frei, A. Freise, R. Frey, T. T. Fricke, D. Friedrich, P. Fritschel, V. V. Frolov, P. Fulda, M. Fyffe, M. Galimberti, L. Gammaitoni, J. A. Garofoli, F. Garufi, G. Gemme, E. Genin, A. Gennai, I. Gholami, S. Ghosh, J. A. Giaime, S. Giampanis, K. D. Giardina, A. Giazotto, C. Gill, E. Goetz, L. M. Goggin, G. Gonz\'alez, M. L. Gorodetsky, S. Go{\ss}ler, R. Gouaty, C. Graef, M. Granata, A. Grant, S. Gras, C. Gray, R. J. S. Greenhalgh, A. M. Gretarsson, C. Greverie, R. Grosso, H. Grote, S. Grunewald, G. M. Guidi, E. K. Gustafson, R. Gustafson, B. Hage, P. Hall, J. M. Hallam, D. Hammer, G. Hammond, J. Hanks, C. Hanna, J. Hanson, J. Harms, G. M. Harry, I. W. Harry, E. D. Harstad, K. Haughian, K. Hayama, J.-F. Hayau, T. Hayler, J. Heefner, H. Heitmann, P. Hello, I. S. Heng, A. W. Heptonstall, M. Hewitson, S. Hild, E. Hirose, D. Hoak, K. A. Hodge, K. Holt, D. J. Hosken, J. Hough, E. J. Howell, D. Hoyland, D. Huet, B. Hughey, S. Husa, S. H. Huttner, T. Huynh--Dinh, D. R. Ingram, R. Inta, T. Isogai, A. Ivanov, P. Jaranowski, W. W. Johnson, D. I. Jones, G. Jones, R. Jones, L. Ju, P. Kalmus, V. Kalogera, S. Kandhasamy, J. B. Kanner, E. Katsavounidis, K. Kawabe, S. Kawamura, F. Kawazoe, W. Kells, D. G. Keppel, A. Khalaidovski, F. Y. Khalili, E. A. Khazanov, H. Kim, P. J. King, D. L. Kinzel, J. S. Kissel, S. Klimenko, V. Kondrashov, R. Kopparapu, S. Koranda, I. Kowalska, D. Kozak, T. Krause, V. Kringel, S. Krishnamurthy, B. Krishnan, A. Kr\'olak, G. Kuehn, J. Kullman, R. Kumar, P. Kwee, M. Landry, M. Lang, B. Lantz, N. Lastzka, A. Lazzarini, P. Leaci, J. Leong, I. Leonor, N. Leroy, N. Letendre, J. Li, T. G. F. Li, N. Liguori, H. Lin, P. E. Lindquist, N. A. Lockerbie, D. Lodhia, M. Lorenzini, V. Loriette, M. Lormand, G. Losurdo, P. Lu, J. Luan, M. Lubinski, A. Lucianetti, H. L\"uck, A. D. Lundgren, B. Machenschalk, M. MacInnis, M. Mageswaran, K. Mailand, E. Majorana, C. Mak, I. Maksimovic, N. Man, I. Mandel, V. Mandic, M. Mantovani, F. Marchesoni, F. Marion, S. M\'arka, Z. M\'arka, E. Maros, J. Marque, F. Martelli, I. W. Martin, R. M. Martin, J. N. Marx, K. Mason, A. Masserot, F. Matichard, L. Matone, R. A. Matzner, N. Mavalvala, R. McCarthy, D. E. McClelland, S. C. McGuire, G. McIntyre, G. McIvor, D. J. A. McKechan, G. Meadors, M. Mehmet, T. Meier, A. Melatos, A. C. Melissinos, G. Mendell, D. F. Men\'endez, R. A. Mercer, L. Merill, S. Meshkov, C. Messenger, M. S. Meyer, H. Miao, C. Michel, L. Milano, J. Miller, Y. Minenkov, Y. Mino, S. Mitra, V. P. Mitrofanov, G. Mitselmakher, R. Mittleman, B. Moe, M. Mohan, S. D. Mohanty, S. R. P. Mohapatra, D. Moraru, J. Moreau, G. Moreno, N. Morgado, A. Morgia, T. Morioka, K. Mors, S. Mosca, V. Moscatelli, K. Mossavi, B. Mours, C. M. Mow--Lowry, G. Mueller, S. Mukherjee, A. Mullavey, H. M\"uller-Ebhardt, J. Munch, P. G. Murray, T. Nash, R. Nawrodt, J. Nelson, I. Neri, G. Newton, A. Nishizawa, F. Nocera, D. Nolting, E. Ochsner, J. O'Dell, G. H. Ogin, R. G. Oldenburg, B. O'Reilly, R. O'Shaughnessy, C. Osthelder, D. J. Ottaway, R. S. Ottens, H. Overmier, B. J. Owen, A. Page, G. Pagliaroli, L. Palladino, C. Palomba, Y. Pan, C. Pankow, F. Paoletti, M. A. Papa, S. Pardi, M. Pareja, M. Parisi, A. Pasqualetti, R. Passaquieti, D. Passuello, P. Patel, D. Pathak, M. Pedraza, L. Pekowsky, S. Penn, C. Peralta, A. Perreca, G. Persichetti, M. Pichot, M. Pickenpack, F. Piergiovanni, M. Pietka, L. Pinard, I. M. Pinto, M. Pitkin, H. J. Pletsch, M. V. Plissi, R. Poggiani, F. Postiglione, M. Prato, V. Predoi, L. R. Price, M. Prijatelj, M. Principe, R. Prix, G. A. Prodi, L. Prokhorov, O. Puncken, M. Punturo, P. Puppo, V. Quetschke, F. J. Raab, D. S. Rabeling, T. Radke, H. Radkins, P. Raffai, M. Rakhmanov, B. Rankins, P. Rapagnani, V. Raymond, V. Re, C. M. Reed, T. Reed, T. Regimbau, S. Reid, D. H. Reitze, F. Ricci, R. Riesen, K. Riles, P. Roberts, N. A. Robertson, F. Robinet, C. Robinson, E. L. Robinson, A. Rocchi, S. Roddy, C. R\"over, L. Rolland, J. Rollins, J. D. Romano, R. Romano, J. H. Romie, D. Rosi\'nska, S. Rowan, A. R\"udiger, P. Ruggi, K. Ryan, S. Sakata, M. Sakosky, F. Salemi, L. Sammut, L. Sancho de la Jordana, V. Sandberg, V. Sannibale, L. Santamar\'ia, G. Santostasi, S. Saraf, B. Sassolas, B. S. Sathyaprakash, S. Sato, M. Satterthwaite, P. R. Saulson, R. Savage, R. Schilling, R. Schnabel, R. M. S. Schofield, B. Schulz, B. F. Schutz, P. Schwinberg, J. Scott, S. M. Scott, A. C. Searle, F. Seifert, D. Sellers, A. S. Sengupta, D. Sentenac, A. Sergeev, D. A. Shaddock, B. Shapiro, P. Shawhan, D. H. Shoemaker, A. Sibley, X. Siemens, D. Sigg, A. Singer, A. M. Sintes, G. Skelton, B. J. J. Slagmolen, J. Slutsky, J. R. Smith, M. R. Smith, N. D. Smith, K. Somiya, B. Sorazu, F. C. Speirits, L. Sperandio, A. J. Stein, L. C. Stein, S. Steinlechner, S. Steplewski, A. Stochino, R. Stone, K. A. Strain, S. Strigin, A. S. Stroeer, R. Sturani, A. L. Stuver, T. Z. Summerscales, M. Sung, S. Susmithan, P. J. Sutton, B. Swinkels, G. P. Szokoly, D. Talukder, D. B. Tanner, S. P. Tarabrin, J. R. Taylor, R. Taylor, P. Thomas, K. A. Thorne, K. S. Thorne, E. Thrane, A. Th\"uring, C. Titsler, K. V. Tokmakov, A. Toncelli, M. Tonelli, O. Torre, C. Torres, C. I. Torrie, E. Tournefier, F. Travasso, G. Traylor, M. Trias, J. Trummer, K. Tseng, L. Turner, D. Ugolini, K. Urbanek, H. Vahlbruch, B. Vaishnav, G. Vajente, M. Vallisneri, J. F. J. van den Brand, C. Van Den Broeck, S. van der Putten, M. V. van der Sluys, A. A. van Veggel, S. Vass, R. Vaulin, M. Vavoulidis, A. Vecchio, G. Vedovato, J. Veitch, P. J. Veitch, C. Veltkamp, D. Verkindt, F. Vetrano, A. Vicer\'e, A. E. Villar, J.-Y. Vinet, H. Vocca, C. Vorvick, S. P. Vyachanin, S. J. Waldman, L. Wallace, A. Wanner, R. L. Ward, M. Was, P. Wei, M. Weinert, A. J. Weinstein, R. Weiss, L. Wen, S. Wen, P. Wessels, M. West, T. Westphal, K. Wette, J. T. Whelan, S. E. Whitcomb, D. White, B. F. Whiting, C. Wilkinson, P. A. Willems, L. Williams, B. Willke, L. Winkelmann, W. Winkler, C. C. Wipf, A. G. Wiseman, G. Woan, R. Wooley, J. Worden, I. Yakushin, H. Yamamoto, K. Yamamoto, D. Yeaton-Massey, S. Yoshida, P. Yu, M. Yvert, M. Zanolin, L. Zhang, Z. Zhang, C. Zhao, N. Zotov, M. E. Zucker, J. Zweizig | Directional limits on persistent gravitational waves using LIGO S5 science data | 10 pages, 4 figures | Phys. Rev. Lett. 107, 271102 (2011) | 10.1103/PhysRevLett.107.271102 | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The gravitational-wave (GW) sky may include nearby pointlike sources as well as astrophysical and cosmological stochastic backgrounds. Since the relative strength and angular distribution of the many possible sources of GWs are not well constrained, searches for GW signals must be performed in a model-independent way. To that end we perform two directional searches for persistent GWs using data from the LIGO S5 science run: one optimized for pointlike sources and one for arbitrary extended sources. The latter result is the first of its kind. Finding no evidence to support the detection of GWs, we present 90% confidence level (CL) upper-limit maps of GW strain power with typical values between 2-20x10^-50 strain^2 Hz^-1 and 5-35x10^-49 strain^2 Hz^-1 sr^-1 for pointlike and extended sources respectively. The limits on pointlike sources constitute a factor of 30 improvement over the previous best limits. We also set 90% CL limits on the narrow-band root-mean-square GW strain from interesting targets including Sco X-1, SN1987A and the Galactic Center as low as ~7x10^-25 in the most sensitive frequency range near 160 Hz. These limits are the most constraining to date and constitute a factor of 5 improvement over the previous best limits. | [
{
"version": "v1",
"created": "Thu, 08 Sep 2011 19:10:40 GMT"
},
{
"version": "v2",
"created": "Fri, 09 Sep 2011 15:32:15 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Abadie",
"J.",
""
],
[
"Abbott",
"B. P.",
""
],
[
"Abbott",
"R.",
""
],
[
"Abernathy",
"M.",
""
],
[
"Accadia",
"T.",
""
],
[
"Acernese",
"F.",
""
],
[
"Adams",
"C.",
""
],
[
"Adhikari",
"R.",
""
],
[
"Ajith",
"P.",
""
],
[
"Allen",
"B.",
""
],
[
"Allen",
"G. S.",
""
],
[
"Ceron",
"E. Amador",
""
],
[
"Amin",
"R. S.",
""
],
[
"Anderson",
"S. B.",
""
],
[
"Anderson",
"W. G.",
""
],
[
"Antonucci",
"F.",
""
],
[
"Arain",
"M. A.",
""
],
[
"Araya",
"M. C.",
""
],
[
"Aronsson",
"M.",
""
],
[
"Arun",
"K. G.",
""
],
[
"Aso",
"Y.",
""
],
[
"Aston",
"S. M.",
""
],
[
"Astone",
"P.",
""
],
[
"Atkinson",
"D.",
""
],
[
"Aufmuth",
"P.",
""
],
[
"Aulbert",
"C.",
""
],
[
"Babak",
"S.",
""
],
[
"Baker",
"P.",
""
],
[
"Ballardin",
"G.",
""
],
[
"Ballmer",
"S.",
""
],
[
"Barker",
"D.",
""
],
[
"Barnum",
"S.",
""
],
[
"Barone",
"F.",
""
],
[
"Barr",
"B.",
""
],
[
"Barriga",
"P.",
""
],
[
"Barsotti",
"L.",
""
],
[
"Barsuglia",
"M.",
""
],
[
"Barton",
"M. A.",
""
],
[
"Bartos",
"I.",
""
],
[
"Bassiri",
"R.",
""
],
[
"Bastarrika",
"M.",
""
],
[
"Bauchrowitz",
"J.",
""
],
[
"Bauer",
"Th. S.",
""
],
[
"Behnke",
"B.",
""
],
[
"Beker",
"M. G.",
""
],
[
"Belletoile",
"A.",
""
],
[
"Benacquista",
"M.",
""
],
[
"Bertolini",
"A.",
""
],
[
"Betzwieser",
"J.",
""
],
[
"Beveridge",
"N.",
""
],
[
"Beyersdorf",
"P. T.",
""
],
[
"Bigotta",
"S.",
""
],
[
"Bilenko",
"I. A.",
""
],
[
"Billingsley",
"G.",
""
],
[
"Birch",
"J.",
""
],
[
"Birindelli",
"S.",
""
],
[
"Biswas",
"R.",
""
],
[
"Bitossi",
"M.",
""
],
[
"Bizouard",
"M. A.",
""
],
[
"Black",
"E.",
""
],
[
"Blackburn",
"J. K.",
""
],
[
"Blackburn",
"L.",
""
],
[
"Blair",
"D.",
""
],
[
"Bland",
"B.",
""
],
[
"Blom",
"M.",
""
],
[
"Boccara",
"C.",
""
],
[
"Bock",
"O.",
""
],
[
"Bodiya",
"T. P.",
""
],
[
"Bondarescu",
"R.",
""
],
[
"Bondu",
"F.",
""
],
[
"Bonelli",
"L.",
""
],
[
"Bonnand",
"R.",
""
],
[
"Bork",
"R.",
""
],
[
"Born",
"M.",
""
],
[
"Bose",
"S.",
""
],
[
"Bosi",
"L.",
""
],
[
"Bouhou",
"B.",
""
],
[
"Boyle",
"M.",
""
],
[
"Braccini",
"S.",
""
],
[
"Bradaschia",
"C.",
""
],
[
"Brady",
"P. R.",
""
],
[
"Braginsky",
"V. B.",
""
],
[
"Brau",
"J. E.",
""
],
[
"Breyer",
"J.",
""
],
[
"Bridges",
"D. O.",
""
],
[
"Brillet",
"A.",
""
],
[
"Brinkmann",
"M.",
""
],
[
"Brisson",
"V.",
""
],
[
"Britzger",
"M.",
""
],
[
"Brooks",
"A. F.",
""
],
[
"Brown",
"D. A.",
""
],
[
"Budzyński",
"R.",
""
],
[
"Bulik",
"T.",
""
],
[
"Bulten",
"H. J.",
""
],
[
"Buonanno",
"A.",
""
],
[
"Burguet--Castell",
"J.",
""
],
[
"Burmeister",
"O.",
""
],
[
"Buskulic",
"D.",
""
],
[
"Buy",
"C.",
""
],
[
"Byer",
"R. L.",
""
],
[
"Cadonati",
"L.",
""
],
[
"Cagnoli",
"G.",
""
],
[
"Cain",
"J.",
""
],
[
"Calloni",
"E.",
""
],
[
"Camp",
"J. B.",
""
],
[
"Campagna",
"E.",
""
],
[
"Campsie",
"P.",
""
],
[
"Cannizzo",
"J.",
""
],
[
"Cannon",
"K.",
""
],
[
"Canuel",
"B.",
""
],
[
"Cao",
"J.",
""
],
[
"Capano",
"C.",
""
],
[
"Carbognani",
"F.",
""
],
[
"Caride",
"S.",
""
],
[
"Caudill",
"S.",
""
],
[
"Cavaglià",
"M.",
""
],
[
"Cavalier",
"F.",
""
],
[
"Cavalieri",
"R.",
""
],
[
"Cella",
"G.",
""
],
[
"Cepeda",
"C.",
""
],
[
"Cesarini",
"E.",
""
],
[
"Chalermsongsak",
"T.",
""
],
[
"Chalkley",
"E.",
""
],
[
"Charlton",
"P.",
""
],
[
"Chassande-Mottin",
"E.",
""
],
[
"Chelkowski",
"S.",
""
],
[
"Chen",
"Y.",
""
],
[
"Chincarini",
"A.",
""
],
[
"Christensen",
"N.",
""
],
[
"Chua",
"S. S. Y.",
""
],
[
"Chung",
"C. T. Y.",
""
],
[
"Clark",
"D.",
""
],
[
"Clark",
"J.",
""
],
[
"Clayton",
"J. H.",
""
],
[
"Cleva",
"F.",
""
],
[
"Coccia",
"E.",
""
],
[
"Colacino",
"C. N.",
""
],
[
"Colas",
"J.",
""
],
[
"Colla",
"A.",
""
],
[
"Colombini",
"M.",
""
],
[
"Conte",
"R.",
""
],
[
"Cook",
"D.",
""
],
[
"Corbitt",
"T. R.",
""
],
[
"Cornish",
"N.",
""
],
[
"Corsi",
"A.",
""
],
[
"Costa",
"C. A.",
""
],
[
"Coulon",
"J. -P.",
""
],
[
"Coward",
"D. M.",
""
],
[
"Coyne",
"D. C.",
""
],
[
"Creighton",
"J. D. E.",
""
],
[
"Creighton",
"T. D.",
""
],
[
"Cruise",
"A. M.",
""
],
[
"Culter",
"R. M.",
""
],
[
"Cumming",
"A.",
""
],
[
"Cunningham",
"L.",
""
],
[
"Cuoco",
"E.",
""
],
[
"Dahl",
"K.",
""
],
[
"Danilishin",
"S. L.",
""
],
[
"Dannenberg",
"R.",
""
],
[
"D'Antonio",
"S.",
""
],
[
"Danzmann",
"K.",
""
],
[
"Das",
"K.",
""
],
[
"Dattilo",
"V.",
""
],
[
"Daudert",
"B.",
""
],
[
"Davier",
"M.",
""
],
[
"Davies",
"G.",
""
],
[
"Davis",
"A.",
""
],
[
"Daw",
"E. J.",
""
],
[
"Day",
"R.",
""
],
[
"Dayanga",
"T.",
""
],
[
"De Rosa",
"R.",
""
],
[
"DeBra",
"D.",
""
],
[
"Degallaix",
"J.",
""
],
[
"del Prete",
"M.",
""
],
[
"Dergachev",
"V.",
""
],
[
"DeRosa",
"R.",
""
],
[
"DeSalvo",
"R.",
""
],
[
"Devanka",
"P.",
""
],
[
"Dhurandhar",
"S.",
""
],
[
"Di Fiore",
"L.",
""
],
[
"Di Lieto",
"A.",
""
],
[
"Di Palma",
"I.",
""
],
[
"Emilio",
"M. Di Paolo",
""
],
[
"Di Virgilio",
"A.",
""
],
[
"Díaz",
"M.",
""
],
[
"Dietz",
"A.",
""
],
[
"Donovan",
"F.",
""
],
[
"Dooley",
"K. L.",
""
],
[
"Doomes",
"E. E.",
""
],
[
"Dorsher",
"S.",
""
],
[
"Douglas",
"E. S. D.",
""
],
[
"Drago",
"M.",
""
],
[
"Drever",
"R. W. P.",
""
],
[
"Driggers",
"J. C.",
""
],
[
"Dueck",
"J.",
""
],
[
"Dumas",
"J. -C.",
""
],
[
"Eberle",
"T.",
""
],
[
"Edgar",
"M.",
""
],
[
"Edwards",
"M.",
""
],
[
"Effler",
"A.",
""
],
[
"Ehrens",
"P.",
""
],
[
"Engel",
"R.",
""
],
[
"Etzel",
"T.",
""
],
[
"Evans",
"M.",
""
],
[
"Evans",
"T.",
""
],
[
"Fafone",
"V.",
""
],
[
"Fairhurst",
"S.",
""
],
[
"Fan",
"Y.",
""
],
[
"Farr",
"B. F.",
""
],
[
"Fazi",
"D.",
""
],
[
"Fehrmann",
"H.",
""
],
[
"Feldbaum",
"D.",
""
],
[
"Ferrante",
"I.",
""
],
[
"Fidecaro",
"F.",
""
],
[
"Finn",
"L. S.",
""
],
[
"Fiori",
"I.",
""
],
[
"Flaminio",
"R.",
""
],
[
"Flanigan",
"M.",
""
],
[
"Flasch",
"K.",
""
],
[
"Foley",
"S.",
""
],
[
"Forrest",
"C.",
""
],
[
"Forsi",
"E.",
""
],
[
"Fotopoulos",
"N.",
""
],
[
"Fournier",
"J. -D.",
""
],
[
"Franc",
"J.",
""
],
[
"Frasca",
"S.",
""
],
[
"Frasconi",
"F.",
""
],
[
"Frede",
"M.",
""
],
[
"Frei",
"M.",
""
],
[
"Frei",
"Z.",
""
],
[
"Freise",
"A.",
""
],
[
"Frey",
"R.",
""
],
[
"Fricke",
"T. T.",
""
],
[
"Friedrich",
"D.",
""
],
[
"Fritschel",
"P.",
""
],
[
"Frolov",
"V. V.",
""
],
[
"Fulda",
"P.",
""
],
[
"Fyffe",
"M.",
""
],
[
"Galimberti",
"M.",
""
],
[
"Gammaitoni",
"L.",
""
],
[
"Garofoli",
"J. A.",
""
],
[
"Garufi",
"F.",
""
],
[
"Gemme",
"G.",
""
],
[
"Genin",
"E.",
""
],
[
"Gennai",
"A.",
""
],
[
"Gholami",
"I.",
""
],
[
"Ghosh",
"S.",
""
],
[
"Giaime",
"J. A.",
""
],
[
"Giampanis",
"S.",
""
],
[
"Giardina",
"K. D.",
""
],
[
"Giazotto",
"A.",
""
],
[
"Gill",
"C.",
""
],
[
"Goetz",
"E.",
""
],
[
"Goggin",
"L. M.",
""
],
[
"González",
"G.",
""
],
[
"Gorodetsky",
"M. L.",
""
],
[
"Goßler",
"S.",
""
],
[
"Gouaty",
"R.",
""
],
[
"Graef",
"C.",
""
],
[
"Granata",
"M.",
""
],
[
"Grant",
"A.",
""
],
[
"Gras",
"S.",
""
],
[
"Gray",
"C.",
""
],
[
"Greenhalgh",
"R. J. S.",
""
],
[
"Gretarsson",
"A. M.",
""
],
[
"Greverie",
"C.",
""
],
[
"Grosso",
"R.",
""
],
[
"Grote",
"H.",
""
],
[
"Grunewald",
"S.",
""
],
[
"Guidi",
"G. M.",
""
],
[
"Gustafson",
"E. K.",
""
],
[
"Gustafson",
"R.",
""
],
[
"Hage",
"B.",
""
],
[
"Hall",
"P.",
""
],
[
"Hallam",
"J. M.",
""
],
[
"Hammer",
"D.",
""
],
[
"Hammond",
"G.",
""
],
[
"Hanks",
"J.",
""
],
[
"Hanna",
"C.",
""
],
[
"Hanson",
"J.",
""
],
[
"Harms",
"J.",
""
],
[
"Harry",
"G. M.",
""
],
[
"Harry",
"I. W.",
""
],
[
"Harstad",
"E. D.",
""
],
[
"Haughian",
"K.",
""
],
[
"Hayama",
"K.",
""
],
[
"Hayau",
"J. -F.",
""
],
[
"Hayler",
"T.",
""
],
[
"Heefner",
"J.",
""
],
[
"Heitmann",
"H.",
""
],
[
"Hello",
"P.",
""
],
[
"Heng",
"I. S.",
""
],
[
"Heptonstall",
"A. W.",
""
],
[
"Hewitson",
"M.",
""
],
[
"Hild",
"S.",
""
],
[
"Hirose",
"E.",
""
],
[
"Hoak",
"D.",
""
],
[
"Hodge",
"K. A.",
""
],
[
"Holt",
"K.",
""
],
[
"Hosken",
"D. J.",
""
],
[
"Hough",
"J.",
""
],
[
"Howell",
"E. J.",
""
],
[
"Hoyland",
"D.",
""
],
[
"Huet",
"D.",
""
],
[
"Hughey",
"B.",
""
],
[
"Husa",
"S.",
""
],
[
"Huttner",
"S. H.",
""
],
[
"Huynh--Dinh",
"T.",
""
],
[
"Ingram",
"D. R.",
""
],
[
"Inta",
"R.",
""
],
[
"Isogai",
"T.",
""
],
[
"Ivanov",
"A.",
""
],
[
"Jaranowski",
"P.",
""
],
[
"Johnson",
"W. W.",
""
],
[
"Jones",
"D. I.",
""
],
[
"Jones",
"G.",
""
],
[
"Jones",
"R.",
""
],
[
"Ju",
"L.",
""
],
[
"Kalmus",
"P.",
""
],
[
"Kalogera",
"V.",
""
],
[
"Kandhasamy",
"S.",
""
],
[
"Kanner",
"J. B.",
""
],
[
"Katsavounidis",
"E.",
""
],
[
"Kawabe",
"K.",
""
],
[
"Kawamura",
"S.",
""
],
[
"Kawazoe",
"F.",
""
],
[
"Kells",
"W.",
""
],
[
"Keppel",
"D. G.",
""
],
[
"Khalaidovski",
"A.",
""
],
[
"Khalili",
"F. Y.",
""
],
[
"Khazanov",
"E. A.",
""
],
[
"Kim",
"H.",
""
],
[
"King",
"P. J.",
""
],
[
"Kinzel",
"D. L.",
""
],
[
"Kissel",
"J. S.",
""
],
[
"Klimenko",
"S.",
""
],
[
"Kondrashov",
"V.",
""
],
[
"Kopparapu",
"R.",
""
],
[
"Koranda",
"S.",
""
],
[
"Kowalska",
"I.",
""
],
[
"Kozak",
"D.",
""
],
[
"Krause",
"T.",
""
],
[
"Kringel",
"V.",
""
],
[
"Krishnamurthy",
"S.",
""
],
[
"Krishnan",
"B.",
""
],
[
"Królak",
"A.",
""
],
[
"Kuehn",
"G.",
""
],
[
"Kullman",
"J.",
""
],
[
"Kumar",
"R.",
""
],
[
"Kwee",
"P.",
""
],
[
"Landry",
"M.",
""
],
[
"Lang",
"M.",
""
],
[
"Lantz",
"B.",
""
],
[
"Lastzka",
"N.",
""
],
[
"Lazzarini",
"A.",
""
],
[
"Leaci",
"P.",
""
],
[
"Leong",
"J.",
""
],
[
"Leonor",
"I.",
""
],
[
"Leroy",
"N.",
""
],
[
"Letendre",
"N.",
""
],
[
"Li",
"J.",
""
],
[
"Li",
"T. G. F.",
""
],
[
"Liguori",
"N.",
""
],
[
"Lin",
"H.",
""
],
[
"Lindquist",
"P. E.",
""
],
[
"Lockerbie",
"N. A.",
""
],
[
"Lodhia",
"D.",
""
],
[
"Lorenzini",
"M.",
""
],
[
"Loriette",
"V.",
""
],
[
"Lormand",
"M.",
""
],
[
"Losurdo",
"G.",
""
],
[
"Lu",
"P.",
""
],
[
"Luan",
"J.",
""
],
[
"Lubinski",
"M.",
""
],
[
"Lucianetti",
"A.",
""
],
[
"Lück",
"H.",
""
],
[
"Lundgren",
"A. D.",
""
],
[
"Machenschalk",
"B.",
""
],
[
"MacInnis",
"M.",
""
],
[
"Mageswaran",
"M.",
""
],
[
"Mailand",
"K.",
""
],
[
"Majorana",
"E.",
""
],
[
"Mak",
"C.",
""
],
[
"Maksimovic",
"I.",
""
],
[
"Man",
"N.",
""
],
[
"Mandel",
"I.",
""
],
[
"Mandic",
"V.",
""
],
[
"Mantovani",
"M.",
""
],
[
"Marchesoni",
"F.",
""
],
[
"Marion",
"F.",
""
],
[
"Márka",
"S.",
""
],
[
"Márka",
"Z.",
""
],
[
"Maros",
"E.",
""
],
[
"Marque",
"J.",
""
],
[
"Martelli",
"F.",
""
],
[
"Martin",
"I. W.",
""
],
[
"Martin",
"R. M.",
""
],
[
"Marx",
"J. N.",
""
],
[
"Mason",
"K.",
""
],
[
"Masserot",
"A.",
""
],
[
"Matichard",
"F.",
""
],
[
"Matone",
"L.",
""
],
[
"Matzner",
"R. A.",
""
],
[
"Mavalvala",
"N.",
""
],
[
"McCarthy",
"R.",
""
],
[
"McClelland",
"D. E.",
""
],
[
"McGuire",
"S. C.",
""
],
[
"McIntyre",
"G.",
""
],
[
"McIvor",
"G.",
""
],
[
"McKechan",
"D. J. A.",
""
],
[
"Meadors",
"G.",
""
],
[
"Mehmet",
"M.",
""
],
[
"Meier",
"T.",
""
],
[
"Melatos",
"A.",
""
],
[
"Melissinos",
"A. C.",
""
],
[
"Mendell",
"G.",
""
],
[
"Menéndez",
"D. F.",
""
],
[
"Mercer",
"R. A.",
""
],
[
"Merill",
"L.",
""
],
[
"Meshkov",
"S.",
""
],
[
"Messenger",
"C.",
""
],
[
"Meyer",
"M. S.",
""
],
[
"Miao",
"H.",
""
],
[
"Michel",
"C.",
""
],
[
"Milano",
"L.",
""
],
[
"Miller",
"J.",
""
],
[
"Minenkov",
"Y.",
""
],
[
"Mino",
"Y.",
""
],
[
"Mitra",
"S.",
""
],
[
"Mitrofanov",
"V. P.",
""
],
[
"Mitselmakher",
"G.",
""
],
[
"Mittleman",
"R.",
""
],
[
"Moe",
"B.",
""
],
[
"Mohan",
"M.",
""
],
[
"Mohanty",
"S. D.",
""
],
[
"Mohapatra",
"S. R. P.",
""
],
[
"Moraru",
"D.",
""
],
[
"Moreau",
"J.",
""
],
[
"Moreno",
"G.",
""
],
[
"Morgado",
"N.",
""
],
[
"Morgia",
"A.",
""
],
[
"Morioka",
"T.",
""
],
[
"Mors",
"K.",
""
],
[
"Mosca",
"S.",
""
],
[
"Moscatelli",
"V.",
""
],
[
"Mossavi",
"K.",
""
],
[
"Mours",
"B.",
""
],
[
"Mow--Lowry",
"C. M.",
""
],
[
"Mueller",
"G.",
""
],
[
"Mukherjee",
"S.",
""
],
[
"Mullavey",
"A.",
""
],
[
"Müller-Ebhardt",
"H.",
""
],
[
"Munch",
"J.",
""
],
[
"Murray",
"P. G.",
""
],
[
"Nash",
"T.",
""
],
[
"Nawrodt",
"R.",
""
],
[
"Nelson",
"J.",
""
],
[
"Neri",
"I.",
""
],
[
"Newton",
"G.",
""
],
[
"Nishizawa",
"A.",
""
],
[
"Nocera",
"F.",
""
],
[
"Nolting",
"D.",
""
],
[
"Ochsner",
"E.",
""
],
[
"O'Dell",
"J.",
""
],
[
"Ogin",
"G. H.",
""
],
[
"Oldenburg",
"R. G.",
""
],
[
"O'Reilly",
"B.",
""
],
[
"O'Shaughnessy",
"R.",
""
],
[
"Osthelder",
"C.",
""
],
[
"Ottaway",
"D. J.",
""
],
[
"Ottens",
"R. S.",
""
],
[
"Overmier",
"H.",
""
],
[
"Owen",
"B. J.",
""
],
[
"Page",
"A.",
""
],
[
"Pagliaroli",
"G.",
""
],
[
"Palladino",
"L.",
""
],
[
"Palomba",
"C.",
""
],
[
"Pan",
"Y.",
""
],
[
"Pankow",
"C.",
""
],
[
"Paoletti",
"F.",
""
],
[
"Papa",
"M. A.",
""
],
[
"Pardi",
"S.",
""
],
[
"Pareja",
"M.",
""
],
[
"Parisi",
"M.",
""
],
[
"Pasqualetti",
"A.",
""
],
[
"Passaquieti",
"R.",
""
],
[
"Passuello",
"D.",
""
],
[
"Patel",
"P.",
""
],
[
"Pathak",
"D.",
""
],
[
"Pedraza",
"M.",
""
],
[
"Pekowsky",
"L.",
""
],
[
"Penn",
"S.",
""
],
[
"Peralta",
"C.",
""
],
[
"Perreca",
"A.",
""
],
[
"Persichetti",
"G.",
""
],
[
"Pichot",
"M.",
""
],
[
"Pickenpack",
"M.",
""
],
[
"Piergiovanni",
"F.",
""
],
[
"Pietka",
"M.",
""
],
[
"Pinard",
"L.",
""
],
[
"Pinto",
"I. M.",
""
],
[
"Pitkin",
"M.",
""
],
[
"Pletsch",
"H. J.",
""
],
[
"Plissi",
"M. V.",
""
],
[
"Poggiani",
"R.",
""
],
[
"Postiglione",
"F.",
""
],
[
"Prato",
"M.",
""
],
[
"Predoi",
"V.",
""
],
[
"Price",
"L. R.",
""
],
[
"Prijatelj",
"M.",
""
],
[
"Principe",
"M.",
""
],
[
"Prix",
"R.",
""
],
[
"Prodi",
"G. A.",
""
],
[
"Prokhorov",
"L.",
""
],
[
"Puncken",
"O.",
""
],
[
"Punturo",
"M.",
""
],
[
"Puppo",
"P.",
""
],
[
"Quetschke",
"V.",
""
],
[
"Raab",
"F. J.",
""
],
[
"Rabeling",
"D. S.",
""
],
[
"Radke",
"T.",
""
],
[
"Radkins",
"H.",
""
],
[
"Raffai",
"P.",
""
],
[
"Rakhmanov",
"M.",
""
],
[
"Rankins",
"B.",
""
],
[
"Rapagnani",
"P.",
""
],
[
"Raymond",
"V.",
""
],
[
"Re",
"V.",
""
],
[
"Reed",
"C. M.",
""
],
[
"Reed",
"T.",
""
],
[
"Regimbau",
"T.",
""
],
[
"Reid",
"S.",
""
],
[
"Reitze",
"D. H.",
""
],
[
"Ricci",
"F.",
""
],
[
"Riesen",
"R.",
""
],
[
"Riles",
"K.",
""
],
[
"Roberts",
"P.",
""
],
[
"Robertson",
"N. A.",
""
],
[
"Robinet",
"F.",
""
],
[
"Robinson",
"C.",
""
],
[
"Robinson",
"E. L.",
""
],
[
"Rocchi",
"A.",
""
],
[
"Roddy",
"S.",
""
],
[
"Röver",
"C.",
""
],
[
"Rolland",
"L.",
""
],
[
"Rollins",
"J.",
""
],
[
"Romano",
"J. D.",
""
],
[
"Romano",
"R.",
""
],
[
"Romie",
"J. H.",
""
],
[
"Rosińska",
"D.",
""
],
[
"Rowan",
"S.",
""
],
[
"Rüdiger",
"A.",
""
],
[
"Ruggi",
"P.",
""
],
[
"Ryan",
"K.",
""
],
[
"Sakata",
"S.",
""
],
[
"Sakosky",
"M.",
""
],
[
"Salemi",
"F.",
""
],
[
"Sammut",
"L.",
""
],
[
"de la Jordana",
"L. Sancho",
""
],
[
"Sandberg",
"V.",
""
],
[
"Sannibale",
"V.",
""
],
[
"Santamaría",
"L.",
""
],
[
"Santostasi",
"G.",
""
],
[
"Saraf",
"S.",
""
],
[
"Sassolas",
"B.",
""
],
[
"Sathyaprakash",
"B. S.",
""
],
[
"Sato",
"S.",
""
],
[
"Satterthwaite",
"M.",
""
],
[
"Saulson",
"P. R.",
""
],
[
"Savage",
"R.",
""
],
[
"Schilling",
"R.",
""
],
[
"Schnabel",
"R.",
""
],
[
"Schofield",
"R. M. S.",
""
],
[
"Schulz",
"B.",
""
],
[
"Schutz",
"B. F.",
""
],
[
"Schwinberg",
"P.",
""
],
[
"Scott",
"J.",
""
],
[
"Scott",
"S. M.",
""
],
[
"Searle",
"A. C.",
""
],
[
"Seifert",
"F.",
""
],
[
"Sellers",
"D.",
""
],
[
"Sengupta",
"A. S.",
""
],
[
"Sentenac",
"D.",
""
],
[
"Sergeev",
"A.",
""
],
[
"Shaddock",
"D. A.",
""
],
[
"Shapiro",
"B.",
""
],
[
"Shawhan",
"P.",
""
],
[
"Shoemaker",
"D. H.",
""
],
[
"Sibley",
"A.",
""
],
[
"Siemens",
"X.",
""
],
[
"Sigg",
"D.",
""
],
[
"Singer",
"A.",
""
],
[
"Sintes",
"A. M.",
""
],
[
"Skelton",
"G.",
""
],
[
"Slagmolen",
"B. J. J.",
""
],
[
"Slutsky",
"J.",
""
],
[
"Smith",
"J. R.",
""
],
[
"Smith",
"M. R.",
""
],
[
"Smith",
"N. D.",
""
],
[
"Somiya",
"K.",
""
],
[
"Sorazu",
"B.",
""
],
[
"Speirits",
"F. C.",
""
],
[
"Sperandio",
"L.",
""
],
[
"Stein",
"A. J.",
""
],
[
"Stein",
"L. C.",
""
],
[
"Steinlechner",
"S.",
""
],
[
"Steplewski",
"S.",
""
],
[
"Stochino",
"A.",
""
],
[
"Stone",
"R.",
""
],
[
"Strain",
"K. A.",
""
],
[
"Strigin",
"S.",
""
],
[
"Stroeer",
"A. S.",
""
],
[
"Sturani",
"R.",
""
],
[
"Stuver",
"A. L.",
""
],
[
"Summerscales",
"T. Z.",
""
],
[
"Sung",
"M.",
""
],
[
"Susmithan",
"S.",
""
],
[
"Sutton",
"P. J.",
""
],
[
"Swinkels",
"B.",
""
],
[
"Szokoly",
"G. P.",
""
],
[
"Talukder",
"D.",
""
],
[
"Tanner",
"D. B.",
""
],
[
"Tarabrin",
"S. P.",
""
],
[
"Taylor",
"J. R.",
""
],
[
"Taylor",
"R.",
""
],
[
"Thomas",
"P.",
""
],
[
"Thorne",
"K. A.",
""
],
[
"Thorne",
"K. S.",
""
],
[
"Thrane",
"E.",
""
],
[
"Thüring",
"A.",
""
],
[
"Titsler",
"C.",
""
],
[
"Tokmakov",
"K. V.",
""
],
[
"Toncelli",
"A.",
""
],
[
"Tonelli",
"M.",
""
],
[
"Torre",
"O.",
""
],
[
"Torres",
"C.",
""
],
[
"Torrie",
"C. I.",
""
],
[
"Tournefier",
"E.",
""
],
[
"Travasso",
"F.",
""
],
[
"Traylor",
"G.",
""
],
[
"Trias",
"M.",
""
],
[
"Trummer",
"J.",
""
],
[
"Tseng",
"K.",
""
],
[
"Turner",
"L.",
""
],
[
"Ugolini",
"D.",
""
],
[
"Urbanek",
"K.",
""
],
[
"Vahlbruch",
"H.",
""
],
[
"Vaishnav",
"B.",
""
],
[
"Vajente",
"G.",
""
],
[
"Vallisneri",
"M.",
""
],
[
"Brand",
"J. F. J. van den",
""
],
[
"Broeck",
"C. Van Den",
""
],
[
"van der Putten",
"S.",
""
],
[
"van der Sluys",
"M. V.",
""
],
[
"van Veggel",
"A. A.",
""
],
[
"Vass",
"S.",
""
],
[
"Vaulin",
"R.",
""
],
[
"Vavoulidis",
"M.",
""
],
[
"Vecchio",
"A.",
""
],
[
"Vedovato",
"G.",
""
],
[
"Veitch",
"J.",
""
],
[
"Veitch",
"P. J.",
""
],
[
"Veltkamp",
"C.",
""
],
[
"Verkindt",
"D.",
""
],
[
"Vetrano",
"F.",
""
],
[
"Viceré",
"A.",
""
],
[
"Villar",
"A. E.",
""
],
[
"Vinet",
"J. -Y.",
""
],
[
"Vocca",
"H.",
""
],
[
"Vorvick",
"C.",
""
],
[
"Vyachanin",
"S. P.",
""
],
[
"Waldman",
"S. J.",
""
],
[
"Wallace",
"L.",
""
],
[
"Wanner",
"A.",
""
],
[
"Ward",
"R. L.",
""
],
[
"Was",
"M.",
""
],
[
"Wei",
"P.",
""
],
[
"Weinert",
"M.",
""
],
[
"Weinstein",
"A. J.",
""
],
[
"Weiss",
"R.",
""
],
[
"Wen",
"L.",
""
],
[
"Wen",
"S.",
""
],
[
"Wessels",
"P.",
""
],
[
"West",
"M.",
""
],
[
"Westphal",
"T.",
""
],
[
"Wette",
"K.",
""
],
[
"Whelan",
"J. T.",
""
],
[
"Whitcomb",
"S. E.",
""
],
[
"White",
"D.",
""
],
[
"Whiting",
"B. F.",
""
],
[
"Wilkinson",
"C.",
""
],
[
"Willems",
"P. A.",
""
],
[
"Williams",
"L.",
""
],
[
"Willke",
"B.",
""
],
[
"Winkelmann",
"L.",
""
],
[
"Winkler",
"W.",
""
],
[
"Wipf",
"C. C.",
""
],
[
"Wiseman",
"A. G.",
""
],
[
"Woan",
"G.",
""
],
[
"Wooley",
"R.",
""
],
[
"Worden",
"J.",
""
],
[
"Yakushin",
"I.",
""
],
[
"Yamamoto",
"H.",
""
],
[
"Yamamoto",
"K.",
""
],
[
"Yeaton-Massey",
"D.",
""
],
[
"Yoshida",
"S.",
""
],
[
"Yu",
"P.",
""
],
[
"Yvert",
"M.",
""
],
[
"Zanolin",
"M.",
""
],
[
"Zhang",
"L.",
""
],
[
"Zhang",
"Z.",
""
],
[
"Zhao",
"C.",
""
],
[
"Zotov",
"N.",
""
],
[
"Zucker",
"M. E.",
""
],
[
"Zweizig",
"J.",
""
]
] |
1703.06706 | Lorenzo Carlucci | Lorenzo Carlucci | A note on Hindman-type theorems for uncountable cardinals | null | null | null | Roma01.Math | math.CO math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent results of Hindman, Leader and Strauss and of Fern\'andez-Bret\'on and Rinot showed that natural versions of Hindman's Theorem fail {\em for all} uncontable cardinals. On the other hand, Komj\'ath proved a result in the positive direction, showing that {\em there are} arbitrarily large abelian groups satisfying {\em some} Hindman-type property. In this note we show how a family of natural Hindman-type theorems for uncountable cardinals can be obtained by adapting some recent results of the author from their original countable setting. We also show how lower bounds for {\em some} of the variants considered can be obtained. | [
{
"version": "v1",
"created": "Mon, 20 Mar 2017 12:24:24 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Carlucci",
"Lorenzo",
""
]
] |
1802.09805 | Eric Marberg | Zachary Hamaker and Eric Marberg | Atoms for signed permutations | 40 pages, 4 figures, 1 table; v2: several minor corrections; v3: updated references, added index of symbols, other minor revisions; v4: major rewrite, expanded results; v5: corrects a mistake in Proposition 5.6 and fixes two related formulas in Theorems 8.2 and 8.7 | European J. Combin. 94 (2021), 103288 | 10.1016/j.ejc.2020.103288 | null | math.CO math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is a natural analogue of weak Bruhat order on the involutions in any Coxeter group. The saturated chains of intervals in this order correspond to reduced words for a certain set of group elements called atoms. Brion gives a general formula for the cohomology class of a $K$-orbit closure in an arbitrary flag variety, where $K$ is a symmetric subgroup of a complex algebraic group. In type A, the terms in this formula are indexed by atoms for permutations. We study the combinatorics of atoms for involutions in the group of signed permutations. In particular, we give a compact description of the atom set for any signed involution and endow it with the structure of a graded poset. Our main result, as an application, is to identify explicitly the terms in Brion's cohomology formula in types B and C. These descriptions apply to all $K$-orbits in these types and are the first of their kind outside of type A. | [
{
"version": "v1",
"created": "Tue, 27 Feb 2018 10:18:07 GMT"
},
{
"version": "v2",
"created": "Sat, 24 Mar 2018 10:53:30 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jul 2018 08:46:03 GMT"
},
{
"version": "v4",
"created": "Thu, 04 Jun 2020 16:17:49 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 03:34:10 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Hamaker",
"Zachary",
""
],
[
"Marberg",
"Eric",
""
]
] |
1812.09340 | Shaoyang Jia | Shaoyang Jia, James P. Vary | Parton distribution functions from scalar light-front parton gas model | null | null | null | null | nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an application of a microcanonical ensemble with light-front kinematics to model the phase-space distribution of relativistic constituents of a bound state. These constituents denoted by partons are treated as classical spin-zero particles confined inside the bound state with inter-parton collisions as their only interaction. The microcanonical molecular dynamics ensemble is applied to obtain the phase-space distribution of such a thermodynamic system. We sample this phase-space distribution using Monte Carlo algorithms to obtain the parton distribution functions (PDFs) in scenarios with $3$, $4$, and $5$ identical partons. In addition PDFs when a selected number of massless partons are mixed with $3$ massive partons are also presented. | [
{
"version": "v1",
"created": "Fri, 21 Dec 2018 19:08:07 GMT"
},
{
"version": "v2",
"created": "Sun, 28 Jul 2024 22:43:57 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 16:22:05 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Jia",
"Shaoyang",
""
],
[
"Vary",
"James P.",
""
]
] |
1904.00778 | Wenhui Chen | Halit Sevki Aslan and Wenhui Chen | On the Cauchy problem for semilinear thermoelastic plate systems in the $L^q$ framework | It updates the new version with general L^q estimates and nonlinearities | null | 10.3934/eect.2025044 | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We mainly consider semilinear thermoelastic plate systems with general power nonlinearities in the whole space $\mathbb{R}^n$. By applying the Fourier analysis, some sharp $(L^q\cap L^m)-L^q$ estimates of solutions (with any $1\leqslant m\leqslant q\leqslant +\infty$) to the classical thermoelastic plate system are derived, which cover all known results in $\mathbb{R}^n$. Then, we investigate global in time existence of small data $L^q$ solutions (with any $q\in[1,+\infty]$) and blow-up of weak solutions for the semilinear thermoelastic plate systems under suitable conditions on the power exponents, which justify critical exponents for several classes of nonlinearities. | [
{
"version": "v1",
"created": "Mon, 01 Apr 2019 12:41:22 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Jul 2019 06:24:29 GMT"
},
{
"version": "v3",
"created": "Sun, 29 Sep 2024 09:20:57 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Aslan",
"Halit Sevki",
""
],
[
"Chen",
"Wenhui",
""
]
] |
1905.10527 | Ali Zafari | Ali Zafari | Some algebraic properties of the extended graph of the bipartite double of Odd graphs | arXiv admin note: substantial text overlap with arXiv:1905.10525 | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In general, if a graph cannot be a distance regular, then to compute its spectrum and analyze its structure by using the relationship between equitable partition and orbit partition of graphs can be complicated. The aim of this paper, is to obtain all distinct eigenvalues of the extended graph $E(2.O_k)$ of the bipartite double graph $2.O_k$ of Odd graph $O_k$, by using the relationship between equitable partition and orbit partition of graphs, although the study of integrality of this graph goes back much further, see ~\cite{paper6a}. In fact, by another approach we show how we can find, by using this method, the set of all distinct eigenvalues of a class of graphs so that cannot be distance regular. First, we show that $E(2.O_k)$ is a vertex transitive graph with the diameter $k$, whereas the diameter of $2.O_k$ is $2k-1$, although $E(2.O_k)$ cannot be a distance regular graph. Also, we determine the automorphism group of $E(2.O_k)$. Moreover, we show that $E(2.O_k)$ is an integral graph, that is, each of its eigenvalues of the adjacency matrix of $E(2.O_k)$ is an integer. Especially, we determine the multiplicity of all distinct eigenvalues of the extended graph $E(2.O_k)$. | [
{
"version": "v1",
"created": "Sat, 25 May 2019 05:56:33 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Sep 2019 12:06:27 GMT"
},
{
"version": "v3",
"created": "Sun, 08 Sep 2019 00:00:45 GMT"
},
{
"version": "v4",
"created": "Thu, 12 Sep 2019 10:10:09 GMT"
},
{
"version": "v5",
"created": "Thu, 06 Feb 2020 08:57:28 GMT"
},
{
"version": "v6",
"created": "Thu, 05 Jun 2025 14:21:58 GMT"
},
{
"version": "v7",
"created": "Wed, 11 Jun 2025 03:28:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zafari",
"Ali",
""
]
] |
1905.11027 | Ke Sun | Ke Sun and Frank Nielsen | A Geometric Modeling of Occam's Razor in Deep Learning | This work first appeared under the former title "Lightlike Neuromanifolds, Occam's Razor and Deep Learning" in 2019 and has been published in Information Geometry in 2025 | Information Geometry, Special Issue: Half a Century of Information Geometry, Part 2, 2025 | 10.1007/s41884-025-00167-2 | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Why do deep neural networks (DNNs) benefit from very high dimensional parameter spaces? Their huge parameter complexities vs stunning performance in practice is all the more intriguing and not explainable using the standard theory of model selection for regular models. In this work, we propose a geometrically flavored information-theoretic approach to study this phenomenon. With the belief that simplicity is linked to better generalization, as grounded in the theory of minimum description length, the objective of our analysis is to examine and bound the complexity of DNNs. We introduce the locally varying dimensionality of the parameter space of neural network models by considering the number of significant dimensions of the Fisher information matrix, and model the parameter space as a manifold using the framework of singular semi-Riemannian geometry. We derive model complexity measures which yield short description lengths for deep neural network models based on their singularity analysis thus explaining the good performance of DNNs despite their large number of parameters. | [
{
"version": "v1",
"created": "Mon, 27 May 2019 07:57:26 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Feb 2020 05:53:28 GMT"
},
{
"version": "v3",
"created": "Wed, 31 Mar 2021 03:45:04 GMT"
},
{
"version": "v4",
"created": "Thu, 23 Dec 2021 12:01:37 GMT"
},
{
"version": "v5",
"created": "Thu, 28 Dec 2023 08:57:54 GMT"
},
{
"version": "v6",
"created": "Thu, 06 Jun 2024 22:26:17 GMT"
},
{
"version": "v7",
"created": "Thu, 31 Oct 2024 23:09:08 GMT"
},
{
"version": "v8",
"created": "Wed, 26 Mar 2025 05:29:49 GMT"
},
{
"version": "v9",
"created": "Wed, 11 Jun 2025 00:31:46 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Sun",
"Ke",
""
],
[
"Nielsen",
"Frank",
""
]
] |
1909.03820 | Steffen Van Bergerem | Steffen van Bergerem | Learning Concepts Definable in First-Order Logic with Counting | null | 34th Annual ACM/IEEE Symposium on Logic in Computer Science, LICS 2019, Vancouver, BC, Canada, June 24-27, 2019 | 10.1109/LICS.2019.8785811 | null | cs.LO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | We study Boolean classification problems over relational background structures in the logical framework introduced by Grohe and Tur\'an (TOCS 2004). It is known (Grohe and Ritzert, LICS 2017) that classifiers definable in first-order logic over structures of polylogarithmic degree can be learned in sublinear time, where the degree of the structure and the running time are measured in terms of the size of the structure. We generalise the results to the first-order logic with counting FOCN, which was introduced by Kuske and Schweikardt (LICS 2017) as an expressive logic generalising various other counting logics. Specifically, we prove that classifiers definable in FOCN over classes of structures of polylogarithmic degree can be consistently learned in sublinear time. This can be seen as a first step towards extending the learning framework to include numerical aspects of machine learning. We extend the result to agnostic probably approximately correct (PAC) learning for classes of structures of degree at most $(\log \log n)^c$ for some constant $c$. Moreover, we show that bounding the degree is crucial to obtain sublinear-time learning algorithms. That is, we prove that, for structures of unbounded degree, learning is not possible in sublinear time, even for classifiers definable in plain first-order logic. | [
{
"version": "v1",
"created": "Mon, 09 Sep 2019 12:57:29 GMT"
},
{
"version": "v2",
"created": "Sat, 23 Mar 2024 22:14:50 GMT"
},
{
"version": "v3",
"created": "Wed, 29 Jan 2025 18:49:49 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 14:23:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"van Bergerem",
"Steffen",
""
]
] |
2007.02527 | Thomas Ringstrom | Thomas J. Ringstrom, Mohammadhosein Hasanbeig, Alessandro Abate | Goal Kernel Planning: Linearly-Solvable Non-Markovian Policies for Logical Tasks with Goal-Conditioned Options | 52 Pages total. This is an update to a paper we submitted to a Journal and received reviewer feedback for improvement | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | In the domain of hierarchical planning, compositionality, abstraction, and task transfer are crucial for designing algorithms that can efficiently solve a variety of problems with maximal representational reuse. Many real-world problems require non-Markovian policies to handle complex structured tasks with logical conditions, often leading to prohibitively large state representations; this requires efficient methods for breaking these problems down and reusing structure between tasks. To this end, we introduce a compositional framework called Linearly-Solvable Goal Kernel Dynamic Programming (LS-GKDP) to address the complexity of solving non-Markovian Boolean sub-goal tasks with ordering constraints. LS-GKDP combines the Linearly-Solvable Markov Decision Process (LMDP) formalism with the Options Framework of Reinforcement Learning. LMDPs can be efficiently solved as a principal eigenvector problem, and options are policies with termination conditions used as temporally extended actions; with LS-GKDP we expand LMDPs to control over options for logical tasks. This involves decomposing a high-dimensional problem down into a set of goal-condition options for each goal and constructing a goal kernel, which is an abstract transition kernel that jumps from an option's initial-states to its termination-states along with an update of the higher-level task-state. We show how an LMDP with a goal kernel enables the efficient optimization of meta-policies in a lower-dimensional subspace defined by the task grounding. Options can also be remapped to new problems within a super-exponential space of tasks without significant recomputation, and we identify cases where the solution is invariant to the task grounding, permitting zero-shot task transfer. | [
{
"version": "v1",
"created": "Mon, 06 Jul 2020 05:13:20 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:43:11 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ringstrom",
"Thomas J.",
""
],
[
"Hasanbeig",
"Mohammadhosein",
""
],
[
"Abate",
"Alessandro",
""
]
] |
2007.04225 | Alexei Bazavov | Alexei Bazavov | Commutator-free Lie group methods with minimum storage requirements and reuse of exponentials | 41 pages, 10 figures. Minor revisions to bring the preprint in line with the final published version. The published version is slightly shortened in the introductory part | Bit Numer Math (2021) | 10.1007/s10543-021-00892-x | null | math.NA cs.NA hep-lat physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new format for commutator-free Lie group methods is proposed based on explicit classical Runge-Kutta schemes. In this format exponentials are reused at every stage and the storage is required only for two quantities: the right hand side of the differential equation evaluated at a given Runge-Kutta stage and the function value updated at the same stage. The next stage of the scheme is able to overwrite these values. The result is proven for a 3-stage third order method and a conjecture for higher order methods is formulated. Five numerical examples are provided in support of the conjecture. This new class of structure-preserving integrators has a wide variety of applications for numerically solving differential equations on manifolds. | [
{
"version": "v1",
"created": "Wed, 08 Jul 2020 16:06:33 GMT"
},
{
"version": "v2",
"created": "Sun, 26 Jul 2020 03:46:16 GMT"
},
{
"version": "v3",
"created": "Wed, 08 Sep 2021 22:16:24 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Bazavov",
"Alexei",
""
]
] |
2008.05821 | David Fern\'andez Bret\'on | Lorenzo Carlucci and David J. Fern\'andez-Bret\'on | The adjacent Hindman's theorem for uncountable groups | Final corrections so that this is as close as possible to the final published version | Colloquium Mathematicum 173 no. 2 (2023), 273-284 | 10.4064/cm9095-4-2023 | Roma01.Math | math.CO math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent results of Hindman, Leader and Strauss and of the second author and Rinot showed that some natural analogs of Hindman's Theorem fail for all uncountable cardinals. Results in the positive direction were obtained by Komj\'ath, the first author, and the second author and Lee, who showed that there are arbitrarily large Abelian groups satisfying some Hindman-type property. Inspired by an analogous result studied by the first author in the countable setting, we prove a new variant of Hindman's Theorem for uncountable cardinals, called the Adjacent Hindman's Theorem: For every $\kappa$ there is a $\lambda$ such that, whenever a group $G$ of cardinality $\lambda$ is coloured with $\kappa$ colours, there exists a $\lambda$-sized injective sequence of elements of $G$ with all finite products of adjacent terms of the sequence of the same colour. We obtain bounds on $\lambda$ as a function of $\kappa$, and prove that such bounds are optimal. This is the first example of a Hindman-type result for uncountable cardinals that we can prove also in the non-Abelian setting and, furthermore, it is the first such example where monochromatic products (or sums) of unbounded length are guaranteed. | [
{
"version": "v1",
"created": "Thu, 13 Aug 2020 11:18:17 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Feb 2023 05:28:43 GMT"
},
{
"version": "v3",
"created": "Mon, 03 Apr 2023 14:09:08 GMT"
},
{
"version": "v4",
"created": "Wed, 12 Apr 2023 15:12:40 GMT"
},
{
"version": "v5",
"created": "Tue, 11 Jul 2023 18:16:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Carlucci",
"Lorenzo",
""
],
[
"Fernández-Bretón",
"David J.",
""
]
] |
2008.13161 | Alexandru Oancea | Kai Cieliebak, Nancy Hingston, and Alexandru Oancea | Poincar\'e duality for loop spaces | 86 pages, 18 figures. The main result is now phrased as an isomorphism of graded topological Frobenius algebras in the setup of Tate vector spaces. This is the final version of the paper, to be published in Compositio Mathematica | null | null | null | math.SG math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that Rabinowitz Floer homology and cohomology carry the structure of a graded Frobenius algebra for both closed and open strings. We prove a Poincar\'e duality theorem between homology and cohomology that preserves this structure. This lifts to a duality theorem between graded open-closed TQFTs. We use in a systematic way the formalism of Tate vector spaces.
Specializing to the case of cotangent bundles, we define Rabinowitz loop homology and cohomology and explain from a unified perspective pairs of dual results that have been observed over the years in the context of the search for closed geodesics. These concern critical levels, relations to the based loop space, manifolds all of whose geodesics are closed, Bott index iteration, and level-potency. Moreover, the graded Frobenius algebra structure gives meaning and proof to a relation conjectured by Sullivan between the loop product and coproduct. | [
{
"version": "v1",
"created": "Sun, 30 Aug 2020 13:13:13 GMT"
},
{
"version": "v2",
"created": "Wed, 07 Sep 2022 16:27:29 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 15:16:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cieliebak",
"Kai",
""
],
[
"Hingston",
"Nancy",
""
],
[
"Oancea",
"Alexandru",
""
]
] |
2010.00788 | Risto Miikkulainen | Santiago Gonzalez, Xin Qiu, and Risto Miikkulainen | Effective Regularization Through Loss-Function Metalearning | A shorter version of this paper appeared in CEC 2025; this paper includes appendices, expanded references, and corrections | Congress on Evolutionary Computation (CEC), 2025 | null | null | cs.LG cs.NE stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evolutionary computation can be used to optimize several different aspects of neural network architectures. For instance, the TaylorGLO method discovers novel, customized loss functions, resulting in improved performance, faster training, and improved data utilization. A likely reason is that such functions discourage overfitting, leading to effective regularization. This paper demonstrates theoretically that this is indeed the case for TaylorGLO. Learning rule decomposition reveals that evolved loss functions balance two factors: the pull toward zero error, and a push away from it to avoid overfitting. This is a general principle that may be used to understand other regularization techniques as well (as demonstrated in this paper for label smoothing). The theoretical analysis leads to a constraint that can be utilized to find more effective loss functions in practice; the mechanism also results in networks that are more robust (as demonstrated in this paper with adversarial inputs). The analysis in this paper thus constitutes a first step towards understanding regularization, and demonstrates the power of evolutionary neural architecture search in general. | [
{
"version": "v1",
"created": "Fri, 02 Oct 2020 05:22:21 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Oct 2021 04:47:05 GMT"
},
{
"version": "v3",
"created": "Sat, 10 May 2025 23:50:24 GMT"
},
{
"version": "v4",
"created": "Sun, 08 Jun 2025 05:04:52 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 03:52:09 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gonzalez",
"Santiago",
""
],
[
"Qiu",
"Xin",
""
],
[
"Miikkulainen",
"Risto",
""
]
] |
2011.07757 | Wenhui Chen | Wenhui Chen and Yan Liu | Large time behavior for the hyperbolic-parabolic coupled system with the regularity-loss structure | The results have been improved | null | 10.1007/s00028-025-01091-9 | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper considers the hyperbolic-parabolic coupled system, arising from the generalized thermoelastic coupled system, in the whole space $\mathbb{R}^n$. We study some qualitative properties for an energy term by diagonalization procedures, and for the solution by the WKB analysis. Particularly, we derive new large time asymptotic profiles with the regularity-loss structure (from the biharmonic parabolic equation and the diffusion wave equation with the Riesz potential operator) and optimal decay estimates with suitable higher regularities for the Cauchy data. Finally, we discover that the wave equation with the Riesz potential dissipation is a large time approximated model of our hyperbolic-parabolic coupled system. | [
{
"version": "v1",
"created": "Mon, 16 Nov 2020 07:25:54 GMT"
},
{
"version": "v2",
"created": "Mon, 29 Apr 2024 07:45:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Wenhui",
""
],
[
"Liu",
"Yan",
""
]
] |
2012.13954 | Stephan Rosswog | S. Rosswog and P. Diener | SPHINCS_BSSN: A general relativistic Smooth Particle Hydrodynamics code for dynamical spacetimes | 41 pages, 16 figures; accepted for publication in Classical and Quantum Gravity | Chap. 7 in Book "New Frontiers in GRMHD Simulations", Springer Singapore, Eds.Cosimo Bambi, Yosuke Mizuno, Swarnim Shashank, Feng Yuan (2025) | 10.1007/978-981-97-8522-3 | null | gr-qc astro-ph.HE astro-ph.IM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a new methodology for simulating self-gravitating general-relativistic fluids. In our approach the fluid is modelled by means of Lagrangian particles in the framework of a general-relativistic (GR) Smooth Particle Hydrodynamics (SPH) formulation, while the spacetime is evolved on a mesh according to the BSSN formulation that is also frequently used in Eulerian GR-hydrodynamics. To the best of our knowledge this is the first Lagrangian fully general relativistic hydrodynamics code (all previous SPH approaches used approximations to GR-gravity). A core ingredient of our particle-mesh approach is the coupling between the gas (represented by particles) and the spacetime (represented by a mesh) for which we have developed a set of sophisticated interpolation tools that are inspired by other particle-mesh approaches, in particular by vortex-particle methods. One advantage of splitting the methodology between matter and spacetime is that it gives us more freedom in choosing the resolution, so that -- if the spacetime is smooth enough -- we obtain good results already with a moderate number of grid cells and can focus the computational effort on the simulation of the matter. Further advantages of our approach are the ease with which ejecta can be tracked and the fact that the neutron star surface remains well-behaved and does not need any particular treatment. In the hydrodynamics part of the code we use a number of techniques that are new to SPH, such as reconstruction, slope limiting and steering dissipation by monitoring entropy conservation. We describe here in detail the employed numerical methods and demonstrate the code performance in a number of benchmark problems ranging from shock tube tests, over Cowling approximations to the fully dynamical evolution of neutron stars in self-consistently evolved spacetimes. | [
{
"version": "v1",
"created": "Sun, 27 Dec 2020 14:36:45 GMT"
},
{
"version": "v2",
"created": "Thu, 11 Mar 2021 16:39:04 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Rosswog",
"S.",
""
],
[
"Diener",
"P.",
""
]
] |
2102.07894 | Darij Grinberg | Darij Grinberg, Lukas Katth\"an, Joel Brewster Lewis | The path-missing and path-free complexes of a directed graph | 49 pages (37 main, 10 appendix, 2 bib). New version has a few typos fixed, some clearer proofs and a new Figure 1. Comments welcome! | null | null | null | math.CO math.AT | http://creativecommons.org/licenses/by/4.0/ | We study two simplicial complexes arising from a directed graph $G = (V, E)$ with two chosen vertices $s$ and $t$: the *path-free complex*, consisting of all subsets $F \subseteq E$ that contain no path from $s$ to $t$, and the *path-missing complex*, its Alexander dual. Using discrete Morse theory, we prove that both complexes have well-behaved homotopy types -- either contractible or homotopy-equivalent to spheres. | [
{
"version": "v1",
"created": "Mon, 15 Feb 2021 23:34:54 GMT"
},
{
"version": "v2",
"created": "Tue, 05 Mar 2024 22:47:58 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 21:36:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Grinberg",
"Darij",
""
],
[
"Katthän",
"Lukas",
""
],
[
"Lewis",
"Joel Brewster",
""
]
] |
2110.03386 | Otari Sakhelashvili | Otari Sakhelashvili | On dual formulation of Axion solution to strong-CP problem | null | Phys.Rev.D 105 (2022) 8, 085020 | 10.1103/PhysRevD.105.085020 | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | An exact duality between axion with arbitrary potential and antisymmetric form-field has been derived some time ago. Using this duality, the axion solution to strong-CP problem has been formulated as a gauge invariant theory of forms. In this description, the QCD axion is represented by a Kalb-Ramond field which is eaten-up by the Chern-Simons $3$-form of QCD, thereby making it massive. This ensures the CP-invariance of the vacuum. Although viewed as an effective low energy theory, this formulation accomplishes the same goal as ordinary Peccei-Quinn mechanism, due to its gauge invariance, it is protected against unwanted UV-corrections. In the previous work it has been shown that dual formulation is insensitive to UV-physics in the sense that the corrections to CP-conserving vacuum from arbitrary massive sources are strictly zero. By going carefully through duality transformations and source-resolution, we reproduce this curious result and give some further consistency checks. We apply similar analysis to other approaches to naturalness problems based on form fields and axions, such as cosmological relaxation of the standard model Higgs boson mass via the attractor mechanism. | [
{
"version": "v1",
"created": "Thu, 07 Oct 2021 12:25:56 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Sakhelashvili",
"Otari",
""
]
] |
2111.03620 | Otari Sakhelashvili | Gia Dvali, Otari Sakhelashvili | Black-Hole-Like Saturons in Gross-Neveu | null | Phys.Rev.D 105 (2022) 6, 065014 | 10.1103/PhysRevD.105.065014 | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | It has been argued recently that objects of maximal microstate entropy permitted by unitarity, the so-called saturons, have properties similar to black holes. We demonstrate the existence of such objects in Gross-Neveu model. From the large-$N$ scaling of $S$-matrix, we deduce the connection between the entropy of the bound-state and the unitarity of scattering. We observe that upon saturation of unitarity, the bound state exhibits a remarkable correspondence with a black hole. The scaling of its entropy is identical to Bekenstein-Hawking entropy. The saturon decays via Hawking's thermal rate of temperature given by the inverse size. The information retrieval time from the Gross-Neveu saturon is isomorphic to Page's time. Our observations indicate that black hole properties are exhibited by saturated states in simple calculable models. | [
{
"version": "v1",
"created": "Fri, 05 Nov 2021 17:13:33 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Dvali",
"Gia",
""
],
[
"Sakhelashvili",
"Otari",
""
]
] |
2111.07590 | Joseph Helfer | Fran\c{c}ois-Simon Fauteux-Chapleau and Joseph Helfer | Exotic tight contact structures on $\mathbb{R}^n$ | Updated to published version | Journal of Symplectic Geometry, Vol. 23, Issue 1 (2025), pp. 227-283 | 10.4310/JSG.250403022531 | null | math.SG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a variant of contact homology for convex open contact manifolds. As an application, we prove the existence of (in fact, infinitely many) exotic tight contact structures on $\mathbb{R}^{2n-1}$ for all $n>2$. | [
{
"version": "v1",
"created": "Mon, 15 Nov 2021 08:11:15 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Mar 2022 17:26:27 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Feb 2023 01:54:55 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 19:29:57 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Fauteux-Chapleau",
"François-Simon",
""
],
[
"Helfer",
"Joseph",
""
]
] |
2111.12022 | Otari Sakhelashvili | Lasha Berezhiani, Gia Dvali, Otari Sakhelashvili | De Sitter space as BRST invariant coherent state of gravitons | 12 pages; v2: minor corrections, refs. added | Phys.Rev.D 105 (2022) 2, 025022 | 10.1103/PhysRevD.105.025022 | null | hep-th gr-qc | http://creativecommons.org/licenses/by/4.0/ | The $S$-matrix formulation indicates that a consistent embedding of de Sitter state in quantum gravity is possible exclusively as an excited quantum state constructed on top of a valid $S$-matrix vacuum such as Minkowski. In the present paper we offer such a construction of de Sitter in the form of a coherent state of gravitons. Unlike previous realizations of this idea, we focus on BRST invariance as the guiding principle for physicality. In order to establish the universal rules of gauge consistency, we study the BRST-invariant construction of coherent states created by classical and quantum sources in QED and in linearized gravity. Introduction of $N$ copies of scalar matter coupled to gravity allows us to take a special double scaling limit, a so-called species limit, in which our construction of de Sitter becomes exact. In this limit, the irrelevant quantum gravitational effects vanish whereas the collective phenomena, such as Gibbons-Hawking radiation, are calculable. | [
{
"version": "v1",
"created": "Tue, 23 Nov 2021 17:31:28 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Jan 2022 12:41:26 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Berezhiani",
"Lasha",
""
],
[
"Dvali",
"Gia",
""
],
[
"Sakhelashvili",
"Otari",
""
]
] |
2112.05521 | Walid Oukil | Walid Oukil | Non trivial zeros of the Riemann zeta function using the differential equations | The numerical results obtained are consistent with the statement of the theorem | null | null | null | math.GM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider the representation of the Riemann zeta function $\zeta$ defined by Abel's summation formula. Using the differential equations, we show that: if $|\zeta(s)|=0$ then $|\zeta(1-s)|\neq0$ for any point $s$ in the critical strip except the critical line. | [
{
"version": "v1",
"created": "Tue, 07 Dec 2021 09:55:07 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Jan 2022 18:22:44 GMT"
},
{
"version": "v3",
"created": "Thu, 20 Jan 2022 11:09:53 GMT"
},
{
"version": "v4",
"created": "Wed, 09 Feb 2022 12:52:22 GMT"
},
{
"version": "v5",
"created": "Sun, 20 Feb 2022 21:49:42 GMT"
},
{
"version": "v6",
"created": "Wed, 23 Feb 2022 18:17:51 GMT"
},
{
"version": "v7",
"created": "Thu, 14 Apr 2022 14:33:14 GMT"
},
{
"version": "v8",
"created": "Tue, 19 Apr 2022 15:11:03 GMT"
},
{
"version": "v9",
"created": "Thu, 28 Apr 2022 14:38:32 GMT"
},
{
"version": "v10",
"created": "Mon, 02 May 2022 13:50:39 GMT"
},
{
"version": "v11",
"created": "Tue, 10 May 2022 11:46:39 GMT"
},
{
"version": "v12",
"created": "Mon, 11 Jul 2022 10:01:27 GMT"
},
{
"version": "v13",
"created": "Tue, 26 Jul 2022 11:11:13 GMT"
},
{
"version": "v14",
"created": "Tue, 02 Aug 2022 14:53:38 GMT"
},
{
"version": "v15",
"created": "Tue, 16 Aug 2022 12:26:42 GMT"
},
{
"version": "v16",
"created": "Tue, 04 Oct 2022 12:11:29 GMT"
},
{
"version": "v17",
"created": "Tue, 08 Nov 2022 18:14:30 GMT"
},
{
"version": "v18",
"created": "Mon, 14 Nov 2022 18:20:14 GMT"
},
{
"version": "v19",
"created": "Tue, 06 Dec 2022 17:37:25 GMT"
},
{
"version": "v20",
"created": "Mon, 12 Dec 2022 20:01:01 GMT"
},
{
"version": "v21",
"created": "Tue, 03 Jan 2023 20:50:57 GMT"
},
{
"version": "v22",
"created": "Tue, 10 Jan 2023 21:19:48 GMT"
},
{
"version": "v23",
"created": "Thu, 02 Feb 2023 09:40:11 GMT"
},
{
"version": "v24",
"created": "Sat, 18 Feb 2023 15:33:52 GMT"
},
{
"version": "v25",
"created": "Wed, 08 Mar 2023 10:05:46 GMT"
},
{
"version": "v26",
"created": "Wed, 15 Mar 2023 16:40:27 GMT"
},
{
"version": "v27",
"created": "Mon, 20 Mar 2023 14:43:51 GMT"
},
{
"version": "v28",
"created": "Tue, 04 Apr 2023 23:06:26 GMT"
},
{
"version": "v29",
"created": "Thu, 13 Apr 2023 09:54:50 GMT"
},
{
"version": "v30",
"created": "Wed, 09 Aug 2023 16:03:01 GMT"
},
{
"version": "v31",
"created": "Tue, 15 Aug 2023 11:47:13 GMT"
},
{
"version": "v32",
"created": "Mon, 21 Aug 2023 08:26:16 GMT"
},
{
"version": "v33",
"created": "Fri, 10 Nov 2023 09:41:36 GMT"
},
{
"version": "v34",
"created": "Fri, 24 Nov 2023 17:01:39 GMT"
},
{
"version": "v35",
"created": "Wed, 06 Dec 2023 12:12:29 GMT"
},
{
"version": "v36",
"created": "Mon, 18 Dec 2023 10:07:16 GMT"
},
{
"version": "v37",
"created": "Mon, 01 Jan 2024 13:20:54 GMT"
},
{
"version": "v38",
"created": "Thu, 11 Jan 2024 14:04:55 GMT"
},
{
"version": "v39",
"created": "Thu, 18 Jan 2024 10:05:52 GMT"
},
{
"version": "v40",
"created": "Sun, 28 Jan 2024 19:24:45 GMT"
},
{
"version": "v41",
"created": "Sun, 04 Feb 2024 14:58:28 GMT"
},
{
"version": "v42",
"created": "Fri, 16 Feb 2024 07:29:52 GMT"
},
{
"version": "v43",
"created": "Fri, 23 Feb 2024 10:15:10 GMT"
},
{
"version": "v44",
"created": "Fri, 01 Mar 2024 14:00:17 GMT"
},
{
"version": "v45",
"created": "Wed, 13 Mar 2024 11:00:19 GMT"
},
{
"version": "v46",
"created": "Mon, 18 Mar 2024 11:42:51 GMT"
},
{
"version": "v47",
"created": "Thu, 04 Apr 2024 08:34:23 GMT"
},
{
"version": "v48",
"created": "Mon, 08 Apr 2024 11:19:22 GMT"
},
{
"version": "v49",
"created": "Wed, 08 May 2024 20:31:20 GMT"
},
{
"version": "v50",
"created": "Mon, 20 May 2024 12:08:58 GMT"
},
{
"version": "v51",
"created": "Wed, 29 May 2024 12:27:29 GMT"
},
{
"version": "v52",
"created": "Mon, 10 Jun 2024 21:04:37 GMT"
},
{
"version": "v53",
"created": "Thu, 20 Jun 2024 15:59:55 GMT"
},
{
"version": "v54",
"created": "Wed, 26 Jun 2024 23:51:33 GMT"
},
{
"version": "v55",
"created": "Mon, 08 Jul 2024 09:01:49 GMT"
},
{
"version": "v56",
"created": "Mon, 15 Jul 2024 13:46:33 GMT"
},
{
"version": "v57",
"created": "Sat, 27 Jul 2024 17:34:07 GMT"
},
{
"version": "v58",
"created": "Sun, 04 Aug 2024 10:03:17 GMT"
},
{
"version": "v59",
"created": "Thu, 15 Aug 2024 18:30:10 GMT"
},
{
"version": "v60",
"created": "Tue, 17 Sep 2024 21:27:16 GMT"
},
{
"version": "v61",
"created": "Mon, 14 Oct 2024 10:37:04 GMT"
},
{
"version": "v62",
"created": "Sun, 20 Oct 2024 14:41:30 GMT"
},
{
"version": "v63",
"created": "Thu, 05 Dec 2024 08:50:04 GMT"
},
{
"version": "v64",
"created": "Wed, 11 Dec 2024 11:58:16 GMT"
},
{
"version": "v65",
"created": "Thu, 19 Dec 2024 12:20:25 GMT"
},
{
"version": "v66",
"created": "Sun, 16 Feb 2025 18:40:44 GMT"
},
{
"version": "v67",
"created": "Wed, 26 Feb 2025 11:56:20 GMT"
},
{
"version": "v68",
"created": "Fri, 14 Mar 2025 14:57:04 GMT"
},
{
"version": "v69",
"created": "Mon, 31 Mar 2025 12:06:15 GMT"
},
{
"version": "v70",
"created": "Wed, 11 Jun 2025 08:54:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Oukil",
"Walid",
""
]
] |
2112.13738 | Juliette Legrand | Juliette Legrand, Philippe Naveau, Marco Oesting | Evaluation of binary classifiers for asymptotically dependent and independent extremes | null | null | null | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning classification methods usually assume that all possible classes are sufficiently present within the training set. Due to their inherent rarities, extreme events are always under-represented and classifiers tailored for predicting extremes need to be carefully designed to handle this under-representation. In this paper, we address the question of how to assess and compare classifiers with respect to their capacity to capture extreme occurrences. This is also related to the topic of scoring rules used in forecasting literature. In this context, we propose and study a risk function adapted to extremal classifiers. The inferential properties of our empirical risk estimator are derived under the framework of multivariate regular variation and hidden regular variation. A simulation study compares different classifiers and indicates their performance with respect to our risk function. To conclude, we apply our framework to the analysis of extreme river discharges in the Danube river basin. The application compares different predictive algorithms and test their capacity at forecasting river discharges from other river stations. | [
{
"version": "v1",
"created": "Mon, 27 Dec 2021 15:40:25 GMT"
},
{
"version": "v2",
"created": "Tue, 13 Sep 2022 08:59:43 GMT"
},
{
"version": "v3",
"created": "Fri, 08 Mar 2024 16:02:24 GMT"
},
{
"version": "v4",
"created": "Wed, 05 Mar 2025 10:29:46 GMT"
},
{
"version": "v5",
"created": "Mon, 02 Jun 2025 12:26:55 GMT"
},
{
"version": "v6",
"created": "Wed, 11 Jun 2025 06:42:33 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Legrand",
"Juliette",
""
],
[
"Naveau",
"Philippe",
""
],
[
"Oesting",
"Marco",
""
]
] |
2201.12756 | Haowei Zhang | H. W. Zhang, Z. W. Ma, and T. Chen | Effective resistivity for magnetohydrodynamic simulation of collisionless magnetic reconnection | null | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The electron inertia term and the off-diagonal electron pressure terms are well-known for the frozen-in condition breakdown in collisionless magnetic reconnection, which are naturally kinetic and difficult to be employed in magnetohydrodynamic (MHD) simulations. After considering the shortcomings of MHD and Hall MHD in neglecting the important electron dynamics such as the inertia and the nongyrotropic pressure, the kinetic characteristics of electrons and ions in the diffusion region are studied and an effective resistivity model involving dynamics of charged particles is proposed [Z. W. Ma et al. 2018 Sci. Rep. 8 10521]. The amplitude of the effective resistivity is mainly determined by electrons in most realistic situations with large ion-electron mass ratios. In this work, the effective resistivity model for collisionless magnetic reconnection without the guide field is successfully applied in the 2.5D MHD and Hall MHD simulations, which remarkably improves the simulation results compared with traditional MHD models. For the MHD case, the effective resistivity significantly increased the reconnection rate to the reasonable value of ~0.1$B_0v_A$. For the Hall MHD case with effective resistivity, the peak reconnection rate is ~0.25$B_0v_A$, and the major structures of the reconnecting field and the current sheet agree well with the particle-in-cell (PIC) and hybrid simulations. | [
{
"version": "v1",
"created": "Sun, 30 Jan 2022 09:10:14 GMT"
},
{
"version": "v2",
"created": "Mon, 07 Feb 2022 12:48:59 GMT"
},
{
"version": "v3",
"created": "Sat, 26 Apr 2025 10:40:37 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 19:46:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhang",
"H. W.",
""
],
[
"Ma",
"Z. W.",
""
],
[
"Chen",
"T.",
""
]
] |
2201.13001 | Jayanta Dey | Jayanta Dey, Haoyin Xu, Ashwin De Silva, Joshua T. Vogelstein | Simple Calibration via Geodesic Kernels | null | null | null | null | cs.LG cs.AI cs.DS q-bio.NC stat.ML | http://creativecommons.org/licenses/by/4.0/ | Deep discriminative approaches, such as decision forests and deep neural networks, have recently found applications in many important real-world scenarios. However, deploying these learning algorithms in safety-critical applications raises concerns, particularly when it comes to ensuring calibration for both in-distribution and out-of-distribution regions. Many popular methods for in-distribution (ID) calibration, such as isotonic and Platt's sigmoidal regression, exhibit adequate ID calibration performance. However, these methods are not calibrated for the entire feature space, leading to overconfidence in the out-of-distribution (OOD) region. Existing OOD calibration methods generally exhibit poor ID calibration. In this paper, we jointly address the ID and OOD problems. We leveraged the fact that deep models learn to partition feature space into a union of polytopes, that is, flat-sided geometric objects. We introduce a geodesic distance to measure the distance between these polytopes and further distinguish samples within the same polytope using a Gaussian kernel. Our experiments on both tabular and vision benchmarks show that the proposed approaches, namely Kernel Density Forest (KDF) and Kernel Density Network (KDN), obtain well-calibrated posteriors for both ID and OOD samples, while mostly preserving the classification accuracy and extrapolating beyond the training data to handle OOD inputs appropriately. | [
{
"version": "v1",
"created": "Mon, 31 Jan 2022 05:07:16 GMT"
},
{
"version": "v2",
"created": "Sun, 06 Feb 2022 14:38:23 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Feb 2022 14:14:47 GMT"
},
{
"version": "v4",
"created": "Fri, 25 Mar 2022 03:41:23 GMT"
},
{
"version": "v5",
"created": "Fri, 19 May 2023 21:15:28 GMT"
},
{
"version": "v6",
"created": "Thu, 19 Oct 2023 02:27:07 GMT"
},
{
"version": "v7",
"created": "Tue, 12 Mar 2024 12:57:20 GMT"
},
{
"version": "v8",
"created": "Fri, 07 Jun 2024 17:10:00 GMT"
},
{
"version": "v9",
"created": "Wed, 04 Jun 2025 23:15:10 GMT"
},
{
"version": "v10",
"created": "Wed, 11 Jun 2025 15:08:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Dey",
"Jayanta",
""
],
[
"Xu",
"Haoyin",
""
],
[
"De Silva",
"Ashwin",
""
],
[
"Vogelstein",
"Joshua T.",
""
]
] |
2202.03146 | Minheng Xiao | Minheng Xiao | Time-Series K-means in Causal Inference and Mechanism Clustering for Financial Data | null | null | null | null | q-fin.ST | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the application of Time Series K-means (TS-K-means) within the context of causal inference and mechanism clustering of financial time series data. Traditional clustering approaches like K-means often rely on static distance metrics, such as Euclidean distance, which inadequately capture the temporal dependencies intrinsic to financial returns. By incorporating Dynamic Time Warping (DTW) as a distance metric, TS-K-means addresses this limitation, improving the robustness of clustering in time-dependent financial data. This study extends the Additive Noise Model Mixture Model (ANM-MM) framework by integrating TS-K-means, facilitating more accurate causal inference and mechanism clustering. The approach is validated through simulations and applied to real-world financial data, demonstrating its effectiveness in enhancing the analysis of complex financial time series, particularly in identifying causal relationships and clustering data based on underlying generative mechanisms. The results show that TS-K-means outperforms traditional K-means, especially with smaller datasets, while maintaining robust causal direction detection as the dataset size changes. | [
{
"version": "v1",
"created": "Mon, 24 Jan 2022 17:01:59 GMT"
},
{
"version": "v2",
"created": "Wed, 24 Aug 2022 10:56:38 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Nov 2022 19:32:45 GMT"
},
{
"version": "v4",
"created": "Tue, 20 Aug 2024 14:28:06 GMT"
},
{
"version": "v5",
"created": "Fri, 27 Dec 2024 21:22:14 GMT"
},
{
"version": "v6",
"created": "Tue, 28 Jan 2025 02:48:07 GMT"
},
{
"version": "v7",
"created": "Tue, 10 Jun 2025 21:03:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xiao",
"Minheng",
""
]
] |
2203.03486 | Andrei Okounkov | David Kazhdan and Andrei Okounkov | On the unramified Eisenstein spectrum | Comments welcome | null | null | null | math.NT math.AG math.AT math.RT | http://creativecommons.org/licenses/by/4.0/ | For a split reductive group ${}^L G$ over a global field, we determine the spectrum of the spherical Hecke algebra coming from the unramified Eisenstein series for the minimal parabolic ${}^L B$. This is done using a certain decomposition of the Springer stack $T^*(B \backslash G/ B)$ for the Langlands dual group in the additive group of cobordisms of cohomologically proper derived quotient stacks. | [
{
"version": "v1",
"created": "Thu, 03 Mar 2022 22:33:51 GMT"
},
{
"version": "v2",
"created": "Sun, 03 Apr 2022 04:15:25 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 23:34:13 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kazhdan",
"David",
""
],
[
"Okounkov",
"Andrei",
""
]
] |
2203.11873 | Sebastien Coube-Sisqueille | S\'ebastien Coube-Sisqueille, Sudipto Banerjee, Beno\^it Liquet | Nonstationary Spatial Process Models with Spatially Varying Covariance Kernels | null | null | null | null | stat.ME | http://creativecommons.org/licenses/by/4.0/ | Building spatial process models that capture nonstationary behavior while delivering computationally efficient inference is challenging. Nonstationary spatially varying kernels (see, e.g., Paciorek, 2003) offer flexibility and richness, but computation is impeded by high-dimensional parameter spaces resulting from spatially varying process parameters. Matters are exacerbated if the number of locations recording measurements is massive. With limited theoretical tractability, obviating computational bottlenecks requires synergy between model construction and algorithm development. We build a class of scalable nonstationary spatial process models using spatially varying covariance kernels. We implement a Bayesian modeling framework using Hybrid Monte Carlo with nested interweaving. We conduct experiments on synthetic data sets to explore model selection and parameter identifiability, and assess inferential improvements accrued from nonstationary modeling. We illustrate strengths and pitfalls with a data set on remote sensed normalized difference vegetation index. | [
{
"version": "v1",
"created": "Tue, 22 Mar 2022 16:51:50 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Mar 2024 15:23:37 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 15:46:47 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Coube-Sisqueille",
"Sébastien",
""
],
[
"Banerjee",
"Sudipto",
""
],
[
"Liquet",
"Benoît",
""
]
] |
2205.07537 | Mohammed El-Kholany | Mohammed M. S. El-Kholany, Martin Gebser and Konstantin Schekotihin | Decomposition Strategies and Multi-shot ASP Solving for Job-shop Scheduling | This paper is an extended version of our papers presented at the 38th International Conference on Logic Programming (ICLP 2022) and the 24th International Symposium on Practical Aspects of Declarative Languages (PADL 2022), accepted for publication in Logical Methods in Computer Science journal | null | null | null | cs.AI cs.LO | http://creativecommons.org/licenses/by/4.0/ | The Job-shop Scheduling Problem (JSP) is a well-known and challenging combinatorial optimization problem in which tasks sharing a machine are to be arranged in a sequence such that encompassing jobs can be completed as early as possible. In this paper, we investigate problem decomposition into time windows whose operations can be successively scheduled and optimized by means of multi-shot Answer Set Programming (ASP) solving. From a computational perspective, decomposition aims to split highly complex scheduling tasks into better manageable subproblems with a balanced number of operations such that good-quality or even optimal partial solutions can be reliably found in a small fraction of runtime. We devise and investigate a variety of decomposition strategies in terms of the number and size of time windows as well as heuristics for choosing their operations. Moreover, we incorporate time window overlapping and compression techniques into the iterative scheduling process to counteract optimization limitations due to the restriction to window-wise partial schedules. Our experiments on different JSP benchmark sets show that successive optimization by multi-shot ASP solving leads to substantially better schedules within tight runtime limits than single-shot optimization on the full problem. In particular, we find that decomposing initial solutions obtained with proficient heuristic methods into time windows leads to improved solution quality. | [
{
"version": "v1",
"created": "Mon, 16 May 2022 09:33:00 GMT"
},
{
"version": "v2",
"created": "Mon, 23 May 2022 09:32:16 GMT"
},
{
"version": "v3",
"created": "Thu, 08 Aug 2024 13:09:25 GMT"
},
{
"version": "v4",
"created": "Wed, 12 Feb 2025 23:51:26 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 12:24:30 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"El-Kholany",
"Mohammed M. S.",
""
],
[
"Gebser",
"Martin",
""
],
[
"Schekotihin",
"Konstantin",
""
]
] |
2205.08309 | Mark Titleman | Mark Titleman | The Duality of Whittaker Potential Theory: Fundamental Representations of Electromagnetism and Gravity, and Their Orthogonality | null | null | null | null | physics.hist-ph physics.class-ph | http://creativecommons.org/licenses/by/4.0/ | E. T. Whittaker produced two papers in 1903 and 1904 that, although sometimes considered mere mathematical statements (Barrett, 1993), held important implications for physical theory. The Whittaker 1903 paper united electrostatic and gravitational attraction as resulting from longitudinal waves - waves whose wavefronts propagate parallel to their direction. The Whittaker 1904 paper showed that electromagnetic waves resulted from the interference of two such longitudinal waves or scalar potential functions. Although unexplored, the implications of these papers are profound: gravitational lensing, gravitational waves, the Aharonov-Bohm effect, the existence of a hyperspace above or behind normal space, the elimination of gravitational and point charge singularities, MOND, and the expansion of the universe. This last implication can be related to the recent finding that black holes with posited vacuum energy interior solutions alongside cosmological boundaries have a cosmological coupling constant of k=3, meaning that black holes gain mass-proportional to a3 in a parameterization equation within a Robertson-Walker cosmology and are a cosmological accelerated expansion species (Farrah et al., 2023). This expansion and many features of General Relativity can be explained by the mass-proportionality and preferred direction of the longitudinal waves within the two underlying non-local Whittaker potentials (Titleman, 2022). Whittaker potential theory also offers a simple explanation for expansion of the universe - it is produced as longitudinal motion within the Whittaker potentials only when dynamic electromagnetism is separate from time-static gravity in intergalactic space. | [
{
"version": "v1",
"created": "Fri, 13 May 2022 21:43:51 GMT"
},
{
"version": "v2",
"created": "Sun, 22 May 2022 07:02:58 GMT"
},
{
"version": "v3",
"created": "Tue, 04 Oct 2022 17:23:43 GMT"
},
{
"version": "v4",
"created": "Sun, 11 May 2025 14:00:51 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 13:54:53 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Titleman",
"Mark",
""
]
] |
2205.09547 | Orlando Marigliano | Arthur Bik, Orlando Marigliano | Classifying one-dimensional discrete models with maximum likelihood degree one | 39 pages. Accepted version | null | null | null | math.CO math.AG math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a classification of all one-dimensional discrete statistical models with maximum likelihood degree one based on their rational parametrization. We show how all such models can be constructed from members of a smaller class of 'fundamental models' using a finite number of simple operations. We introduce 'chipsplitting games', a class of combinatorial games on a grid which we use to represent fundamental models. This combinatorial perspective enables us to show that there are only finitely many fundamental models in the probability simplex $\Delta_n$ for $n\leq 4$. | [
{
"version": "v1",
"created": "Thu, 19 May 2022 13:18:10 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Feb 2024 15:58:23 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 10:01:23 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Bik",
"Arthur",
""
],
[
"Marigliano",
"Orlando",
""
]
] |
2205.11839 | Pinchen Xie | Pinchen Xie, Yixiao Chen, Xinyu Xu, Zhi Yao, Weinan E, Roberto Car | Ab Initio bulk free energy surface of proper ferroelectrics | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We report a systematic and accurate approach for deriving the bulk free energy surface (FES), a function of temperature, polarization, and strain, from the first-principles density functional theory (DFT) of proper ferroelectrics. The core of our approach is the metadynamics algorithm that extracts the polarization dependence of the FES from all-atom molecular dynamics simulations without an a priori ansatz. The rest of the FES is derived from the metadynamics trajectories that span the relevant phase space. We demonstrate our approach in the case of lead titanate. The errors across the phase transition, due to DFT numerics, all-atom molecular dynamics, and free energy evaluation by enhanced sampling, can be systematically controlled and are of the order of 1meV/atom. The accuracy of the resulting ab initio FES is only limited by the adopted functional approximation of DFT. | [
{
"version": "v1",
"created": "Tue, 24 May 2022 07:00:57 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 23:24:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xie",
"Pinchen",
""
],
[
"Chen",
"Yixiao",
""
],
[
"Xu",
"Xinyu",
""
],
[
"Yao",
"Zhi",
""
],
[
"E",
"Weinan",
""
],
[
"Car",
"Roberto",
""
]
] |
2205.12337 | Yi-Cai Zhang | Yi-Cai Zhang | Proposed realization of critical regions in a one-dimensional flat band lattice with a quasi-periodic potential | null | Sci Rep 14, 17921 (2024) | 10.1038/s41598-024-68851-4 | null | cond-mat.dis-nn cond-mat.quant-gas | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the previous work, the concept of critical region in a generalized Aubry-Andr\'{e} model (Ganeshan-Pixley-Das Sarma's model) has been set up. In this work we propose that the critical region can be realized in a one-dimensional flat band lattice system with a quasi-periodic potential. It is found that the above flat band lattice model can be reduced into an effective Ganeshan-Pixley-Das Sarma's model where the effective parameter $\alpha=V_0/(2E)$ with potential strength $V_0$ and eigenenergy $E$. It is shown that there are very rich physics in this model. Depending on $|\alpha|<1$ or $|\alpha|\geq1$, the effective quasi-periodic potential would be bounded or unbounded. For these two cases, the Lyapunov exponent [$\gamma(E)$], mobility edges ($E_c$) and critical indices ($\nu$) of localized length are obtained exactly. In addition, several localized state regions, extended state regions and critical regions would appear in the parameter $V_0-E$ plane. For a given potential strength $V_0$, the localized-extended and localized-critical transitions can co-exist. Furthermore, we find the critical index of localized length $\xi(E)=1/\gamma(E)$ is $\nu=1$ near localized-extended transitions and $\nu=1/2$ near the localized-critical transitions. Near the transition point between the bound ($|\alpha|<1$) and unbounded ($|\alpha|\geq1$) cases, i.e, $|\alpha|=|V_0/(2E)|= 1$, the derivative of Lypunov exponent of localized states with respect to energy is discontinuous. The localized states in bounded and unbounded cases can be distinguished from each other by Avila's acceleration. At the end, we find that near the transition point, there also exist critical-extended transitions in the phase diagram. | [
{
"version": "v1",
"created": "Fri, 20 May 2022 06:59:33 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhang",
"Yi-Cai",
""
]
] |
2206.11111 | Joshua Frisch R | Anna Erschler, Joshua Frisch | Poisson boundary of group extensions | Criteria formulated more directly in terms of Krull dimension, Substantial changes to make Krull dimension more explicit throughout. Many Small corrections, more unified notation | null | null | null | math.GR math.DS math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given a finitely generated group, the well-known Stability Problem asks whether the non-triviality of the Poisson-Furstenberg boundary (which is equivalent to the existence of non-constant bounded harmonic functions) depends on the choice of simple random walk on the group. This question was far from being understood even in the class of linear groups. Given an amenable group, e.g. a solvable group, there is no known characterisation, even a conjectural one, of when it admits a simple random walk with non-trivial boundary. We provide a characterisation of groups with non-trivial boundary for finitely generated linear groups of characteristic $p$. We prove in particular that the Stability Problem has a positive answer in this class of groups. For linear groups of characteristic $0$, we prove a sufficient condition for the triviality of the boundary which does not depend on the choice of a simple random walk. We conjecture that our sufficient condition is also necessary. Our arguments are based on a new comparison criterion for group extensions, on new $\Delta$-restriction entropy estimates and a criterion for boundary non-triviality, and on a new "cautiousness" criterion for triviality of the boundary. | [
{
"version": "v1",
"created": "Wed, 22 Jun 2022 14:05:25 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jan 2023 16:15:29 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 17:32:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Erschler",
"Anna",
""
],
[
"Frisch",
"Joshua",
""
]
] |
2207.04719 | Rico Henri Herzog | Rico H. Herzog, Juliana E. Gon\c{c}alves, Geertje Slingerland, Reinout Kleinhans, Holger Prang, Frances Brazier, Trivik Verma | Identifying public values and spatial conflicts in urban planning | null | null | 10.1177/00420980231207475 | null | cs.CY cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Identifying the diverse and often competing values of citizens, and resolving the consequent public value conflicts, are of significant importance for inclusive and integrated urban development. Scholars have highlighted that relational, value-laden urban space gives rise to many diverse conflicts that vary both spatially and temporally. Although notions of public value conflicts have been conceived in theory, there are very few empirical studies that identify such values and their conflicts in urban space. Building on public value theory and using a case-study mixed-methods approach, this paper proposes a new approach to empirically investigate public value conflicts in urban space. Using unstructured participatory data of 4,528 citizen contributions from a Public Participation Geographic Information Systems in Hamburg, Germany, natural language processing and spatial clustering techniques are used to identify areas of potential value conflicts. Four expert workshops assess and interpret these quantitative findings. Integrating both quantitative and qualitative results, 19 general public values and a total of 9 archetypical conflicts are identified. On the basis of these results, this paper proposes a new conceptual tool of Public Value Spheres that extends the theoretical notion of public-value conflicts and helps to further account for the value-laden nature of urban space. | [
{
"version": "v1",
"created": "Mon, 11 Jul 2022 08:59:43 GMT"
},
{
"version": "v2",
"created": "Mon, 18 Jul 2022 08:01:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Herzog",
"Rico H.",
""
],
[
"Gonçalves",
"Juliana E.",
""
],
[
"Slingerland",
"Geertje",
""
],
[
"Kleinhans",
"Reinout",
""
],
[
"Prang",
"Holger",
""
],
[
"Brazier",
"Frances",
""
],
[
"Verma",
"Trivik",
""
]
] |
2207.08554 | Lorenzo Carlucci | Lorenzo Carlucci, Leonardo Mainardi | Regressive versions of Hindman's Theorem | Corrected proofs of Proposition 2. Added Theorem 2 | Arch. Math. Logic 63, 447-472 (2024) | 10.1007/s00153-023-00901-6 | Roma01.Math | math.LO math.CO | http://creativecommons.org/licenses/by/4.0/ | When the Canonical Ramsey's Theorem by Erd\H{o}s and Rado is applied to regressive functions one obtains the Regressive Ramsey's Theorem by Kanamori and McAloon. Taylor proved a "canonical" version of Hindman's Theorem, analogous to the Canonical Ramsey's Theorem. We introduce the restriction of Taylor's Canonical Hindman's Theorem to a subclass of the regressive functions, the $\lambda$-regressive functions, relative to an adequate version of min-homogeneity and prove some results about the Reverse Mathematics of this Regressive Hindman's Theorem and of natural restrictions of it.
In particular we prove that the first non-trivial restriction of the principle is equivalent to Arithmetical Comprehension. We furthermore prove that this same principle strongly computably reduces the well-ordering-preservation principle for base-$\omega$ exponentiation. | [
{
"version": "v1",
"created": "Mon, 18 Jul 2022 12:17:22 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Sep 2022 10:26:51 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Jan 2023 21:45:40 GMT"
},
{
"version": "v4",
"created": "Tue, 09 Jan 2024 09:33:28 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Carlucci",
"Lorenzo",
""
],
[
"Mainardi",
"Leonardo",
""
]
] |
2207.14438 | Angus Lowe | Angus Lowe and Ashwin Nayak | Lower Bounds for Learning Quantum States with Single-Copy Measurements | v3. Minor typos fixed and references to subsequent work added. Most of the results in this article were included in the first author's Master's thesis at U. Waterloo (Oct. 2021) and were presented at QIP 2022 (Mar. 2022) | ACM Trans. Comput. Theory 17, 1, Article 7 (March 2025), 42 pages | 10.1145/3717450 | null | quant-ph cs.CC cs.IT cs.LG math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problems of quantum tomography and shadow tomography using measurements performed on individual, identical copies of an unknown $d$-dimensional state. We first revisit a known lower bound due to Haah et al. (2017) on quantum tomography with accuracy $\epsilon$ in trace distance, when the measurements choices are independent of previously observed outcomes (i.e., they are nonadaptive). We give a succinct proof of this result. This leads to stronger lower bounds when the learner uses measurements with a constant number of outcomes. In particular, this rigorously establishes the optimality of the folklore ``Pauli tomography" algorithm in terms of its sample complexity. We also derive novel bounds of $\Omega(r^2 d/\epsilon^2)$ and $\Omega(r^2 d^2/\epsilon^2)$ for learning rank $r$ states using arbitrary and constant-outcome measurements, respectively, in the nonadaptive case.
In addition to the sample complexity, a resource of practical significance for learning quantum states is the number of different measurements used by an algorithm. We extend our lower bounds to the case where the learner performs possibly adaptive measurements from a fixed set of $\exp(O(d))$ measurements. This implies in particular that adaptivity does not give us any advantage using single-copy measurements that are efficiently implementable. We also obtain a similar bound in the case where the goal is to predict the expectation values of a given sequence of observables, a task known as shadow tomography. Finally, in the case of adaptive, single-copy measurements implementable with polynomial-size circuits, we prove that a straightforward strategy based on computing sample means of the given observables is optimal. | [
{
"version": "v1",
"created": "Fri, 29 Jul 2022 02:26:08 GMT"
},
{
"version": "v2",
"created": "Mon, 01 Aug 2022 22:03:52 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 00:10:17 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lowe",
"Angus",
""
],
[
"Nayak",
"Ashwin",
""
]
] |
2208.07552 | Juhyung Park | Juhyung Park, Dongwon Park, Sooyeon Ji, Hyeong-Geol Shin, Se Young Chun, and Jongho Lee | Coil2Coil: Self-supervised MR image denoising using phased-array coil images | 9 pages, 5figures | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Denoising of magnetic resonance images is beneficial in improving the quality of low signal-to-noise ratio images. Recently, denoising using deep neural networks has demonstrated promising results. Most of these networks, however, utilize supervised learning, which requires large training images of noise-corrupted and clean image pairs. Obtaining training images, particularly clean images, is expensive and time-consuming. Hence, methods such as Noise2Noise (N2N) that require only pairs of noise-corrupted images have been developed to reduce the burden of obtaining training datasets. In this study, we propose a new self-supervised denoising method, Coil2Coil (C2C), that does not require the acquisition of clean images or paired noise-corrupted images for training. Instead, the method utilizes multichannel data from phased-array coils to generate training images. First, it divides and combines multichannel coil images into two images, one for input and the other for label. Then, they are processed to impose noise independence and sensitivity normalization such that they can be used for the training images of N2N. For inference, the method inputs a coil-combined image (e.g., DICOM image), enabling a wide application of the method. When evaluated using synthetic noise-added images, C2C shows the best performance against several self-supervised methods, reporting comparable outcomes to supervised methods. When testing the DICOM images, C2C successfully denoised real noise without showing structure-dependent residuals in the error maps. Because of the significant advantage of not requiring additional scans for clean or paired images, the method can be easily utilized for various clinical applications. | [
{
"version": "v1",
"created": "Tue, 16 Aug 2022 05:57:24 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 04:25:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Park",
"Juhyung",
""
],
[
"Park",
"Dongwon",
""
],
[
"Ji",
"Sooyeon",
""
],
[
"Shin",
"Hyeong-Geol",
""
],
[
"Chun",
"Se Young",
""
],
[
"Lee",
"Jongho",
""
]
] |
2209.06175 | Victor Magron | Ngoc Hoang Anh Mai and Victor Magron and Jean-Bernard Lasserre and Kim-Chuan Toh | Tractable hierarchies of convex relaxations for polynomial optimization on the nonnegative orthant | 37 pages, 15 tables | null | null | null | math.OC cs.LG math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider polynomial optimization problems (POP) on a semialgebraic set contained in the nonnegative orthant (every POP on a compact set can be put in this format by a simple translation of the origin). Such a POP can be converted to an equivalent POP by squaring each variable. Using even symmetry and the concept of factor width, we propose a hierarchy of semidefinite relaxations based on the extension of P\'olya's Positivstellensatz by Dickinson-Povh. As its distinguishing and crucial feature, the maximal matrix size of each resulting semidefinite relaxation can be chosen arbitrarily and in addition, we prove that the sequence of values returned by the new hierarchy converges to the optimal value of the original POP at the rate $O(\varepsilon^{-c})$ if the semialgebraic set has nonempty interior. When applied to (i) robustness certification of multi-layer neural networks and (ii) computation of positive maximal singular values, our method based on P\'olya's Positivstellensatz provides better bounds and runs several hundred times faster than the standard Moment-SOS hierarchy. | [
{
"version": "v1",
"created": "Tue, 13 Sep 2022 17:23:50 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 15:28:57 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Mai",
"Ngoc Hoang Anh",
""
],
[
"Magron",
"Victor",
""
],
[
"Lasserre",
"Jean-Bernard",
""
],
[
"Toh",
"Kim-Chuan",
""
]
] |
2209.10934 | Xuanran Zhu | Xuanran Zhu, Chao Zhang, Chenfeng Cao, Youning Li, Yiu Tung Poon, Bei Zeng | Detecting Entanglement by Pure Bosonic Extension | 11 pages, 8 figures | Phys. Rev. Research 6, 013249 (2024) | 10.1103/PhysRevResearch.6.013249 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the realm of quantum information theory, the detection and quantification of quantum entanglement stand as paramount tasks. The relative entropy of entanglement (REE) serves as a prominent measure of entanglement, with extensive applications spanning numerous related fields. The positive partial transpose (PPT) criterion, while providing an efficient method for the computation of REE, unfortunately, falls short when dealing with bound entanglement. In this study, we propose a method termed "pure bosonic extension" to enhance the practicability of $k$-bosonic extensions, which approximates the set of separable states from the "outside", through a hierarchical structure. It enables efficient characterization of the set of $k$-bosonic extendible states, facilitating the derivation of accurate lower bounds for REE. Compared to the Semi-Definite Programming (SDP) approach, such as the symmetric/bosonic extension function in QETLAB, our algorithm supports much larger dimensions and higher values of extension $k$. | [
{
"version": "v1",
"created": "Thu, 22 Sep 2022 11:32:45 GMT"
},
{
"version": "v2",
"created": "Wed, 20 Sep 2023 07:52:17 GMT"
},
{
"version": "v3",
"created": "Wed, 24 Jan 2024 05:25:22 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhu",
"Xuanran",
""
],
[
"Zhang",
"Chao",
""
],
[
"Cao",
"Chenfeng",
""
],
[
"Li",
"Youning",
""
],
[
"Poon",
"Yiu Tung",
""
],
[
"Zeng",
"Bei",
""
]
] |
2210.01697 | Soledad Fern\'andez-Garc\'ia | Luca Bonaventura, Soledad Fern\'andez-Garc\'ia and Macarena G\'omez-M\'armol | Efficient implicit solvers for models of neuronal networks | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce economical versions of standard implicit ODE solvers that are specifically tailored for the efficient and accurate simulation of neural networks. These reformulations allow to achieve a significant increase in the efficiency of network simulations, by reducing the size of the algebraic systems effectively solved at each time step. While we focus here specifically on Explicit first step, Diagonally Implicit Runge Kutta methods (ESDIRK), similar simplifications can also be applied to any implicit ODE solver. In order to demonstrate the capabilities of the proposed methods, we consider networks based on three different single-cell models with slow-fast dynamics, including the classical FitzHugh-Nagumo model, a Intracellular Calcium Concentration model and the Hindmarsh-Rose model. Numerical experiments on the simulation of networks of increasing size based on these models demonstrate the superior efficiency of the proposed economical methods. | [
{
"version": "v1",
"created": "Tue, 04 Oct 2022 15:52:07 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:20:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Bonaventura",
"Luca",
""
],
[
"Fernández-García",
"Soledad",
""
],
[
"Gómez-Mármol",
"Macarena",
""
]
] |
2210.05856 | Qingyun Zeng | Qingyun Zeng | Derived Lie $\infty$-groupoids and algebroids in higher differential geometry | Thesis draft (2021.9). Minor improvements. Comments welcome! | null | null | null | math.DG math.AG math.AT | http://creativecommons.org/licenses/by/4.0/ | We study various problems arising in higher differential geometry using {\it derived Lie $\infty$-groupoids and algebroids}.We first study Lie $\infty$-groupoids in various categories of derived geometric objects in differential geometry, including derived manifolds, derived analytic spaces, derived noncommutative spaces, and derived Banach manifolds. We construct category of fibrant objects (CFO) structures in the category of derived Lie $\infty$-groupoids. Then we study $L_{\infty}$-algebroids which are the infinitesimal counterpart of derived Lie $\infty$-groupoids. We then study the homotopical algebras for derived Lie $\infty$-groupoids and algebroids and study their homotopy-coherent representations, which we call $\infty$-representations. We relate $\infty$-representations of $L_{\infty}$-algebroids to (quasi-) cohesive modules developed by Block, and $\infty$-representations of Lie $\infty$-groupoids to $\infty$-local system introduced by Block-Smith. Then we apply these tools in studying singular foliations and their characteristic classes. We construct Atiyah classes for $L_{\infty}$-algebroids pairs. We study singular foliations and their holonomies. We construct $\L_{\infty}$-algebroids for holomorphic singular foliations, and then We study elliptic involutive structures and prove an dg-enhancement of $V$-analytic coherent sheaves. These examples inspire us to define {\it perfect singular foliations}, which is a subcategory of singular foliation but with better homological algebras. Next, we construct various Lie $\infty$-groupoids for singular foliations. Then we study foliations on stacks and higher groupoids. Finally, we prove an $A_{\infty}$ de Rham theorem for foliations, and Riemann-Hilbert correspondence for foliated $\infty$-local system foliated manifolds. | [
{
"version": "v1",
"created": "Wed, 12 Oct 2022 01:44:39 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 03:20:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zeng",
"Qingyun",
""
]
] |
2211.04654 | Rodrigo Andrade e Silva | Rodrigo Andrade e Silva | General theory of swimming in curved spacetimes | 25 pages, 3 figures | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Swimming in curved spacetimes is a phenomenon whereby free bodies in curved spacetimes are able to propel themselves by performing cyclic internal motions. When originally proposed, it was further suggested that, in the limit of fast internal cycles, the net motion would display a simple geometric-phase character, in which the displacement per cycle would not depend on the time progression of the internal motions but only on the sequence of shapes assumed by the body, like a swimmer in a non-turbulent viscous fluid (low Reynolds number). In this paper we develop a general, covariant theory of swimming in curved spacetimes, describing a technique to study the motion of free, small, light, articulated bodies in general relativity by mapping the problem to an analogue in special relativity. We give considerable attention to the limit of fast cycles and investigate the conditions in which the overall motion could display such geometric-phase behavior. The conclusion, however, is that this simple behavior is only realized in very specific circumstances, depending on the structure of the body, characteristics of internal motions, initial conditions, and symmetries of the spacetime; whereas, in general, our formulas predict a more complicated dynamics. | [
{
"version": "v1",
"created": "Wed, 09 Nov 2022 03:04:13 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:34:57 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Silva",
"Rodrigo Andrade e",
""
]
] |
2211.13396 | Huang Zhiqiang | Zhiqiang Huang and Xiao-Kan Guo | Leggett-Garg inequalities for multitime processes | 13 pages, 1 figure | Quantum Inf Process 24, 172 (2025) | 10.1007/s11128-025-04790-x | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study some aspects of the Leggett-Garg inequalities by using the operator-state formalism for multitime processes. The process tensor in its Choi-state form, which we call process state, is employed to investigate the Leggett-Garg inequalities and their violations. We find the sufficient conditions on process states for the Leggett-Garg inequalities to hold, based on which we find a new way of characterizing the influences on the violation of Leggett-Garg inequalities through the structure of process states. | [
{
"version": "v1",
"created": "Thu, 24 Nov 2022 03:25:50 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Feb 2023 00:31:41 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Feb 2023 01:52:03 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 01:40:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Huang",
"Zhiqiang",
""
],
[
"Guo",
"Xiao-Kan",
""
]
] |
2211.16207 | Jean-Stefan Koskivirta | Wushi Goldring, Naoki Imai, Jean-Stefan Koskivirta | Weights of mod $p$ automorphic forms and partial Hasse invariants | 48 pages, appendix by Wushi Goldring | null | null | null | math.NT math.AG math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For a connected, reductive group $G$ over a finite field endowed with a cocharacter $\mu$, we define the zip cone of $(G,\mu)$ as the cone of all possible weights of mod $p$ automorphic forms on the stack of $G$-zips. This cone is conjectured to coincide with the cone of weights of characteristic $p$ automorphic forms for Hodge-type Shimura varieties of good reduction. We prove in full generality that the cone of weights of characteristic $0$ automorphic forms is contained in the zip cone, which gives further evidence to this conjecture. Furthermore, we determine exactly when the zip cone is generated by the weights of partial Hasse invariants, which is a group-theoretical generalization of a result of Diamond--Kassaei and Goldring--Koskivirta. | [
{
"version": "v1",
"created": "Tue, 29 Nov 2022 13:43:09 GMT"
},
{
"version": "v2",
"created": "Sat, 14 Oct 2023 10:27:13 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 10:20:00 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Goldring",
"Wushi",
""
],
[
"Imai",
"Naoki",
""
],
[
"Koskivirta",
"Jean-Stefan",
""
]
] |
2212.08152 | Lee Kennard | Lee Kennard, Michael Wiemeler, Burkhard Wilking | Positive curvature, torus symmetry, and matroids | 55 pages, 10 figures | J. Eur. Math. Soc. (JEMS), 2025 | 10.4171/JEMS/1656 | null | math.DG math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We identify a link between regular matroids and torus representations all of whose isotropy groups have an odd number of components. Applying Seymour's 1980 classification of the former objects, we obtain a classification of the latter. In addition, we prove optimal upper bounds for the cogirth of regular matroids up to rank nine, and we apply this to prove the existence of fixed-point sets of circles with large dimension in a torus representation with this property up to rank nine. Finally, we apply these results to prove new obstructions to the existence of Riemannian metrics with positive sectional curvature and torus symmetry. | [
{
"version": "v1",
"created": "Thu, 15 Dec 2022 21:35:29 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:13:36 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kennard",
"Lee",
""
],
[
"Wiemeler",
"Michael",
""
],
[
"Wilking",
"Burkhard",
""
]
] |
2301.03113 | Quoc Tran-Dinh | Quoc Tran-Dinh and Yang Luo | Randomized Block-Coordinate Optimistic Gradient Algorithms for Root-Finding Problems | 38 pages, and 6 figures | Mathematics of Operations Research, 2025 | null | UNC-STOR-2022.07.RBCM | math.OC stat.ML | http://creativecommons.org/licenses/by/4.0/ | In this paper, we develop two new randomized block-coordinate optimistic gradient algorithms to approximate a solution of nonlinear equations in large-scale settings, which are called root-finding problems. Our first algorithm is non-accelerated with constant stepsizes, and achieves $\mathcal{O}(1/k)$ best-iterate convergence rate on $\mathbb{E}[ \Vert Gx^k\Vert^2]$ when the underlying operator $G$ is Lipschitz continuous and satisfies a weak Minty solution condition, where $\mathbb{E}[\cdot]$ is the expectation and $k$ is the iteration counter. Our second method is a new accelerated randomized block-coordinate optimistic gradient algorithm. We establish both $\mathcal{O}(1/k^2)$ and $o(1/k^2)$ last-iterate convergence rates on both $\mathbb{E}[ \Vert Gx^k\Vert^2]$ and $\mathbb{E}[ \Vert x^{k+1} - x^{k}\Vert^2]$ for this algorithm under the co-coerciveness of $G$. In addition, we prove that the iterate sequence $\{x^k\}$ converges to a solution almost surely, and $k\Vert Gx^k\Vert$ attains a $o(1/k)$ almost sure convergence rate. Then, we apply our methods to a class of large-scale finite-sum inclusions, which covers prominent applications in machine learning, statistical learning, and network optimization, especially in federated learning. We obtain two new federated learning-type algorithms and their convergence rate guarantees for solving this problem class. | [
{
"version": "v1",
"created": "Sun, 08 Jan 2023 21:46:27 GMT"
},
{
"version": "v2",
"created": "Tue, 07 Mar 2023 19:03:32 GMT"
},
{
"version": "v3",
"created": "Sat, 23 Sep 2023 05:46:22 GMT"
},
{
"version": "v4",
"created": "Mon, 27 Jan 2025 16:36:46 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 15:17:06 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tran-Dinh",
"Quoc",
""
],
[
"Luo",
"Yang",
""
]
] |
2302.06526 | Margherita Solci | Margherita Solci | Nonlocal-interaction vortices | null | SIAM J. Math. Anal. 56(3), 2024 | 10.1137/23M1563438 | null | math.AP math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We consider sequences of quadratic non-local functionals, depending on a small parameter $\e$, that approximate the Dirichlet integral by a well-known result by Bourgain, Brezis and Mironescu. Similarly to what is done for hard-core approximations to vortex energies in the case of the Dirichlet integral, we further scale such energies by $|\log\e|^{-1}$ and restrict them to $S^1$-valued functions. We introduce a notion of convergence of functions to integral currents with respect to which such energies are equi-coercive, and show the converge to a vortex energy, similarly to the limit behaviour of Ginzburg-Landau energies at the vortex scaling. | [
{
"version": "v1",
"created": "Mon, 13 Feb 2023 17:13:19 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 15:10:08 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Solci",
"Margherita",
""
]
] |
2302.07944 | Brandon Trabucco | Brandon Trabucco, Kyle Doherty, Max Gurinas, Ruslan Salakhutdinov | Effective Data Augmentation With Diffusion Models | Update to ICLR 2024 manuscript (https://openreview.net/forum?id=ZWzUA9zeAg), add leafy spurge citations | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data augmentation is one of the most prevalent tools in deep learning, underpinning many recent advances, including those from classification, generative models, and representation learning. The standard approach to data augmentation combines simple transformations like rotations and flips to generate new images from existing ones. However, these new images lack diversity along key semantic axes present in the data. Current augmentations cannot alter the high-level semantic attributes, such as animal species present in a scene, to enhance the diversity of data. We address the lack of diversity in data augmentation with image-to-image transformations parameterized by pre-trained text-to-image diffusion models. Our method edits images to change their semantics using an off-the-shelf diffusion model, and generalizes to novel visual concepts from a few labelled examples. We evaluate our approach on few-shot image classification tasks, and on a real-world weed recognition task, and observe an improvement in accuracy in tested domains. | [
{
"version": "v1",
"created": "Tue, 07 Feb 2023 20:42:28 GMT"
},
{
"version": "v2",
"created": "Thu, 25 May 2023 18:54:38 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:01:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Trabucco",
"Brandon",
""
],
[
"Doherty",
"Kyle",
""
],
[
"Gurinas",
"Max",
""
],
[
"Salakhutdinov",
"Ruslan",
""
]
] |
2303.12200 | Thomas Koerber | Michael Eichmair, Thomas Koerber | Schoen's conjecture for limits of isoperimetric surfaces | Final version to appear in J. Differential Geom | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | Let $(M,g)$ be an $n$-dimensional asymptotically flat Riemannian manifold with nonnegative scalar curvature that admits a noncompact area-minimizing hypersurface $\Sigma \subset M$. In the case where $n = 3$, O. Chodosh and the first-named author have proven that $(M, g)$ is necessarily isometric to Euclidean space, confirming a conjecture of R. Schoen. In this paper, we extend this result to dimension $3 < n \leq 7$ provided that $\Sigma$ arises as a limit of isoperimetric surfaces. By contrast, we prove that when $3 < n \leq 7$, there is no such result for general noncompact area-minimizing $\Sigma \subset M$, even when additional assumptions on the stability of $\Sigma$ are imposed. | [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 21:15:58 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 15:16:57 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Eichmair",
"Michael",
""
],
[
"Koerber",
"Thomas",
""
]
] |
2304.02678 | Laura Monk | Nalini Anantharaman and Laura Monk | Friedman-Ramanujan functions in random hyperbolic geometry and application to spectral gaps | 103 pages, 21 figures, minor corrections | null | null | null | math.SP math.GT | http://creativecommons.org/licenses/by/4.0/ | This two-part paper studies the Weil-Petersson measure on the moduli space of compact hyperbolic surfaces of genus $g$. In this first part, we define "volume functions" $V_g^{T}(l)$ associated with arbitrary topological types $T$ of closed geodesics, generalising the "volume polynomials" studied by M. Mirzakhani for simple closed geodesics. Our programme is to study the structure of these functions, focusing on their behaviour as a function of $l$ in the limit $g\rightarrow +\infty$.
In this first article, motivated by J. Friedman's work on random graphs, we prove that volume functions admit asymptotic expansions to any order in powers of $1/g$, and claim that the coefficients in these expansions belong to a newly-introduced class of functions called "Friedman-Ramanujan functions". We prove the claim for closed geodesics filling a surface of Euler characteristic $0$ and $-1$. This result is then applied to prove that a random hyperbolic surface has spectral gap $\geq 2/9-\epsilon$ with high probability as $g \rightarrow +\infty$, using the trace method and cancellation properties of Friedman--Ramanujan functions. | [
{
"version": "v1",
"created": "Wed, 05 Apr 2023 18:16:09 GMT"
},
{
"version": "v2",
"created": "Mon, 31 Jul 2023 11:25:24 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 12:58:36 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Anantharaman",
"Nalini",
""
],
[
"Monk",
"Laura",
""
]
] |
2305.03038 | Ranjan Maitra | Sattwik Ghosal and Ranjan Maitra | The envelope of a complex Gaussian random variable | 25 pages, 4 figures, 1 table | null | null | null | math.ST eess.SP math.PR stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The envelope of an elliptical Gaussian complex vector, or equivalently, the amplitude or norm of a bivariate normal random vector has application in many weather and signal processing contexts. We explicitly characterize its distribution in the general case through its probability density, cumulative distribution and moment generating function. Moments and limiting distributions are also derived. These derivations are exploited to also characterize the special cases where the bivariate Gaussian mean vector and covariance matrix have a simpler structure, providing new additional insights in many cases. Simulations illustrate the benefits of using our formulae over Monte Carlo methods. We also use our derivations to get a better initial characterization of the distribution of the observed values in structural Magnetic Resonance Imaging datasets, and of wind speed. | [
{
"version": "v1",
"created": "Thu, 04 May 2023 17:57:33 GMT"
},
{
"version": "v2",
"created": "Sun, 07 May 2023 13:36:11 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Jun 2025 16:26:06 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 21:32:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ghosal",
"Sattwik",
""
],
[
"Maitra",
"Ranjan",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.