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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.10834 | Cornelius Carlsson | Cornelius Carlsson, Jaime Saez-Mollejo, Federico Fedele, Stefano Calcaterra, Daniel Chrastina, Giovanni Isella, Georgios Katsaros, Natalia Ares | Automated All-RF Tuning for Spin Qubit Readout and Control | null | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Efficient tuning of spin qubits remains a major bottleneck in scaling semiconductor quantum dot-based quantum processors. A key challenge is the rapid identification of gate voltage regimes suitable for qubit initialisation, control, and readout. Here, we leverage radio-frequency charge sensing to automate spin qubit tuning, achieving a median tuning time of approximately 15 minutes. In a single continuous run, our routine identifies spin qubits at 12 distinct charge transitions in under 17 hours. Beyond tuning, our routine autonomously acquires data revealing the gate-voltage dependence of the exchange interaction, dephasing time, and quality factor -- quantities that vary substantially between charge configurations. These results represent a step change in high-throughput spin qubit tuning and provide a foundation for a systematic and automated exploration of semiconductor quantum circuits. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:52:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Carlsson",
"Cornelius",
""
],
[
"Saez-Mollejo",
"Jaime",
""
],
[
"Fedele",
"Federico",
""
],
[
"Calcaterra",
"Stefano",
""
],
[
"Chrastina",
"Daniel",
""
],
[
"Isella",
"Giovanni",
""
],
[
"Katsaros",
"Georgios",
""
],
[
"Ares",
"Natalia",
""
]
] |
2506.10835 | Francisco G. Montoya | Francisco G. Montoya and Santiago S\'anchez Acevedo | General Reference Frame Identification and Transformation in Unbalanced Power Systems | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Various domains such as power system stability analysis, electric machine modeling, and control of power electronic converters have significantly benefited from the application of coordinate transformations. One of the main benefits is the dimensional reduction, which reduces the complexity of the problems. This paper introduces a novel general transformation based on a geometric framework that directly identifies the plane containing the locus for unbalanced quantities through bivector analysis using Geometric Algebra. The proposed method provides a direct transformation valid for any degree of unbalance in $n$-phase, $(n+1)$-wire sinusoidal systems. The transformation requires only two measurements (voltage or current) taken at different time instants, making it computationally efficient. Moreover, we demonstrate through pure geometric reasoning that our approach is general and encompasses other techniques, such as the classical Clarke transformation. Numerical simulations and experimental validation using a real-time digital simulator and a physical laboratory setup demonstrate the effectiveness of the proposed method. This generalization to multi-dimensional systems, combined with the reduced measurement requirements, represents a significant advancement over existing approaches that are typically restricted to three-phase applications or suffer from computational limitations. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:52:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Montoya",
"Francisco G.",
""
],
[
"Acevedo",
"Santiago Sánchez",
""
]
] |
2506.10836 | Ayel\'en Prado Ms | Ayel\'en Prado, Diego Perez-Morelo, Leonardo Salazar Alarc\'on, Hern\'an Pastoriza | Low-Power Optical Actuation of n-GaAs Cantilevers via Surface Piezoelectric Coupling | null | null | null | null | physics.app-ph physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The mechanical behavior of any semiconductor microstructure is inevitably coupled to light. In the case of micro mechanical resonators (MEMs), carrier generation can affect the quality factor of the resonance through electron-phonon scattering and ohmic losses, but it may also inject energy into the structure and induce movement. Thus, semiconductor MEMs may be regarded as intrinsically optomechanical systems. Here, we report on the optical actuation of a simple-clamped n-GaAs cantilever. This is achieved through modulated nanowatt LED illumination at the resonance frequency. We propose that the mechanism responsible for the coupling between light and movement is piezoelectrically induced stress in the surface depletion layer. Motion may be detected using two methods: by measuring the piezoelectric voltage generated due to deformation or the current arising due to the capacitance changes. In the latter case, a bias voltage must be applied to the device, which leads to nonlinear dynamics. Our results indicate that photothermal and electrostatic effects can be ruled out because i) we measure a very small drift of the resonance frequency with the light modulation offset which, together with finite-element simulations, indicates photothermal effects are not responsible for actuation, ii) we observe a phase shift under different bias voltage polarities, which rules out electrostatic actuation and iii), static measurements performed using an optical profilometer confirm the piezoelectric nature of the interaction with light. In summary, we find homogeneous low power LED illumination to be an effective, simple and convenient method of actuation for piezoelectric semiconductor based MEMs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:52:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Prado",
"Ayelén",
""
],
[
"Perez-Morelo",
"Diego",
""
],
[
"Alarcón",
"Leonardo Salazar",
""
],
[
"Pastoriza",
"Hernán",
""
]
] |
2506.10837 | Vaibhav Pandey | Pulak Banerjee, Chinmoy Dey, M. C. Kumar and Vaibhav Pandey | Threshold resummation for $W$-boson pair production at NNLO+NNLL | 25 pages, 4 figures, 3 tables. arXiv admin note: text overlap with arXiv:2409.16375 | null | null | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | We present results for threshold resummation of the invariant mass distribution, for on-shell production of a pair of $W$-bosons at next-to-next-to-leading order + next-to-next-to-leading logarithmic (NNLO+NNLL) accuracy in QCD. Owing to its sensitivity to the self-interactions between gauge bosons, this process is important to investigate at the energies of the Large Hadron Collider (LHC). We achieve this resummation by exploiting the factorization properties of the soft and virtual parts of the partonic cross-section. Our analysis has been carried out for the invariant mass distribution up to $Q$ = 2500 GeV. At this highest $Q$ we find that, for 13.6 TeV LHC, the NNLL resummation enhances the NNLO cross-sections by about $6.3\%$ and reduces the conventional scale uncertainties from 6.8\% at NNLO to 4.1\% at NNLO+NNLL. We also estimate the intrinsic uncertainties due to the non-perturbative parton distribution functions at the highest perturbative order, for both fixed-order and resummed results, to be around 3\% for $Q \sim$ 2000 GeV. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:53:50 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Banerjee",
"Pulak",
""
],
[
"Dey",
"Chinmoy",
""
],
[
"Kumar",
"M. C.",
""
],
[
"Pandey",
"Vaibhav",
""
]
] |
2506.10838 | Min Sha | Zhiqian Liu, Xiaoting Li, Wenheng Liu, Min Sha | Three quantities arising from B\'{e}zout's identity and resultants of integer polynomials | 14 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study three quantities arising naturally from B\'{e}zout's identity, the resultant and the reduced resultant of two non-zero coprime integer polynomials. We establish several new divisibility relations among them. We also pose two conjectures by making computations. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:55:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liu",
"Zhiqian",
""
],
[
"Li",
"Xiaoting",
""
],
[
"Liu",
"Wenheng",
""
],
[
"Sha",
"Min",
""
]
] |
2506.10839 | Maciej Maliborski | Filip Ficek and Maciej Maliborski | New class of time-periodic solutions to the 1D cubic wave equation | 18 pages, 2 figures. Includes Mathematica code and data files | null | null | null | math.AP math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent papers (arXiv:2407.16507, arXiv:2408.05158) we presented results suggesting the existence of a new class of time-periodic solutions to the defocusing cubic wave equation on a one-dimensional interval with Dirichlet boundary conditions. Here we confirm these findings by rigorously constructing solutions from this class. The proof uses rational arithmetic computations to verify essential operator bounds. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:56:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ficek",
"Filip",
""
],
[
"Maliborski",
"Maciej",
""
]
] |
2506.10840 | Tianrui Zhu | Tianrui Zhu, Houyuan Chen, Ruihao Gong, Michele Magno, Haotong Qin, Kai Zhang | Post-Training Quantization for Video Matting | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video matting is crucial for applications such as film production and virtual reality, yet deploying its computationally intensive models on resource-constrained devices presents challenges. Quantization is a key technique for model compression and acceleration. As an efficient approach, Post-Training Quantization (PTQ) is still in its nascent stages for video matting, facing significant hurdles in maintaining accuracy and temporal coherence. To address these challenges, this paper proposes a novel and general PTQ framework specifically designed for video matting models, marking, to the best of our knowledge, the first systematic attempt in this domain. Our contributions include: (1) A two-stage PTQ strategy that combines block-reconstruction-based optimization for fast, stable initial quantization and local dependency capture, followed by a global calibration of quantization parameters to minimize accuracy loss. (2) A Statistically-Driven Global Affine Calibration (GAC) method that enables the network to compensate for cumulative statistical distortions arising from factors such as neglected BN layer effects, even reducing the error of existing PTQ methods on video matting tasks up to 20%. (3) An Optical Flow Assistance (OFA) component that leverages temporal and semantic priors from frames to guide the PTQ process, enhancing the model's ability to distinguish moving foregrounds in complex scenes and ultimately achieving near full-precision performance even under ultra-low-bit quantization. Comprehensive quantitative and visual results show that our PTQ4VM achieves the state-of-the-art accuracy performance across different bit-widths compared to the existing quantization methods. We highlight that the 4-bit PTQ4VM even achieves performance close to the full-precision counterpart while enjoying 8x FLOP savings. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:57:14 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhu",
"Tianrui",
""
],
[
"Chen",
"Houyuan",
""
],
[
"Gong",
"Ruihao",
""
],
[
"Magno",
"Michele",
""
],
[
"Qin",
"Haotong",
""
],
[
"Zhang",
"Kai",
""
]
] |
2506.10841 | Esmaeil Kavousi Ghafi | Esmaeil Kavousi Ghafi, Oliver Lang, Matthias Wagner, Alexander Melzer, Mario Huemer | Automotive Radar Online Channel Imbalance Estimation via NLMS | null | null | null | null | eess.SP | http://creativecommons.org/licenses/by/4.0/ | Automotive radars are one of the essential enablers of advanced driver assistance systems (ADASs). Continuous monitoring of the functional safety and reliability of automotive radars is a crucial requirement to prevent accidents and increase road safety. One of the most critical aspects to monitor in this context is radar channel imbalances, as they are a key parameter regarding the reliability of the radar. These imbalances may originate from several parameter variations or hardware fatigues, e.g., a solder ball break (SBB), and may affect some radar processing steps, such as the angle of arrival estimation. In this work, a novel method for online estimation of automotive radar channel imbalances is proposed. The proposed method exploits a normalized least mean squares (NLMS) algorithm as a block in the processing chain of the radar to estimate the channel imbalances. The input of this block is the detected targets in the range-Doppler map of the radar on the road without any prior knowledge on the angular parameters of the targets. This property in combination with low computational complexity of the NLMS, makes the proposed method suitable for online channel imbalance estimation, in parallel to the normal operation of the radar. Furthermore, it features reduced dependency on specific targets of interest and faster update rates of the channel imbalance estimation compared to the majority of state-of-the-art methods. This improvement is achieved by allowing for multiple targets in the angular spectrum, whereas most other methods are restricted to only single targets in the angular spectrum. The performance of the proposed method is validated using various simulation scenarios and is supported by measurement results. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:57:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ghafi",
"Esmaeil Kavousi",
""
],
[
"Lang",
"Oliver",
""
],
[
"Wagner",
"Matthias",
""
],
[
"Melzer",
"Alexander",
""
],
[
"Huemer",
"Mario",
""
]
] |
2506.10842 | Syed Ali Reza | Nudrat Fariha, Md Nazmuddin Moin Khan, Md Iqbal Hossain, Syed Ali Reza, Joy Chakra Bortty, Kazi Sharmin Sultana, Md Shadidur Islam Jawad, Saniah Safat, Md Abdul Ahad, Maksuda Begum | Advanced fraud detection using machine learning models: enhancing financial transaction security | null | null | 10.14419/c73kcb17 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The rise of digital payments has accelerated the need for intelligent and scalable systems to detect fraud. This research presents an end-to-end, feature-rich machine learning framework for detecting credit card transaction anomalies and fraud using real-world data. The study begins by merging transactional, cardholder, merchant, and merchant category datasets from a relational database to create a unified analytical view. Through the feature engineering process, we extract behavioural signals such as average spending, deviation from historical patterns, transaction timing irregularities, and category frequency metrics. These features are enriched with temporal markers such as hour, day of week, and weekend indicators to expose all latent patterns that indicate fraudulent behaviours. Exploratory data analysis reveals contextual transaction trends across all the dataset features. Using the transactional data, we train and evaluate a range of unsupervised models: Isolation Forest, One Class SVM, and a deep autoencoder trained to reconstruct normal behavior. These models flag the top 1% of reconstruction errors as outliers. PCA visualizations illustrate each models ability to separate anomalies into a two-dimensional latent space. We further segment the transaction landscape using K-Means clustering and DBSCAN to identify dense clusters of normal activity and isolate sparse, suspicious regions. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 15:59:25 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Fariha",
"Nudrat",
""
],
[
"Khan",
"Md Nazmuddin Moin",
""
],
[
"Hossain",
"Md Iqbal",
""
],
[
"Reza",
"Syed Ali",
""
],
[
"Bortty",
"Joy Chakra",
""
],
[
"Sultana",
"Kazi Sharmin",
""
],
[
"Jawad",
"Md Shadidur Islam",
""
],
[
"Safat",
"Saniah",
""
],
[
"Ahad",
"Md Abdul",
""
],
[
"Begum",
"Maksuda",
""
]
] |
2506.10843 | Feline Lindeboom | Feline Lindeboom, Martijn Brehm, Davide Grossi, Pradeep Murukannaiah | A voice for minorities: diversity in approval-based committee elections under incomplete or inaccurate information | 26 pages, 3 figures | null | null | null | cs.GT | http://creativecommons.org/licenses/by/4.0/ | We study diversity in approval-based committee elections with incomplete or inaccurate information. As standard in the literature on approval-based multi-winner voting, we define diversity according to the maximum coverage problem, which is known to be NP-complete, with a best attainable polynomial time approximation ratio of $1-1/\e$. In the incomplete information model, voters can vote on only a small portion of the candidates. We suggest a greedy algorithm and a local search algorithm that query voters and use the query responses to approximate the total population's opinion. For both algorithms, we prove an upper bound on the number of queries required to get a close to $(1-1/\e)$-approximate solution with high probability. We also provide a lower bound for the query complexity of non-adaptive algorithms, that cannot adapt their querying strategy to readily obtained information. In the inaccurate information setting, voters' responses are corrupted with a probability $p\in(0,\frac{1}{2})$. We provide both an upper and a lower bound for the number of queries required to attain a $(1-1/\e)$-approximate solution with high probability. Finally, using real data from Polis, we see that our algorithms perform remarkably better than the theoretical results suggest, both with incomplete and inaccurate information. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:01:34 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lindeboom",
"Feline",
""
],
[
"Brehm",
"Martijn",
""
],
[
"Grossi",
"Davide",
""
],
[
"Murukannaiah",
"Pradeep",
""
]
] |
2506.10844 | Alireza Salemi | Alireza Salemi, Mukta Maddipatla, Hamed Zamani | CIIR@LiveRAG 2025: Optimizing Multi-Agent Retrieval Augmented Generation through Self-Training | null | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper presents mRAG, a multi-agent retrieval-augmented generation (RAG) framework composed of specialized agents for subtasks such as planning, searching, reasoning, and coordination. Our system uses a self-training paradigm with reward-guided trajectory sampling to optimize inter-agent collaboration and enhance response generation. Evaluated on DataMorgana-derived datasets during the SIGIR 2025 LiveRAG competition, mRAG outperforms conventional RAG baselines. We further analyze competition outcomes and showcase the framework's strengths with case studies, demonstrating its efficacy for complex, real-world RAG tasks. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:02:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Salemi",
"Alireza",
""
],
[
"Maddipatla",
"Mukta",
""
],
[
"Zamani",
"Hamed",
""
]
] |
2506.10845 | Fabian Kuhn | Salwa Faour and Fabian Kuhn | Faster CONGEST Approximation Algorithms for Maximum Weighted Independent Set in Sparse Graphs | 23 pages | null | null | null | cs.DS cs.DC | http://creativecommons.org/licenses/by/4.0/ | The maximum independent set problem is a classic optimization problem that has also been studied quite intensively in the distributed setting. While the problem is hard to approximate in general, there are good approximation algorithms known for several sparse graph families. In this paper, we consider deterministic distributed CONGEST algorithms for the weighted version of the problem in trees and graphs of bounded arboricity.
For trees, we prove that the task of deterministically computing a $(1-\epsilon)$-approximate solution to the maximum weight independent set (MWIS) problem has a tight $\Theta(\log^*(n) / \epsilon)$ complexity. The lower bound already holds on unweighted oriented paths. On the upper bound side, we show that the bound can be achieved even in unrooted trees.
For graphs $G=(V,E)$ of arboricity $\beta>1$, we give two algorithms. If the sum of all node weights is $w(V)$, we show that for any $\epsilon>0$, an independent set of weight at least $(1-\epsilon)\cdot \frac{w(V)}{4\beta}$ can be computed in $O(\log^2(\beta/\epsilon)/\epsilon + \log^* n)$ rounds. This result is obtained by a direct application of the local rounding framework of Faour, Ghaffari, Grunau, Kuhn, and Rozho\v{n} [SODA '23]. We further show that for any $\epsilon>0$, an independent set of weight at least $(1-\epsilon)\cdot\frac{w(V)}{2\beta+1}$ can be computed in $O(\log^3(\beta)\cdot\log(1/\epsilon)/\epsilon^2 \cdot\log n)$ rounds. This improves on a recent result of Gil [OPODIS '23], who showed that a $1/\lfloor(2+\epsilon)\beta\rfloor$-approximation to the MWIS problem can be computed in $O(\beta\cdot\log n)$ rounds. As an intermediate step, we design an algorithm to compute an independent set of total weight at least $(1-\epsilon)\cdot\sum_{v\in V}\frac{w(v)}{deg(v)+1}$ in time $O(\log^3(\Delta)\cdot\log(1/\epsilon)/\epsilon + \log^* n)$, where $\Delta$ is the maximum degree of the graph. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:05:36 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Faour",
"Salwa",
""
],
[
"Kuhn",
"Fabian",
""
]
] |
2506.10846 | Chengcheng Xin | Chengcheng Xin, Maximiliano Isi, Will M. Farr, Zolt\'an Haiman | Identifying Compact Chirping SMBHBs in LSST using Bayesian Analysis | null | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Legacy Survey of Space and Time (LSST) is expected to observe up to ${\sim}100$ million quasars in the next decade. In this work, we show that it is possible to use such data to measure the characteristic frequency evolution of a "chirp" induced by gravitational waves, which can serve as robust evidence for the presence of a compact supermassive black-hole binary. Following the LSST specifications, we generate mock lightcurves consisting of (i) a post-Newtonian chirp produced by orbital motion through, e.g., relativistic Doppler boosting, (ii) a damped random walk representing intrinsic quasar variability, and (iii) Gaussian photometric errors, while assuming non-uniform observations with extended gaps over a period of 10 yr. Through a fully-Bayesian analysis, we show that we can simultaneously measure the chirp and noise parameters with little degeneracy between the two. For chirp signals with an amplitude of $A = 0.5$ mag and a range of times to merger ($t_m = 15{-}10^4$ yr), we can typically measure a non-zero amplitude and positive frequency derivative with over $5\sigma$ credibility. For binaries with $t_m = 50$ yr, we achieve $3\sigma$ ($5\sigma$) confidence that the signal is chirping for $A \gtrsim 0.1$ ($A > 0.2$). Our analysis can take as little as 35 s (and typically $<$ 10 min) to run, making it scalable to a large number of lightcurves. This implies that LSST could, on its own, establish the presence of a compact supermassive black-hole binary, and thus discover gravitational wave sources detectable by LISA and by Pulsar Timing Arrays. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:05:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xin",
"Chengcheng",
""
],
[
"Isi",
"Maximiliano",
""
],
[
"Farr",
"Will M.",
""
],
[
"Haiman",
"Zoltán",
""
]
] |
2506.10847 | Roldao da Rocha | R. Casadio, C. Noberto Souza, R. da Rocha | Quantum gravitational corrections at third-order curvature, acoustic analog black holes and their quasinormal modes | 45 pages, 18 figures | null | null | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quasinormal modes for bosonic (scalar, electromagnetic, and axial gravitational) and fermionic field perturbations, radiated from black holes that carry quantum gravitational corrections at third order in the curvature to the Schwarzschild solution, are scrutinized from the propagation of analog transonic sound waves across a de Laval nozzle. The thermodynamic variables, the nozzle geometry, the Mach number, and the thrust coefficient are computed as functions of the parameter driving the effective action for quantum gravity containing a dimension-six local operator beyond general relativity. The quasinormal modes for quantum gravitational corrected analog black holes are also determined for higher overtones, yielding a more precise description of the quantum-corrected ringdown process and the gravitational waveform way before the fundamental mode sets in. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:07:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Casadio",
"R.",
""
],
[
"Souza",
"C. Noberto",
""
],
[
"da Rocha",
"R.",
""
]
] |
2506.10848 | Dongrui Liu | Qingyan Wei, Yaojie Zhang, Zhiyuan Liu, Dongrui Liu, Linfeng Zhang | Accelerating Diffusion Large Language Models with SlowFast: The Three Golden Principles | 11 pages; 5 figures; | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion-based language models (dLLMs) have emerged as a promising alternative to traditional autoregressive LLMs by enabling parallel token generation and significantly reducing inference latency. However, existing sampling strategies for dLLMs, such as confidence-based or semi-autoregressive decoding, often suffer from static behavior, leading to suboptimal efficiency and limited flexibility. In this paper, we propose SlowFast Sampling, a novel dynamic sampling strategy that adaptively alternates between exploratory and accelerated decoding stages. Our method is guided by three golden principles: certainty principle, convergence principle, and positional principle, which govern when and where tokens can be confidently and efficiently decoded. We further integrate our strategy with dLLM-Cache to reduce redundant computation. Extensive experiments across benchmarks and models show that SlowFast Sampling achieves up to 15.63$\times$ speedup on LLaDA with minimal accuracy drop, and up to 34.22$\times$ when combined with caching. Notably, our approach outperforms strong autoregressive baselines like LLaMA3 8B in throughput, demonstrating that well-designed sampling can unlock the full potential of dLLMs for fast and high-quality generation. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:08:28 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wei",
"Qingyan",
""
],
[
"Zhang",
"Yaojie",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Liu",
"Dongrui",
""
],
[
"Zhang",
"Linfeng",
""
]
] |
2506.10849 | Ma\"el Le Treust | Luis M. Brice\~no-Arias and Ma\"el Le Treust | Bregman proximal gradient method for linear optimization under entropic constraints | null | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | In this paper, we present an efficient algorithm for solving a linear optimization problem with entropic constraints -- a class of problems that arises in game theory and information theory. Our analysis distinguishes between the cases of active and inactive constraints, addressing each using a Bregman proximal gradient method with entropic Legendre functions, for which we establish an ergodic convergence rate of $O(1/n)$ in objective values. For a specific cost structure, our framework provides a theoretical justification for the well-known Blahut-Arimoto algorithm. In the active constraint setting, we include a bisection procedure to approximate the strictly positive Lagrange multiplier. The efficiency of the proposed method is illustrated through comparisons with standard optimization solvers on a representative example from game theory, including extensions to higher-dimensional settings. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:08:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Briceño-Arias",
"Luis M.",
""
],
[
"Treust",
"Maël Le",
""
]
] |
2506.10850 | Derek Benham | Derek Benham, Easton Potokar, Joshua G. Mangelson | Invariant Extended Kalman Filter for Autonomous Surface Vessels with Partial Orientation Measurements | Presented at the 2025 IEEE ICRA Workshop on Field Robotics. 8 pages, 4 figures, 2 tables | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Autonomous surface vessels (ASVs) are increasingly vital for marine science, offering robust platforms for underwater mapping and inspection. Accurate state estimation, particularly of vehicle pose, is paramount for precise seafloor mapping, as even small surface deviations can have significant consequences when sensing the seafloor below. To address this challenge, we propose an Invariant Extended Kalman Filter (InEKF) framework designed to integrate partial orientation measurements. While conventional estimation often relies on relative position measurements to fixed landmarks, open ocean ASVs primarily observe a receding horizon. We leverage forward-facing monocular cameras to estimate roll and pitch with respect to this horizon, which provides yaw-ambiguous partial orientation information. To effectively utilize these measurements within the InEKF, we introduce a novel framework for incorporating such partial orientation data. This approach contrasts with traditional InEKF implementations that assume full orientation measurements and is particularly relevant for planar vehicle motion constrained to a "seafaring plane." This paper details the developed InEKF framework; its integration with horizon-based roll/pitch observations and dual-antenna GPS heading measurements for ASV state estimation; and provides a comparative analysis against the InEKF using full orientation and a Multiplicative EKF (MEKF). Our results demonstrate the efficacy and robustness of the proposed partial orientation measurements for accurate ASV state estimation in open ocean environments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:08:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Benham",
"Derek",
""
],
[
"Potokar",
"Easton",
""
],
[
"Mangelson",
"Joshua G.",
""
]
] |
2506.10851 | Adel Chehade | Adel Chehade, Edoardo Ragusa, Paolo Gastaldo, Rodolfo Zunino | Energy-Efficient Deep Learning for Traffic Classification on Microcontrollers | Accepted at IEEE ISCC 2025 | null | null | null | cs.NI cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we present a practical deep learning (DL) approach for energy-efficient traffic classification (TC) on resource-limited microcontrollers, which are widely used in IoT-based smart systems and communication networks. Our objective is to balance accuracy, computational efficiency, and real-world deployability. To that end, we develop a lightweight 1D-CNN, optimized via hardware-aware neural architecture search (HW-NAS), which achieves 96.59% accuracy on the ISCX VPN-NonVPN dataset with only 88.26K parameters, a 20.12K maximum tensor size, and 10.08M floating-point operations (FLOPs). Moreover, it generalizes across various TC tasks, with accuracies ranging from 94% to 99%. To enable deployment, the model is quantized to INT8, suffering only a marginal 1-2% accuracy drop relative to its Float32 counterpart. We evaluate real-world inference performance on two microcontrollers: the high-performance STM32F746G-DISCO and the cost-sensitive Nucleo-F401RE. The deployed model achieves inference latencies of 31.43ms and 115.40ms, with energy consumption of 7.86 mJ and 29.10 mJ per inference, respectively. These results demonstrate the feasibility of on-device encrypted traffic analysis, paving the way for scalable, low-power IoT security solutions. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:10:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chehade",
"Adel",
""
],
[
"Ragusa",
"Edoardo",
""
],
[
"Gastaldo",
"Paolo",
""
],
[
"Zunino",
"Rodolfo",
""
]
] |
2506.10852 | Mathias Braun | Mathias Braun, Clemens S\"amann | Gromov's reconstruction theorem and measured Gromov-Hausdorff convergence in Lorentzian geometry | 35 pages. Comments welcome | null | null | null | math.DG gr-qc math-ph math.MG math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish Gromov's celebrated reconstruction theorem in Lorentzian geometry. Alongside this result, we introduce and study a natural concept of isomorphy of normalized bounded Lorentzian metric measure spaces. We outline applications to the spacetime reconstruction problem from causal set theory. Lastly, we propose three notions of convergence of (isomorphism classes of) normalized bounded Lorentzian metric measure spaces, for which we prove several fundamental properties. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:11:20 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Braun",
"Mathias",
""
],
[
"Sämann",
"Clemens",
""
]
] |
2506.10853 | Yu Zhang | Yu Zhang, Yang Hu, and De Wang | A Study on Individual Spatiotemporal Activity Generation Method Using MCP-Enhanced Chain-of-Thought Large Language Models | null | null | null | null | cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Human spatiotemporal behavior simulation is critical for urban planning research, yet traditional rule-based and statistical approaches suffer from high computational costs, limited generalizability, and poor scalability. While large language models (LLMs) show promise as "world simulators," they face challenges in spatiotemporal reasoning including limited spatial cognition, lack of physical constraint understanding, and group homogenization tendencies. This paper introduces a framework integrating chain-of-thought (CoT) reasoning with Model Context Protocol (MCP) to enhance LLMs' capability in simulating spatiotemporal behaviors that correspond with validation data patterns. The methodology combines human-like progressive reasoning through a five-stage cognitive framework with comprehensive data processing via six specialized MCP tool categories: temporal management, spatial navigation, environmental perception, personal memory, social collaboration, and experience evaluation. Experiments in Shanghai's Lujiazui district validate the framework's effectiveness across 1,000 generated samples. Results demonstrate high similarity with real mobile signaling data, achieving generation quality scores of 7.86 to 8.36 across different base models. Parallel processing experiments show efficiency improvements, with generation times decreasing from 1.30 to 0.17 minutes per sample when scaling from 2 to 12 processes. This work contributes to integrating CoT reasoning with MCP for urban behavior modeling, advancing LLMs applications in urban computing and providing a practical approach for synthetic mobility data generation. The framework offers a foundation for smart city planning, transportation forecasting, and participatory urban design applications. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:14:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Yu",
""
],
[
"Hu",
"Yang",
""
],
[
"Wang",
"De",
""
]
] |
2506.10854 | P\'al Andr\'as Papp | P\'al Andr\'as Papp, Aleksandros Sobczyk, A. N. Yzelman | The Impact of Partial Computations on the Red-Blue Pebble Game | Published in the 37th ACM Symposium on Parallelism in Algorithms and Architectures (SPAA 2025) | null | 10.1145/3694906.3743320 | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study an extension of the well-known red-blue pebble game (RBP) with partial computation steps, inspired by the recent work of Sobczyk. While the original RBP assumes that we need to have all the inputs of an operation in fast memory at the same time, in many concrete computations, the inputs can be aggregated one by one into the final output value. These partial computation steps can enable pebbling strategies with much smaller I/O cost, and in settings where such a step-by-step aggregation is possible, this extended red-blue pebble game offers a much more realistic cost model.
We establish the fundamental properties of this partial-computing red-blue pebble game (PRBP), and compare it to the original RBP. We begin with some simple examples where allowing partial computations can decrease the optimal I/O cost. It is also shown that the cost can decrease by up to a linear factor this way, but in general, it is NP-hard to decide whether partial computations allow for a smaller cost in a specific DAG. We then discuss how $S$-partitions, a crucial tool for deriving I/O lower bounds in RBP, can be adapted to the PRBP model. These new tools are then used to establish lower bounds on the I/O cost of some prominent computational tasks. Finally, we also adapt a hardness result from RBP, showing that the optimum cost is still NP-hard to approximate in PRBP to any reasonable factor. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:16:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Papp",
"Pál András",
""
],
[
"Sobczyk",
"Aleksandros",
""
],
[
"Yzelman",
"A. N.",
""
]
] |
2506.10855 | Michele Gubian | Michele Gubian and Ioana Krehan and Oli Liu and James Kirby and Sharon Goldwater | Analyzing the relationships between pretraining language, phonetic, tonal, and speaker information in self-supervised speech models | null | null | null | null | cs.CL eess.AS | http://creativecommons.org/licenses/by/4.0/ | Analyses of self-supervised speech models have begun to reveal where and how they represent different types of information. However, almost all analyses have focused on English. Here, we examine how wav2vec2 models trained on four different languages encode both language-matched and non-matched speech. We use probing classifiers and geometric analyses to examine how phones, lexical tones, and speaker information are represented. We show that for all pretraining and test languages, the subspaces encoding phones, tones, and speakers are largely orthogonal, and that layerwise patterns of probing accuracy are similar, with a relatively small advantage for matched-language phone and tone (but not speaker) probes in the later layers. Our findings suggest that the structure of representations learned by wav2vec2 is largely independent of the speech material used during pretraining. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:16:38 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Gubian",
"Michele",
""
],
[
"Krehan",
"Ioana",
""
],
[
"Liu",
"Oli",
""
],
[
"Kirby",
"James",
""
],
[
"Goldwater",
"Sharon",
""
]
] |
2506.10856 | Julie Zhang | Julie Zhang, Noah A. Rosenberg, Julia A. Palacios | The space of multifurcating ranked tree shapes: enumeration, lattice structure, and Markov chains | 42 pages, 17 figures | null | null | null | math.PR math.CO q-bio.PE | http://creativecommons.org/licenses/by/4.0/ | Coalescent models of bifurcating genealogies are used to infer evolutionary parameters from molecular data. However, there are many situations where bifurcating genealogies do not accurately reflect the true underlying ancestral history of samples, and a multifurcating genealogy is required. The space of multifurcating genealogical trees, where nodes can have more than two descendants, is largely underexplored in the setting of coalescent inference. In this paper, we examine the space of rooted, ranked, and unlabeled multifurcating trees. We recursively enumerate the space and then construct a partial ordering which induces a lattice on the space of multifurcating ranked tree shapes. The lattice structure lends itself naturally to defining Markov chains that permit exploration on the space of multifurcating ranked tree shapes. Finally, we prove theoretical bounds for the mixing time of two Markov chains defined on the lattice, and we present simulation results comparing the distribution of trees and tree statistics under various coalescent models to the uniform distribution on this tree space. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:16:49 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Julie",
""
],
[
"Rosenberg",
"Noah A.",
""
],
[
"Palacios",
"Julia A.",
""
]
] |
2506.10857 | Jiashuo Yu | Jiashuo Yu, Yue Wu, Meng Chu, Zhifei Ren, Zizheng Huang, Pei Chu, Ruijie Zhang, Yinan He, Qirui Li, Songze Li, Zhenxiang Li, Zhongying Tu, Conghui He, Yu Qiao, Yali Wang, Yi Wang, Limin Wang | VRBench: A Benchmark for Multi-Step Reasoning in Long Narrative Videos | Technical Report | null | null | null | cs.CV cs.AI cs.MM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present VRBench, the first long narrative video benchmark crafted for evaluating large models' multi-step reasoning capabilities, addressing limitations in existing evaluations that overlook temporal reasoning and procedural validity. It comprises 1,010 long videos (with an average duration of 1.6 hours), along with 9,468 human-labeled multi-step question-answering pairs and 30,292 reasoning steps with timestamps. These videos are curated via a multi-stage filtering process including expert inter-rater reviewing to prioritize plot coherence. We develop a human-AI collaborative framework that generates coherent reasoning chains, each requiring multiple temporally grounded steps, spanning seven types (e.g., event attribution, implicit inference). VRBench designs a multi-phase evaluation pipeline that assesses models at both the outcome and process levels. Apart from the MCQs for the final results, we propose a progress-level LLM-guided scoring metric to evaluate the quality of the reasoning chain from multiple dimensions comprehensively. Through extensive evaluations of 12 LLMs and 16 VLMs on VRBench, we undertake a thorough analysis and provide valuable insights that advance the field of multi-step reasoning. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:17:17 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yu",
"Jiashuo",
""
],
[
"Wu",
"Yue",
""
],
[
"Chu",
"Meng",
""
],
[
"Ren",
"Zhifei",
""
],
[
"Huang",
"Zizheng",
""
],
[
"Chu",
"Pei",
""
],
[
"Zhang",
"Ruijie",
""
],
[
"He",
"Yinan",
""
],
[
"Li",
"Qirui",
""
],
[
"Li",
"Songze",
""
],
[
"Li",
"Zhenxiang",
""
],
[
"Tu",
"Zhongying",
""
],
[
"He",
"Conghui",
""
],
[
"Qiao",
"Yu",
""
],
[
"Wang",
"Yali",
""
],
[
"Wang",
"Yi",
""
],
[
"Wang",
"Limin",
""
]
] |
2506.10858 | Zhenhuan Zhou | Zhenhuan Zhou | Med-URWKV: Pure RWKV With ImageNet Pre-training For Medical Image Segmentation | Preprint Draft, 5 pages. This paper will be updated with a formal version in the future, Copyright: College of Computer Science, Nankai University. All rights reserved | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Medical image segmentation is a fundamental and key technology in computer-aided diagnosis and treatment. Previous methods can be broadly classified into three categories: convolutional neural network (CNN) based, Transformer based, and hybrid architectures that combine both. However, each of them has its own limitations, such as restricted receptive fields in CNNs or the computational overhead caused by the quadratic complexity of Transformers. Recently, the Receptance Weighted Key Value (RWKV) model has emerged as a promising alternative for various vision tasks, offering strong long-range modeling capabilities with linear computational complexity. Some studies have also adapted RWKV to medical image segmentation tasks, achieving competitive performance. However, most of these studies focus on modifications to the Vision-RWKV (VRWKV) mechanism and train models from scratch, without exploring the potential advantages of leveraging pre-trained VRWKV models for medical image segmentation tasks. In this paper, we propose Med-URWKV, a pure RWKV-based architecture built upon the U-Net framework, which incorporates ImageNet-based pretraining to further explore the potential of RWKV in medical image segmentation tasks. To the best of our knowledge, Med-URWKV is the first pure RWKV segmentation model in the medical field that can directly reuse a large-scale pre-trained VRWKV encoder. Experimental results on seven datasets demonstrate that Med-URWKV achieves comparable or even superior segmentation performance compared to other carefully optimized RWKV models trained from scratch. This validates the effectiveness of using a pretrained VRWKV encoder in enhancing model performance. The codes will be released. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:19:18 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhou",
"Zhenhuan",
""
]
] |
2506.10859 | Kehan Long | Kehan Long, Shasha Li, Chen Xu, Jintao Tang, Ting Wang | Precise Zero-Shot Pointwise Ranking with LLMs through Post-Aggregated Global Context Information | Accepted by SIGIR 2025 | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements have successfully harnessed the power of Large Language Models (LLMs) for zero-shot document ranking, exploring a variety of prompting strategies. Comparative approaches like pairwise and listwise achieve high effectiveness but are computationally intensive and thus less practical for larger-scale applications. Scoring-based pointwise approaches exhibit superior efficiency by independently and simultaneously generating the relevance scores for each candidate document. However, this independence ignores critical comparative insights between documents, resulting in inconsistent scoring and suboptimal performance. In this paper, we aim to improve the effectiveness of pointwise methods while preserving their efficiency through two key innovations: (1) We propose a novel Global-Consistent Comparative Pointwise Ranking (GCCP) strategy that incorporates global reference comparisons between each candidate and an anchor document to generate contrastive relevance scores. We strategically design the anchor document as a query-focused summary of pseudo-relevant candidates, which serves as an effective reference point by capturing the global context for document comparison. (2) These contrastive relevance scores can be efficiently Post-Aggregated with existing pointwise methods, seamlessly integrating essential Global Context information in a training-free manner (PAGC). Extensive experiments on the TREC DL and BEIR benchmark demonstrate that our approach significantly outperforms previous pointwise methods while maintaining comparable efficiency. Our method also achieves competitive performance against comparative methods that require substantially more computational resources. More analyses further validate the efficacy of our anchor construction strategy. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:20:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Long",
"Kehan",
""
],
[
"Li",
"Shasha",
""
],
[
"Xu",
"Chen",
""
],
[
"Tang",
"Jintao",
""
],
[
"Wang",
"Ting",
""
]
] |
2506.10860 | Vinit Vijay Deshpande | Vinit Vijay Deshpande and Romana Piat | Failure modes of irregular ceramic foam under compression: Development of a new image based strut segmentation strategy | null | null | null | null | cond-mat.soft physics.app-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The work investigates the failure modes of the microstructure of an irregular ceramic foam subjected to uniaxial compression loading. The foam material is manufactured using the direct foaming method and has polydispersed pores homogeneously distributed in the microstructure. The effective stress-strain curve and the macroscopic strength of the material differs significantly with the predictions of the Gibson-Ashby model that assumes a regular microstructure. The objective of this work is to determine the reasons. The work builds upon an image segmentation algorithm that utilizes a skeletonization method followed by geometric pruning strategies to isolate the struts in the foam microstructure. In this work, a novel pruning strategy defined by a physics-based significance measure is proposed, which identifies the struts whose failure leads to macroscopic failure of the material. The stresses in the struts are calculated by a finite element simulation, which is then utilized to determine their failure modes. This also reveals the relationship between the strut orientation and their failure modes.The energy dissipated by the individual failure modes as the loading is increased shows that there are two dominant modes which are different from the bending failure reported in the Gibson-Ashby model. These failure modes are mixed compression-tension and pure compression. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:22:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Deshpande",
"Vinit Vijay",
""
],
[
"Piat",
"Romana",
""
]
] |
2506.10861 | Yingxin Qi | Yingxin Qi, Hanyu Wang, and Qinghua Guo, Zhihong Zhu and Biao Yang | Photonic chiral bulk transports manipulated by boundary freedom in three-dimensional meta-crystals | null | null | null | null | cond-mat.mes-hall physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In topological physics, one of the most intriguing phenomena is the presence of topological boundary states, accurately predicted by the well-established bulk-edge correspondence. For example, in three-dimensional Weyl semimetals, Fermi arcs emerge to connect projected Weyl points on the surface due to inheriting the bulk-edge correspondence from the integer quantum Hall effect. However, limited attention has been paid to exploring the reverse mechanism in topological crystals. In this study, we propose that boundaries can serve as an alternative degree of freedom to manipulate topological bulk transports. We analytically and experimentally validate our concept using a finite-thickness photonic meta-crystal that supports bulk nodal lines, with its zeroth modes exhibiting opposite chiral bulk transports under different boundary conditions. Notably, the mirror symmetry remains preserved across both configurations. These findings are applicable to other topological systems, providing new insights into systems with varied boundary conditions and offering the potential for the design of more compact and spatially efficient topological photonic devices. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:22:52 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Qi",
"Yingxin",
""
],
[
"Wang",
"Hanyu",
""
],
[
"Guo",
"Qinghua",
""
],
[
"Zhu",
"Zhihong",
""
],
[
"Yang",
"Biao",
""
]
] |
2506.10862 | Rui Zhang | Rui Zhang, Qi Meng, Han Wan, Yang Liu, Zhi-Ming Ma, Hao Sun | OmniFluids: Unified Physics Pre-trained Modeling of Fluid Dynamics | null | null | null | null | physics.flu-dyn cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-fidelity and efficient simulation of fluid dynamics drive progress in various scientific and engineering applications. Traditional computational fluid dynamics methods offer strong interpretability and guaranteed convergence, but rely on fine spatial and temporal meshes, incurring prohibitive computational costs. Physics-informed neural networks (PINNs) and neural operators aim to accelerate PDE solvers using deep learning techniques. However, PINNs require extensive retraining and careful tuning, and purely data-driven operators demand large labeled datasets. Hybrid physics-aware methods embed numerical discretizations into network architectures or loss functions, but achieve marginal speed gains and become unstable when balancing coarse priors against high-fidelity measurements. To this end, we introduce OmniFluids, a unified physics pre-trained operator learning framework that integrates physics-only pre-training, coarse-grid operator distillation, and few-shot fine-tuning, which enables fast inference and accurate prediction under limited or zero data supervision. For architectural design, the key components of OmniFluids include a mixture of operators, a multi-frame decoder, and factorized Fourier layers, which enable efficient and scalable modeling of diverse physical tasks while maintaining seamless integration with physics-based supervision. Across a broad range of two- and three-dimensional benchmarks, OmniFluids significantly outperforms state-of-the-art AI-driven methods in flow field reconstruction and turbulence statistics accuracy, delivering 10-100x speedups compared to classical solvers, and accurately recovers unknown physical parameters from sparse, noisy data. This work establishes a new paradigm for efficient and generalizable surrogate modeling in complex fluid systems under limited data availability. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:23:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Rui",
""
],
[
"Meng",
"Qi",
""
],
[
"Wan",
"Han",
""
],
[
"Liu",
"Yang",
""
],
[
"Ma",
"Zhi-Ming",
""
],
[
"Sun",
"Hao",
""
]
] |
2506.10863 | Nicholas Williams | Nicholas Williams, Kara Rudolph, Iv\'an D\'iaz | Nonparametric estimation of an optimal treatment rule with fused randomized trials and missing effect modifiers | null | null | null | null | stat.AP stat.ME | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A fundamental principle of clinical medicine is that a treatment should only be administered to those patients who would benefit from it. Treatment strategies that assign treatment to patients as a function of their individual characteristics are known as dynamic treatment rules. The dynamic treatment rule that optimizes the outcome in the population is called the optimal dynamic treatment rule. Randomized clinical trials are considered the gold standard for estimating the marginal causal effect of a treatment on an outcome; they are often not powered to detect heterogeneous treatment effects, and thus, may rarely inform more personalized treatment decisions. The availability of multiple trials studying a common set of treatments presents an opportunity for combining data, often called data-fusion, to better estimate dynamic treatment rules. However, there may be a mismatch in the set of patient covariates measured across trials. We address this problem here; we propose a nonparametric estimator for the optimal dynamic treatment rule that leverages information across the set of randomized trials. We apply the estimator to fused randomized trials of medications for the treatment of opioid use disorder to estimate a treatment rule that would match patient subgroups with the medication that would minimize risk of return to regular opioid use. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:24:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Williams",
"Nicholas",
""
],
[
"Rudolph",
"Kara",
""
],
[
"Díaz",
"Iván",
""
]
] |
2506.10864 | Roman Smirnov | Timothy F. Power and Roman G. Smirnov | Rethinking Growth: An Extension of the Solow-Swan Model | null | null | null | null | econ.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The aggregate Cobb-Douglas production function stands as a central element in the renowned Solow-Swan model in economics, providing a crucial theoretical framework for comprehending the determinants of economic growth. This model not only guides policymakers and economists but also influences their decisions, fostering sustainable and inclusive development. In this study, we utilize a one-input version of a new generalization of the Cobb-Douglas production function proposed recently, thereby extending the Solow-Swan model to incorporate energy production as a factor. We offer a rationale for this extension and conduct a comprehensive analysis employing advanced mathematical tools to explore solutions to this new model. This approach allows us to effectively integrate environmental considerations related to energy production into economic growth strategies, fostering long-term sustainability. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:25:27 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Power",
"Timothy F.",
""
],
[
"Smirnov",
"Roman G.",
""
]
] |
2506.10865 | Maurizio Quadrio | Marco Atzori, Gabriele Dini Ciacci and Maurizio Quadrio | Understanding the mismatch between in-vivo and in-silico rhinomanometry | null | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Numerical simulations and clinical measurements of nasal resistance are in quantitative disagreement. Bias introduced by the design of medical devices has not been considered until now as a possible explanation. The aim of present paper is to study the effect of the location of the probe on the rhinomanometer that is meant to measure the ambient pressure. Rhinomanometry is carried out on a 3D silicone model of a patient-specific anatomy; a clinical device and dedicated sensors are employed side-by-side for mutual validation. The same anatomy is also employed for numerical simulations, with approaches spanning a wide range of fidelity levels. We find that the intrinsic uncertainty of the numerical simulations is of minor importance. To the contrary, the position of the pressure tap intended to acquire the external pressure in the clinical device is crucial, and can cause a mismatch comparable to that generally observed between in-silico and in-vivo rhinomanometry data. A source of systematic bias may therefore exist in rhinomanometers, designed under the assumption that measurements of the nasal resistance are unaffected by the flow development within the instruments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:27:34 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Atzori",
"Marco",
""
],
[
"Ciacci",
"Gabriele Dini",
""
],
[
"Quadrio",
"Maurizio",
""
]
] |
2506.10866 | Hanqing Zhang | Hanqing Zhang, Junyu Mao, Mohammad Fahim Shakib, and Giordano Scarciotti | Data-Driven Model Reduction by Moment Matching for Linear and Nonlinear Parametric Systems | 16 pages, 6 figures, submitted to IEEE Transactions on Automatic Control | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Theory and methods to obtain parametric reduced-order models by moment matching are presented. The definition of the parametric moment is introduced, and methods (model-based and data-driven) for the approximation of the parametric moment of linear and nonlinear parametric systems are proposed. These approximations are exploited to construct families of parametric reduced-order models that match the approximate parametric moment of the system to be reduced and preserve key system properties such as asymptotic stability and dissipativity. The use of the model reduction methods is illustrated by means of a parametric benchmark model for the linear case and a large-scale wind farm model for the nonlinear case. In the illustration, a comparison of the proposed approximation methods is drawn and their advantages/disadvantages are discussed. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:27:49 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Hanqing",
""
],
[
"Mao",
"Junyu",
""
],
[
"Shakib",
"Mohammad Fahim",
""
],
[
"Scarciotti",
"Giordano",
""
]
] |
2506.10867 | Philip Bittihn | Lukas Hupe, Jonas Isensee, Ramin Golestanian, Philip Bittihn | Prediction and control of geometry-induced nematic order in growing multicellular systems | 12 pages, 6 figures | null | null | null | cond-mat.soft physics.bio-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In densely-packed two-dimensional systems of growing cells, such as rod-shaped bacteria, a number of experimental and numerical studies report distinct patterns of nematic orientational order in the presence of confinement. So far, these effects have been explained using variations of growing active nematic continuum theories, which incorporate feedback between growth-induced active stresses, the resulting material flow and nematic orientation, and were adapted to the specific geometry under investigation. Here, we first show that a direct, analytical prediction of orientation patterns based on a simple isotropic-growth assumption and the shear rate tensor of the expansion flow already covers previously observed cases. We use this method to tune orientation patterns and net topological defect charge in a systematic way using domain geometry, confirmed by agent-based simulations. We then show how this framework can be extended to quantitatively capture alignment strength, and explore its potential for cross-prediction across different geometries. Our simplified and unifying theoretical framework highlights the role of domain geometry in shaping nematic order of growing systems, and thereby provides a way to forward-engineer desired orientation patterns. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:28:30 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Hupe",
"Lukas",
""
],
[
"Isensee",
"Jonas",
""
],
[
"Golestanian",
"Ramin",
""
],
[
"Bittihn",
"Philip",
""
]
] |
2506.10868 | Simon Lang | Simon Lang, Martin Leutbecher, Pedro Maciel | A multi-scale loss formulation for learning a probabilistic model with proper score optimisation | null | null | null | null | physics.ao-ph cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | We assess the impact of a multi-scale loss formulation for training probabilistic machine-learned weather forecasting models. The multi-scale loss is tested in AIFS-CRPS, a machine-learned weather forecasting model developed at the European Centre for Medium-Range Weather Forecasts (ECMWF). AIFS-CRPS is trained by directly optimising the almost fair continuous ranked probability score (afCRPS). The multi-scale loss better constrains small scale variability without negatively impacting forecast skill. This opens up promising directions for future work in scale-aware model training. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:30:18 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lang",
"Simon",
""
],
[
"Leutbecher",
"Martin",
""
],
[
"Maciel",
"Pedro",
""
]
] |
2506.10869 | Quinn Thibeault | Quinn Thibeault and Giulia Pedrielli | MultiCoSim: A Python-based Multi-Fidelity Co-Simulation Framework | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Simulation is a foundational tool for the analysis and testing of cyber-physical systems (CPS), underpinning activities such as algorithm development, runtime monitoring, and system verification. As CPS grow in complexity and scale, particularly in safety-critical and learning-enabled settings, accurate analysis and synthesis increasingly rely on the rapid use of simulation experiments. Because CPS inherently integrate hardware, software, and physical processes, simulation platforms must support co-simulation of heterogeneous components at varying levels of fidelity. Despite recent advances in high-fidelity modeling of hardware, firmware, and physics, co-simulation in diverse environments remains challenging. These limitations hinder the development of reusable benchmarks and impede the use of simulation for automated and comparative evaluation.
Existing simulation tools often rely on rigid configurations, lack automation support, and present obstacles to portability and modularity. Many are configured through static text files or impose constraints on how simulation components are represented and connected, making it difficult to flexibly compose systems or integrate components across platforms.
To address these challenges, we introduce MultiCoSim, a Python-based simulation framework that enables users to define, compose, and configure simulation components programmatically. MultiCoSim supports distributed, component-based co-simulation and allows seamless substitution and reconfiguration of components. We demonstrate the flexibility of MultiCoSim through case studies that include co-simulations involving custom automaton-based controllers, as well as integration with off-the-shelf platforms like the PX4 autopilot for aerial robotics. These examples highlight MultiCoSim's capability to streamline CPS simulation pipelines for research and development. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:31:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Thibeault",
"Quinn",
""
],
[
"Pedrielli",
"Giulia",
""
]
] |
2506.10870 | Xiaojun Chang | Yuxin Li, Meijie Yang and Xiaojun Chang | Normalized solutions for a Sobolev critical quasilinear Schr\"odinger equation | 57pages | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study the existence of normalized solutions for the following quasilinear Schr\"odinger equation with critical exponent:
\begin{equation*}
-\Delta u-u\Delta (u^2)+\lambda u=\tau|u|^{q-2}u+|u|^{2\cdot2^*-2}u,~~~~x\in\R^N,
\end{equation*}
under the mass constraint $\int_{\R^N}|u|^2dx=c$ for some prescribed $c>0$. Here $\tau\in \mathbb{R}$ is a parameter, $\lambda\in\R$ appears as a Lagrange multiplier, $N\ge3$, $2^*:=\frac{2N}{N-2}$ and $2<q<2\cdot2^*$. By deriving precise energy level estimates and establishing new convergence theorems, we apply the perturbation method to establish several existence results for $\tau>0$ in the Sobolev critical regime:
[label=(\alph*)]
\item For the case of $2<q<2+\frac{4}{N}$, we obtain the existence of two solutions, one of which is a local minimizer, and the other is a mountain pass type solution, under explicit conditions on $c>0$;
\item For the case of $2+\frac{4}{N}\leq q<4+\frac{4}{N}$, we obtain the existence of normalized solutions of mountain pass type under different conditions on $c>0$;
\item For the case of $4+\frac{4}{N}\leq q<2\cdot2^*$, we obtain the existence of a ground state normalized solution under different conditions on $c>0$.
Moreover, when $\tau\le 0$, we derive the non-existence result for $2<q<2\cdot2^*$ and all $c>0$. Our research provides a comprehensive analysis across the entire range $q\in(2, 2 \cdot 2^*)$ and for all $N\ge3$. The methods we have developed are flexible and can be extended to a broader class of nonlinearities. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:34:10 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Li",
"Yuxin",
""
],
[
"Yang",
"Meijie",
""
],
[
"Chang",
"Xiaojun",
""
]
] |
2506.10871 | Pierre-Fran\c{c}ois Massiani | Pierre-Fran\c{c}ois Massiani and Alexander von Rohr and Lukas Haverbeck and Sebastian Trimpe | Viability of Future Actions: Robust Safety in Reinforcement Learning via Entropy Regularization | 24 pages, 11 figures, 2 tables. Accepted for publication at ECML-PKDD 2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the many recent advances in reinforcement learning (RL), the question of learning policies that robustly satisfy state constraints under unknown disturbances remains open. In this paper, we offer a new perspective on achieving robust safety by analyzing the interplay between two well-established techniques in model-free RL: entropy regularization, and constraints penalization. We reveal empirically that entropy regularization in constrained RL inherently biases learning toward maximizing the number of future viable actions, thereby promoting constraints satisfaction robust to action noise. Furthermore, we show that by relaxing strict safety constraints through penalties, the constrained RL problem can be approximated arbitrarily closely by an unconstrained one and thus solved using standard model-free RL. This reformulation preserves both safety and optimality while empirically improving resilience to disturbances. Our results indicate that the connection between entropy regularization and robustness is a promising avenue for further empirical and theoretical investigation, as it enables robust safety in RL through simple reward shaping. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:34:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Massiani",
"Pierre-François",
""
],
[
"von Rohr",
"Alexander",
""
],
[
"Haverbeck",
"Lukas",
""
],
[
"Trimpe",
"Sebastian",
""
]
] |
2506.10872 | Ziv Scully | Ziv Scully, Alexander Terenin | The Gittins Index: A Design Principle for Decision-Making Under Uncertainty | null | null | null | null | math.OC cs.LG cs.PF math.PR stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Gittins index is a tool that optimally solves a variety of decision-making problems involving uncertainty, including multi-armed bandit problems, minimizing mean latency in queues, and search problems like the Pandora's box model. However, despite the above examples and later extensions thereof, the space of problems that the Gittins index can solve perfectly optimally is limited, and its definition is rather subtle compared to those of other multi-armed bandit algorithms. As a result, the Gittins index is often regarded as being primarily a concept of theoretical importance, rather than a practical tool for solving decision-making problems.
The aim of this tutorial is to demonstrate that the Gittins index can be fruitfully applied to practical problems. We start by giving an example-driven introduction to the Gittins index, then walk through several examples of problems it solves - some optimally, some suboptimally but still with excellent performance. Two practical highlights in the latter category are applying the Gittins index to Bayesian optimization, and applying the Gittins index to minimizing tail latency in queues. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:38:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Scully",
"Ziv",
""
],
[
"Terenin",
"Alexander",
""
]
] |
2506.10873 | Finn Schmolke | Finn Schmolke | Asymptotic Fate of Continuously Monitored Quantum Systems | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A quantum trajectory is the natural response of a quantum system subject to external perturbations due to continuous indirect measurement. We completely characterize the asymptotic behavior of continuously monitored quantum systems in finite dimensions and show that generically, spontaneous irreversible localization transitions on the level of individual realizations occur, where the evolution becomes effectively constrained to one of the irreducible components of the total Hilbert space. More generally, localization can be either complete, where the strongest possible confinement is achieved, or incomplete, where localization terminates prematurely. On the trajectory level, symmetries and conserved quantities are no longer respected and localization transitions occur concurrently with violations of ergodicity. As a result, a generalized update rule emerges that effectively projects the system onto one of several possible time evolutions. The update comes equipped with a generalized Born rule that assigns probabilities to these irreversible events. Spontaneous transitions thus occur probabilistically and can deviate considerably from the behavior of the ensemble. In particular, time and ensemble average no longer commute which gives rise to global violations of ergodicity, while on a local level, ergodicity is restored. We explicitly illustrate the asymptotic behavior of continuously monitored systems in a series of examples demonstrating stabilization of many-body scar states and generation of Bell states from local measurement. Finally, we present two algorithms, one based on simultaneous block diagonalization and one based on quantum trajectories to identify all the minimal orthogonal subspaces of the Lindblad equation and all the extremal stationary states which are supported on them. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:40:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Schmolke",
"Finn",
""
]
] |
2506.10874 | Sarah Toonsi | Sarah A. Toonsi and Jeff S. Shamma | Higher-Order Uncoupled Learning Dynamics and Nash Equilibrium | null | null | null | null | cs.MA cs.GT cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study learnability of mixed-strategy Nash Equilibrium (NE) in general finite games using higher-order replicator dynamics as well as classes of higher-order uncoupled heterogeneous dynamics. In higher-order uncoupled learning dynamics, players have no access to utilities of opponents (uncoupled) but are allowed to use auxiliary states to further process information (higher-order). We establish a link between uncoupled learning and feedback stabilization with decentralized control. Using this association, we show that for any finite game with an isolated completely mixed-strategy NE, there exist higher-order uncoupled learning dynamics that lead (locally) to that NE. We further establish the lack of universality of learning dynamics by linking learning to the control theoretic concept of simultaneous stabilization. We construct two games such that any higher-order dynamics that learn the completely mixed-strategy NE of one of these games can never learn the completely mixed-strategy NE of the other. Next, motivated by imposing natural restrictions on allowable learning dynamics, we introduce the Asymptotic Best Response (ABR) property. Dynamics with the ABR property asymptotically learn a best response in environments that are asymptotically stationary. We show that the ABR property relates to an internal stability condition on higher-order learning dynamics. We provide conditions under which NE are compatible with the ABR property. Finally, we address learnability of mixed-strategy NE in the bandit setting using a bandit version of higher-order replicator dynamics. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:42:01 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Toonsi",
"Sarah A.",
""
],
[
"Shamma",
"Jeff S.",
""
]
] |
2506.10875 | Guanjin Wang | Guanjin Wang, Xiangxue Zhao, Shapour Azarm, Balakumar Balachandran | Data-Driven Prediction of Dynamic Interactions Between Robot Appendage and Granular Material | null | null | null | null | cs.RO cs.AI cs.LG cs.NA math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An alternative data-driven modeling approach has been proposed and employed to gain fundamental insights into robot motion interaction with granular terrain at certain length scales. The approach is based on an integration of dimension reduction (Sequentially Truncated Higher-Order Singular Value Decomposition), surrogate modeling (Gaussian Process), and data assimilation techniques (Reduced Order Particle Filter). This approach can be used online and is based on offline data, obtained from the offline collection of high-fidelity simulation data and a set of sparse experimental data. The results have shown that orders of magnitude reduction in computational time can be obtained from the proposed data-driven modeling approach compared with physics-based high-fidelity simulations. With only simulation data as input, the data-driven prediction technique can generate predictions that have comparable accuracy as simulations. With both simulation data and sparse physical experimental measurement as input, the data-driven approach with its embedded data assimilation techniques has the potential in outperforming only high-fidelity simulations for the long-horizon predictions. In addition, it is demonstrated that the data-driven modeling approach can also reproduce the scaling relationship recovered by physics-based simulations for maximum resistive forces, which may indicate its general predictability beyond a case-by-case basis. The results are expected to help robot navigation and exploration in unknown and complex terrains during both online and offline phases. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:43:21 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Guanjin",
""
],
[
"Zhao",
"Xiangxue",
""
],
[
"Azarm",
"Shapour",
""
],
[
"Balachandran",
"Balakumar",
""
]
] |
2506.10876 | Pritam Chattopadhyay | Pritam Chattopadhyay, Avijit Misra, Tanmoy Pandit, and Goutam Paul | Landauer Principle and Thermodynamics of Computation | null | null | null | null | quant-ph cond-mat.stat-mech cs.CC cs.FL | http://creativecommons.org/licenses/by/4.0/ | According to the Landauer principle, any logically irreversible process accompanies entropy production, which results in heat dissipation in the environment. Erasing of information, one of the primary logically irreversible processes, has a lower bound on heat dissipated into the environment, called the Landauer bound (LB). However, the practical erasure processes dissipate much more heat than the LB. Recently, there have been a few experimental investigations to reach this bound both in the classical and quantum domains. There has also been a spate of activities to enquire about this LB in finite time, with finite-size heat baths, non-Markovian and nonequilibrium environment in the quantum regime where the effects of fluctuations and correlation of the systems with the bath can no longer be ignored. This article provides a comprehensive review of the recent progress on the Landauer bound, which serves as a fundamental principle in the thermodynamics of computation. We also provide a perspective for future endeavors in these directions. Furthermore, we review the recent exploration toward establishing energetic bounds of a computational process. We also review the thermodynamic aspects of error correction, which is an indispensable part of information processing and computations. In doing so, we briefly discuss the basics of these fields to provide a complete picture. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:43:28 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chattopadhyay",
"Pritam",
""
],
[
"Misra",
"Avijit",
""
],
[
"Pandit",
"Tanmoy",
""
],
[
"Paul",
"Goutam",
""
]
] |
2506.10877 | Hongda Sun | Hongda Sun, Jiaren Peng, Wenzhong Yang, Liang He, Bo Du, Rui Yan | Enhancing Medical Dialogue Generation through Knowledge Refinement and Dynamic Prompt Adjustment | ACL 2025 Findings | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Medical dialogue systems (MDS) have emerged as crucial online platforms for enabling multi-turn, context-aware conversations with patients. However, existing MDS often struggle to (1) identify relevant medical knowledge and (2) generate personalized, medically accurate responses. To address these challenges, we propose MedRef, a novel MDS that incorporates knowledge refining and dynamic prompt adjustment. First, we employ a knowledge refining mechanism to filter out irrelevant medical data, improving predictions of critical medical entities in responses. Additionally, we design a comprehensive prompt structure that incorporates historical details and evident details. To enable real-time adaptability to diverse patient conditions, we implement two key modules, Triplet Filter and Demo Selector, providing appropriate knowledge and demonstrations equipped in the system prompt. Extensive experiments on MedDG and KaMed benchmarks show that MedRef outperforms state-of-the-art baselines in both generation quality and medical entity accuracy, underscoring its effectiveness and reliability for real-world healthcare applications. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:44:25 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sun",
"Hongda",
""
],
[
"Peng",
"Jiaren",
""
],
[
"Yang",
"Wenzhong",
""
],
[
"He",
"Liang",
""
],
[
"Du",
"Bo",
""
],
[
"Yan",
"Rui",
""
]
] |
2506.10878 | Haoxiong Yan | Haoxiong Yan, Allen Zang, Joel Grebel, Xuntao Wu, Ming-Han Chou, Gustav Andersson, Christopher R. Conner, Yash J. Joshi, Shiheng Li, Jacob M. Miller, Rhys G. Povey, Hong Qiao, Eric Chitambar, Andrew N. Cleland | Quantum secret sharing in a triangular superconducting quantum network | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a three-node quantum communication testbed with a triangular topology, each side of the triangle formed by a 1.3-meter-long transmission line. We demonstrate state transfer and entanglement generation between any two nodes, generate genuine multipartite entangled GHZ states, and implement quantum secret sharing (QSS) of classical information. Our experiments show that the QSS protocol can effectively detect eavesdropping, ensuring the secure sharing of secrets. This device forms a testbed for robust and secure quantum networking, enabling testing of more advanced quantum communication protocols. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:44:28 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yan",
"Haoxiong",
""
],
[
"Zang",
"Allen",
""
],
[
"Grebel",
"Joel",
""
],
[
"Wu",
"Xuntao",
""
],
[
"Chou",
"Ming-Han",
""
],
[
"Andersson",
"Gustav",
""
],
[
"Conner",
"Christopher R.",
""
],
[
"Joshi",
"Yash J.",
""
],
[
"Li",
"Shiheng",
""
],
[
"Miller",
"Jacob M.",
""
],
[
"Povey",
"Rhys G.",
""
],
[
"Qiao",
"Hong",
""
],
[
"Chitambar",
"Eric",
""
],
[
"Cleland",
"Andrew N.",
""
]
] |
2506.10879 | Pratik Worah | Pratik Worah | A Goemans-Williamson type algorithm for identifying subcohorts in clinical trials | null | null | null | null | q-bio.QM cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We design an efficient algorithm that outputs a linear classifier for identifying homogeneous subsets (equivalently subcohorts) from large inhomogeneous datasets. Our theoretical contribution is a rounding technique, similar to that of Goemans and Williamson (1994), that approximates the optimal solution of the underlying optimization problem within a factor of $0.82$. As an application, we use our algorithm to design a simple test that can identify homogeneous subcohorts of patients, that are mainly comprised of metastatic cases, from the RNA microarray dataset for breast cancer by Curtis et al. (2012). Furthermore, we also use the test output by the algorithm to systematically identify subcohorts of patients in which statistically significant changes in methylation levels of tumor suppressor genes co-occur with statistically significant changes in nuclear receptor expression. Identifying such homogeneous subcohorts of patients can be useful for the discovery of disease pathways and therapeutics, specific to the subcohort. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:44:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Worah",
"Pratik",
""
]
] |
2506.10880 | Viviana Giunzioni | V. Giunzioni, A. Merlini, F. P. Andriulli | Spectral Analysis of Discretized Boundary Integral Operators in 3D: a High-Frequency Perspective | null | null | null | null | cs.CE cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | When modeling propagation and scattering phenomena using integral equations discretized by the boundary element method, it is common practice to approximate the boundary of the scatterer with a mesh comprising elements of size approximately equal to a fraction of the wavelength $\lambda$ of the incident wave, e.g., $\lambda/10$. In this work, by analyzing the spectra of the operator matrices, we show a discrepancy with respect to the continuous operators which grows with the simulation frequency, challenging the common belief that the aforementioned widely used discretization approach is sufficient to maintain the accuracy of the solution constant when increasing the frequency. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:45:13 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Giunzioni",
"V.",
""
],
[
"Merlini",
"A.",
""
],
[
"Andriulli",
"F. P.",
""
]
] |
2506.10881 | Rui Albuquerque | R. Albuquerque | The mirror map and other natural tangent tensors | 16 pages | null | null | null | math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We review the intrinsic geometry of the tangent bundle of a differentiable manifold $M$, aside from any non-natural structures. We recall the properties of the mirror map $B\in\mathrm{End}(TTM)$, known also as the canonical endomorphism or the almost-tangent structure, solve some cohomological questions and raise other induced by the $\mathrm{d}_B$ operator. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:45:26 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Albuquerque",
"R.",
""
]
] |
2506.10882 | Alejandro Rivero-Santamaria | Maxime Infuso, Samuel Del Fr\'e, Gilberto A. Alou, Mathieu Bertin, Jean-Hugues Fillion, Alejandro Rivero Santamar\'ia, and Maurice Monnerville | Deep Potential-Driven Molecular Dynamics of CO Ice Analogues: Investigating Desorption Following Vibrational Excitation | null | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a new deep learning-based machine learning potential (MLP) for molecular dynamics simulations of solid carbon monoxide (CO), capable of accurately describing CO vibrations both in the fundamental state and in highly excited vibrational states, up to approximately v = 40. The MLP is based on the combination of high-dimensional neural network atomic potentials using the DeePMD-kit package, trained on prior ab initio molecular dynamics (AIMD) data, with selective treatment of the excited molecule allowing us to capture complex energy redistribution dynamics in condensed-phase environments. In particular, the MLP is capable of accurately describing the desorption process of a single CO molecule within an aggregate of 50 CO molecules, in excellent agreement with both previous theoretical predictions and experimental measurements. The MLP provides a much finer description of the translational and rotational energy distributions, capturing their character with high fidelity and allowing a more detailed comparison with experimental results. Furthermore, the analysis of the rotational energy, resolved over specific translational energies, revealed new insights into the coupling between translational and rotational degrees of freedom during the photodesorption process. This novel approach opens new perspectives for extensive statistical studies on desorption energies and detailed investigations of surface molecule excitations and the exploration of larger-scale models incorporating periodic boundary conditions to simulate more realistic CO aggregates. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:46:35 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Infuso",
"Maxime",
""
],
[
"Del Fré",
"Samuel",
""
],
[
"Alou",
"Gilberto A.",
""
],
[
"Bertin",
"Mathieu",
""
],
[
"Fillion",
"Jean-Hugues",
""
],
[
"Santamaría",
"Alejandro Rivero",
""
],
[
"Monnerville",
"Maurice",
""
]
] |
2506.10883 | Sai Mu | Xuecong Wang, Sai Mu | Intrinsic defects and 4d/5d transition metal defects in Cr$_2$O$_3$: pathways to enhance the N\'eel temperature | 12 pages, 11 figures | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | First-principles calculations are employed to explore avenues to increase the N\'eel temperature ($T_{\mathrm{N}}$) of the magnetoelectric antiferromagnet Cr$_2$O$_3$ through doping. Employing the hybrid functional method, we calculate the formation energy of intrinsic defects and transition metal dopants (Mo, W, Nb, Ta, Zr, and Hf) to assess their likelihood of formation. Intrinsic defect calculations indicate that Cr interstitials and oxygen vacancies dominate under Cr-rich conditions, whereas Cr vacancies prevail under O-rich conditions. Notably, under Cr-rich conditions, the Fermi level can be pinned slightly above mid-gap due to the formation of Cr interstitials and oxygen vacancies. To assess the influence of dopant on $T_{\mathrm{N}}$ of Cr$_2$O$_3$, we calculate the enhancement of the exchange energy for the spin on the dopant site or on adjacent Cr site using the supercell method. Our study identifies isovalent Mo and W substitution on Cr site as the most promising candidates to increase N\'eel temperature due to the impurity-mediated enhanced exchange interaction for half-filled bands. Formation energy calculations indicate that Mo and W substitution on Cr are easier to form under Cr-rich conditions and a Fermi level near or slightly above the midgap renders a desirable neutral Mo and W defect. This is assisted by the formation of intrinsic Cr interstitial and O vacancy under Cr-rich conditions. These findings offer a route to utilize defects for higher $T_{\mathrm{N}}$ and enhanced performance of Cr$_2$O$_3$ in magnetoelectric devices and furnish invaluable insights for directing subsequent experimental endeavors. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:47:48 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Xuecong",
""
],
[
"Mu",
"Sai",
""
]
] |
2506.10884 | Dong Hae Mangalindan | Dong Hae Mangalindan, Karthik Kandikonda, Ericka Rovira, and Vaibhav Srivastava | Modeling Trust Dynamics in Robot-Assisted Delivery: Impact of Trust Repair Strategies | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With increasing efficiency and reliability, autonomous systems are becoming valuable assistants to humans in various tasks. In the context of robot-assisted delivery, we investigate how robot performance and trust repair strategies impact human trust. In this task, while handling a secondary task, humans can choose to either send the robot to deliver autonomously or manually control it. The trust repair strategies examined include short and long explanations, apology and promise, and denial.
Using data from human participants, we model human behavior using an Input-Output Hidden Markov Model (IOHMM) to capture the dynamics of trust and human action probabilities. Our findings indicate that humans are more likely to deploy the robot autonomously when their trust is high. Furthermore, state transition estimates show that long explanations are the most effective at repairing trust following a failure, while denial is most effective at preventing trust loss.
We also demonstrate that the trust estimates generated by our model are isomorphic to self-reported trust values, making them interpretable. This model lays the groundwork for developing optimal policies that facilitate real-time adjustment of human trust in autonomous systems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:49:30 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Mangalindan",
"Dong Hae",
""
],
[
"Kandikonda",
"Karthik",
""
],
[
"Rovira",
"Ericka",
""
],
[
"Srivastava",
"Vaibhav",
""
]
] |
2506.10885 | Qingda Mai | Qingda (Michael) Mai | Slimming Down LLMs Without Losing Their Minds | 10 pages | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates and validates the impact of fine-tuning on large language model performance, focusing on parameter-efficient methods (LoRA and QLoRA). We evaluate model capabilities across three key domains: (1) commonsense reasoning (HellaSwag), (2) mathematical reasoning (GSM8K), and (3) multi-domain knowledge (MMLU-CS).
Our findings demonstrate that: (1) LoRA-based methods effectively improve task-specific performance while maintaining computational efficiency, and (2) performance strongly depends on alignment between fine-tuning dataset and benchmark tasks. The study provides both theoretical insights into parameter-efficient mechanisms and practical guidance for developers implementing efficient LLM adaptation with limited resources. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:49:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Qingda",
"",
"",
"Michael"
],
[
"Mai",
"",
""
]
] |
2506.10886 | Alex Poliakov | Steven Vasquez-Grinnell and Alex Poliakov | S3 Mirror: S3Mirror: Making Genomic Data Transfers Fast, Reliable, and Observable with DBOS | null | null | null | null | cs.DB q-bio.GN | http://creativecommons.org/licenses/by/4.0/ | To meet the needs of a large pharmaceutical organization, we set out to create S3Mirror - an application for transferring large genomic sequencing datasets between S3 buckets quickly, reliably, and observably. We used the DBOSTransact durable execution framework to achieve these goals and benchmarked the performance and cost of the application. S3Mirror is an open source DBOS Python application that can run in a variety of environments, including DBOS Cloud Pro where it runs as much as 40x faster than AWS DataSync at a fraction of the cost. Moreover, S3Mirror is resilient to failures and allows for real-time filewise observability of ongoing and past transfers. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:50:04 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Vasquez-Grinnell",
"Steven",
""
],
[
"Poliakov",
"Alex",
""
]
] |
2506.10887 | Yixiao Huang | Yixiao Huang, Hanlin Zhu, Tianyu Guo, Jiantao Jiao, Somayeh Sojoudi, Michael I. Jordan, Stuart Russell, Song Mei | Generalization or Hallucination? Understanding Out-of-Context Reasoning in Transformers | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) can acquire new knowledge through fine-tuning, but this process exhibits a puzzling duality: models can generalize remarkably from new facts, yet are also prone to hallucinating incorrect information. However, the reasons for this phenomenon remain poorly understood. In this work, we argue that both behaviors stem from a single mechanism known as out-of-context reasoning (OCR): the ability to deduce implications by associating concepts, even those without a causal link. Our experiments across five prominent LLMs confirm that OCR indeed drives both generalization and hallucination, depending on whether the associated concepts are causally related. To build a rigorous theoretical understanding of this phenomenon, we then formalize OCR as a synthetic factual recall task. We empirically show that a one-layer single-head attention-only transformer with factorized output and value matrices can learn to solve this task, while a model with combined weights cannot, highlighting the crucial role of matrix factorization. Our theoretical analysis shows that the OCR capability can be attributed to the implicit bias of gradient descent, which favors solutions that minimize the nuclear norm of the combined output-value matrix. This mathematical structure explains why the model learns to associate facts and implications with high sample efficiency, regardless of whether the correlation is causal or merely spurious. Ultimately, our work provides a theoretical foundation for understanding the OCR phenomenon, offering a new lens for analyzing and mitigating undesirable behaviors from knowledge injection. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:50:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Yixiao",
""
],
[
"Zhu",
"Hanlin",
""
],
[
"Guo",
"Tianyu",
""
],
[
"Jiao",
"Jiantao",
""
],
[
"Sojoudi",
"Somayeh",
""
],
[
"Jordan",
"Michael I.",
""
],
[
"Russell",
"Stuart",
""
],
[
"Mei",
"Song",
""
]
] |
2506.10888 | Lucas Gnecco Heredia | Lucas Gnecco-Heredia and Benjamin Negrevergne and Yann Chevaleyre | Lattice Climber Attack: Adversarial attacks for randomized mixtures of classifiers | 17 pages including bibliography + 13 pages of supplementary material. Extended version of the article accepted at ECML 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Finite mixtures of classifiers (a.k.a. randomized ensembles) have been proposed as a way to improve robustness against adversarial attacks. However, existing attacks have been shown to not suit this kind of classifier. In this paper, we discuss the problem of attacking a mixture in a principled way and introduce two desirable properties of attacks based on a geometrical analysis of the problem (effectiveness and maximality). We then show that existing attacks do not meet both of these properties. Finally, we introduce a new attack called {\em lattice climber attack} with theoretical guarantees in the binary linear setting, and demonstrate its performance by conducting experiments on synthetic and real datasets. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:53:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Gnecco-Heredia",
"Lucas",
""
],
[
"Negrevergne",
"Benjamin",
""
],
[
"Chevaleyre",
"Yann",
""
]
] |
2506.10889 | Waylon Luo | Waylon Luo (1), Jiapeng Zhao (2), Tong Zhan (3), Qiang Guan (1) ((1) Kent State University, (2) Cisco, (3) Meta) | Adaptive Job Scheduling in Quantum Clouds Using Reinforcement Learning | 10 pages, 6 figures, ICPP 2025 | null | null | null | cs.DC quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Present-day quantum systems face critical bottlenecks, including limited qubit counts, brief coherence intervals, and high susceptibility to errors-all of which obstruct the execution of large and complex circuits. The advancement of quantum algorithms has outpaced the capabilities of existing quantum hardware, making it difficult to scale computations effectively. Additionally, inconsistencies in hardware performance and pervasive quantum noise undermine system stability and computational accuracy. To optimize quantum workloads under these constraints, strategic approaches to task scheduling and resource coordination are essential. These methods must aim to accelerate processing, retain operational fidelity, and reduce the communication burden inherent to distributed setups. One of the persistent challenges in this domain is how to efficiently divide and execute large circuits across multiple quantum processors (QPUs), especially in error-prone environments. In response, we introduce a simulation-based tool that supports distributed scheduling and concurrent execution of quantum jobs on networked QPUs connected via real-time classical channels. The tool models circuit decomposition for workloads that surpass individual QPU limits, allowing for parallel execution through inter-processor communication. Using this simulation environment, we compare four distinct scheduling techniques-among them, a model informed by reinforcement learning. These strategies are evaluated across multiple metrics, including runtime efficiency, fidelity preservation, and communication costs. Our analysis underscores the trade-offs inherent in each approach and highlights how parallelized, noise-aware scheduling can meaningfully improve computational throughput in distributed quantum infrastructures. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:54:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Luo",
"Waylon",
"",
"Kent State University"
],
[
"Zhao",
"Jiapeng",
"",
"Cisco"
],
[
"Zhan",
"Tong",
"",
"Meta"
],
[
"Guan",
"Qiang",
"",
"Kent State University"
]
] |
2506.10890 | Zhao Zhang | Zhao Zhang, Yutao Cheng, Dexiang Hong, Maoke Yang, Gonglei Shi, Lei Ma, Hui Zhang, Jie Shao, Xinglong Wu | CreatiPoster: Towards Editable and Controllable Multi-Layer Graphic Design Generation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Graphic design plays a crucial role in both commercial and personal contexts, yet creating high-quality, editable, and aesthetically pleasing graphic compositions remains a time-consuming and skill-intensive task, especially for beginners. Current AI tools automate parts of the workflow, but struggle to accurately incorporate user-supplied assets, maintain editability, and achieve professional visual appeal. Commercial systems, like Canva Magic Design, rely on vast template libraries, which are impractical for replicate. In this paper, we introduce CreatiPoster, a framework that generates editable, multi-layer compositions from optional natural-language instructions or assets. A protocol model, an RGBA large multimodal model, first produces a JSON specification detailing every layer (text or asset) with precise layout, hierarchy, content and style, plus a concise background prompt. A conditional background model then synthesizes a coherent background conditioned on this rendered foreground layers. We construct a benchmark with automated metrics for graphic-design generation and show that CreatiPoster surpasses leading open-source approaches and proprietary commercial systems. To catalyze further research, we release a copyright-free corpus of 100,000 multi-layer designs. CreatiPoster supports diverse applications such as canvas editing, text overlay, responsive resizing, multilingual adaptation, and animated posters, advancing the democratization of AI-assisted graphic design. Project homepage: https://github.com/graphic-design-ai/creatiposter | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:54:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Zhao",
""
],
[
"Cheng",
"Yutao",
""
],
[
"Hong",
"Dexiang",
""
],
[
"Yang",
"Maoke",
""
],
[
"Shi",
"Gonglei",
""
],
[
"Ma",
"Lei",
""
],
[
"Zhang",
"Hui",
""
],
[
"Shao",
"Jie",
""
],
[
"Wu",
"Xinglong",
""
]
] |
2506.10891 | Ritik Batra | Ritik Batra, Lydia Kim, Ilan Mandel, Amritansh Kwatra, Jane L.E., Steven J. Jackson, Thijs Roumen | (De)composing Craft: An Elementary Grammar for Sharing Expertise in Craft Workflows | 29 pages, 7 figures | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Craft practices rely on evolving archives of skill and knowledge, developed through generations of craftspeople experimenting with designs, materials, and techniques. Better documentation of these practices enables the sharing of knowledge and expertise between sites and generations. However, most documentation focuses solely on the linear steps leading to final artifacts, neglecting the tacit knowledge necessary to improvise, or adapt workflows to meet the unique demands of each craft project. This omission limits knowledge sharing and reduces craft to a mechanical endeavor, rather than a sophisticated way of seeing, thinking, and doing. Drawing on expert interviews and literature from HCI, CSCW and the social sciences, we develop an elementary grammar to document improvisational actions of real-world craft practices. We demonstrate the utility of this grammar with an interface called CraftLink that can be used to analyze expert videos and semi-automatically generate documentation to convey material and contextual variations of craft practices. Our user study with expert crocheters (N=7) using this interface evaluates our grammar's effectiveness in capturing and sharing expert knowledge with other craftspeople, offering new pathways for computational systems to support collaborative archives of knowledge and practice within communities. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:54:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Batra",
"Ritik",
""
],
[
"Kim",
"Lydia",
""
],
[
"Mandel",
"Ilan",
""
],
[
"Kwatra",
"Amritansh",
""
],
[
"E.",
"Jane L.",
""
],
[
"Jackson",
"Steven J.",
""
],
[
"Roumen",
"Thijs",
""
]
] |
2506.10892 | Subham Sekhar Sahoo | Subham Sekhar Sahoo, Justin Deschenaux, Aaron Gokaslan, Guanghan Wang, Justin Chiu, Volodymyr Kuleshov | The Diffusion Duality | ICML 2025. We provide the code at: https://github.com/s-sahoo/duo | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Uniform-state discrete diffusion models hold the promise of fast text generation due to their inherent ability to self-correct. However, they are typically outperformed by autoregressive models and masked diffusion models. In this work, we narrow this performance gap by leveraging a key insight: Uniform-state diffusion processes naturally emerge from an underlying Gaussian diffusion. Our method, Duo, transfers powerful techniques from Gaussian diffusion to improve both training and sampling. First, we introduce a curriculum learning strategy guided by the Gaussian process, doubling training speed by reducing variance. Models trained with curriculum learning surpass autoregressive models in zero-shot perplexity on 3 of 7 benchmarks. Second, we present Discrete Consistency Distillation, which adapts consistency distillation from the continuous to the discrete setting. This algorithm unlocks few-step generation in diffusion language models by accelerating sampling by two orders of magnitude. We provide the code and model checkpoints on the project page: http://s-sahoo.github.io/duo | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:55:35 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sahoo",
"Subham Sekhar",
""
],
[
"Deschenaux",
"Justin",
""
],
[
"Gokaslan",
"Aaron",
""
],
[
"Wang",
"Guanghan",
""
],
[
"Chiu",
"Justin",
""
],
[
"Kuleshov",
"Volodymyr",
""
]
] |
2506.10893 | Davide Fazio | Davide Fazio and Raffaele Mascella | Considerations on Everett J. Nelson's connexive logic | null | null | null | null | math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work explores Everett John Nelson's connexive logic, outlined in his PhD thesis and partially summarized in his 1930 paper \emph{Intensional Relations}, which is obtained by extending the system $\mathsf{NL}$ (reconstructed by E. Mares and F. Paoli) with a weak conjunction elimination rule explicitly assumed in the former but not in the latter. After a preliminary analysis of Nelson's philosophical ideas, we provide an algebraic-relational semantics for his logic and we investigate possible extensions thereof which are able to cope with Nelson's ideas with much more accuracy than the original system. For example, we will inquire into extensions whose algebraic-relational models are endowed with irreflexive incompatibility relations, or determine a ``weakly'' transitive entailment. Such an investigation will allow us to establish relationships between some of the trademarks of Nelson's thought and concepts of prominent importance for connexive logic, as e.g. Kapsner's strong connexivity and superconnexivity, as well as between the algebraic-relational semantics of Nelsonian logics and ordered structures that have gained great attention over the past years, namely partially ordered involutive residuate groupoids and (non-orthomodular) orthoposets. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 16:57:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Fazio",
"Davide",
""
],
[
"Mascella",
"Raffaele",
""
]
] |
2506.10894 | Cristian Guillermo Gebhardt | Pedro B. Bazon, Cristian G. Gebhardt, Gustavo C. Buscaglia and Roberto F. Ausas | Numerical approximation of a PDE-constrained Optimization problem that appears in Data-Driven Computational Mechanics | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We investigate an optimization problem that arises when working within the paradigm of Data-Driven Computational Mechanics. In the context of the diffusion-reaction problem, such an optimization problem seeks for the continuous primal fields (gradient and flux) that are closest to some predefined discrete fields taken from a material data set. The optimization is performed over primal fields that satisfy the physical conservation law and the geometrical compatibility. We consider a reaction term in the conservation law, which has the effect of coupling all the optimality conditions. We first establish the well-posedness in the continuous setting. Then, we propose stable finite element discretizations that consistently approximate the continuous formulation, preserving its saddle-point structure and allowing for equal-order interpolation of all fields. Finally, we demonstrate the effectiveness of the proposed methods through a set of numerical examples. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:00:18 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Bazon",
"Pedro B.",
""
],
[
"Gebhardt",
"Cristian G.",
""
],
[
"Buscaglia",
"Gustavo C.",
""
],
[
"Ausas",
"Roberto F.",
""
]
] |
2506.10895 | Guimeng Liu | Guimeng Liu, Milad Abdollahzadeh, Ngai-Man Cheung | AIR: Zero-shot Generative Model Adaptation with Iterative Refinement | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Zero-shot generative model adaptation (ZSGM) aims to adapt a pre-trained generator to a target domain using only text guidance and without any samples from the target domain. Central to recent ZSGM approaches are directional loss which use the text guidance in the form of aligning the image offset with text offset in the embedding space of a vision-language model like CLIP. This is similar to the analogical reasoning in NLP where the offset between one pair of words is used to identify a missing element in another pair by aligning the offset between these two pairs. However, a major limitation of existing ZSGM methods is that the learning objective assumes the complete alignment between image offset and text offset in the CLIP embedding space, resulting in quality degrade in generated images. Our work makes two main contributions. Inspired by the offset misalignment studies in NLP, as our first contribution, we perform an empirical study to analyze the misalignment between text offset and image offset in CLIP embedding space for various large publicly available datasets. Our important finding is that offset misalignment in CLIP embedding space is correlated with concept distance, i.e., close concepts have a less offset misalignment. To address the limitations of the current approaches, as our second contribution, we propose Adaptation with Iterative Refinement (AIR) which is the first ZSGM approach to focus on improving target domain image quality based on our new insight on offset misalignment.Qualitative, quantitative, and user study in 26 experiment setups consistently demonstrate the proposed AIR approach achieves SOTA performance. Additional experiments are in Supp. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:00:50 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liu",
"Guimeng",
""
],
[
"Abdollahzadeh",
"Milad",
""
],
[
"Cheung",
"Ngai-Man",
""
]
] |
2506.10896 | Thomas Sounack | Thomas Sounack, Joshua Davis, Brigitte Durieux, Antoine Chaffin, Tom J. Pollard, Eric Lehman, Alistair E. W. Johnson, Matthew McDermott, Tristan Naumann, Charlotta Lindvall | BioClinical ModernBERT: A State-of-the-Art Long-Context Encoder for Biomedical and Clinical NLP | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Encoder-based transformer models are central to biomedical and clinical Natural Language Processing (NLP), as their bidirectional self-attention makes them well-suited for efficiently extracting structured information from unstructured text through discriminative tasks. However, encoders have seen slower development compared to decoder models, leading to limited domain adaptation in biomedical and clinical settings. We introduce BioClinical ModernBERT, a domain-adapted encoder that builds on the recent ModernBERT release, incorporating long-context processing and substantial improvements in speed and performance for biomedical and clinical NLP. BioClinical ModernBERT is developed through continued pretraining on the largest biomedical and clinical corpus to date, with over 53.5 billion tokens, and addresses a key limitation of prior clinical encoders by leveraging 20 datasets from diverse institutions, domains, and geographic regions, rather than relying on data from a single source. It outperforms existing biomedical and clinical encoders on four downstream tasks spanning a broad range of use cases. We release both base (150M parameters) and large (396M parameters) versions of BioClinical ModernBERT, along with training checkpoints to support further research. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:01:11 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sounack",
"Thomas",
""
],
[
"Davis",
"Joshua",
""
],
[
"Durieux",
"Brigitte",
""
],
[
"Chaffin",
"Antoine",
""
],
[
"Pollard",
"Tom J.",
""
],
[
"Lehman",
"Eric",
""
],
[
"Johnson",
"Alistair E. W.",
""
],
[
"McDermott",
"Matthew",
""
],
[
"Naumann",
"Tristan",
""
],
[
"Lindvall",
"Charlotta",
""
]
] |
2506.10897 | Daniel Borrajo | Daniel Borrajo and Giuseppe Canonaco and Tom\'as de la Rosa and Alfredo Garrach\'on and Sriram Gopalakrishnan and Simerjot Kaur and Marianela Morales and Sunandita Patra and Alberto Pozanco and Keshav Ramani and Charese Smiley and Pietro Totis and Manuela Veloso | GenPlanX. Generation of Plans and Execution | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Classical AI Planning techniques generate sequences of actions for complex tasks. However, they lack the ability to understand planning tasks when provided using natural language. The advent of Large Language Models (LLMs) has introduced novel capabilities in human-computer interaction. In the context of planning tasks, LLMs have shown to be particularly good in interpreting human intents among other uses. This paper introduces GenPlanX that integrates LLMs for natural language-based description of planning tasks, with a classical AI planning engine, alongside an execution and monitoring framework. We demonstrate the efficacy of GenPlanX in assisting users with office-related tasks, highlighting its potential to streamline workflows and enhance productivity through seamless human-AI collaboration. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:02:27 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Borrajo",
"Daniel",
""
],
[
"Canonaco",
"Giuseppe",
""
],
[
"de la Rosa",
"Tomás",
""
],
[
"Garrachón",
"Alfredo",
""
],
[
"Gopalakrishnan",
"Sriram",
""
],
[
"Kaur",
"Simerjot",
""
],
[
"Morales",
"Marianela",
""
],
[
"Patra",
"Sunandita",
""
],
[
"Pozanco",
"Alberto",
""
],
[
"Ramani",
"Keshav",
""
],
[
"Smiley",
"Charese",
""
],
[
"Totis",
"Pietro",
""
],
[
"Veloso",
"Manuela",
""
]
] |
2506.10898 | Bowy La Rivi\`ere | Rik Mulder, Bowy M. La Rivi\`ere, Natalia Chepiga | Exact zero modes in the interacting Majorana X- and Y-junctions | 10 pages, 8 figures (5 in main text, 3 in appendices) | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We report the appearance of exact zero modes in junctions of interacting Majorana wires. We consider two, three and four short Majorana chains coupled to each other at one edge and forming a longer chain with an impurity bond, Y- and X- junctions correspondingly. The exact zero modes emerge as a result of incommensurate short-range correlations resulting from interacting Majorana fermions and manifest themselves as unavoided crossing of energy levels of in-gap states continuously tuned by hopping and interaction strength. For the Y- and X-junctions, we detect four in-gap states that are clearly visible below the bulk excitations. In the X-junction, the energy levels cross pairwise; in the Y-junction, the crossings happen simultaneously in all four levels. We argue that this peculiarity is a result of fractional degrees of freedom localized at the center of junctions with an odd number of arms, which mediate the interaction of the outer edge states. Furthermore, we show that zero modes in multiple-arm junctions can be effectively modeled by a single chain with a strong impurity bond. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:05:46 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Mulder",
"Rik",
""
],
[
"La Rivière",
"Bowy M.",
""
],
[
"Chepiga",
"Natalia",
""
]
] |
2506.10899 | Dimitri Meunier | Dimitri Meunier, Antoine Moulin, Jakub Wornbard, Vladimir R. Kostic, Arthur Gretton | Demystifying Spectral Feature Learning for Instrumental Variable Regression | null | null | null | null | stat.ML cs.LG stat.ME | http://creativecommons.org/licenses/by/4.0/ | We address the problem of causal effect estimation in the presence of hidden confounders, using nonparametric instrumental variable (IV) regression. A leading strategy employs spectral features - that is, learned features spanning the top eigensubspaces of the operator linking treatments to instruments. We derive a generalization error bound for a two-stage least squares estimator based on spectral features, and gain insights into the method's performance and failure modes. We show that performance depends on two key factors, leading to a clear taxonomy of outcomes. In a good scenario, the approach is optimal. This occurs with strong spectral alignment, meaning the structural function is well-represented by the top eigenfunctions of the conditional operator, coupled with this operator's slow eigenvalue decay, indicating a strong instrument. Performance degrades in a bad scenario: spectral alignment remains strong, but rapid eigenvalue decay (indicating a weaker instrument) demands significantly more samples for effective feature learning. Finally, in the ugly scenario, weak spectral alignment causes the method to fail, regardless of the eigenvalues' characteristics. Our synthetic experiments empirically validate this taxonomy. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:06:43 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Meunier",
"Dimitri",
""
],
[
"Moulin",
"Antoine",
""
],
[
"Wornbard",
"Jakub",
""
],
[
"Kostic",
"Vladimir R.",
""
],
[
"Gretton",
"Arthur",
""
]
] |
2506.10900 | Valdemar Farre Vf | Valdemar Farr\'e, Juan Estrada, David Vega, Luis F Urquiza-Aguiar, Juan A. V\'asquez Peralvo, Symeon Chatzinotas | Dynamic Beyond 5G and 6G Connectivity: Leveraging NTN and RIS Synergies for Optimized Coverage and Capacity in High-Density Environments | 6 pages, 6 figures, 11 tables | null | null | null | cs.NI cs.SY eess.SP eess.SY | http://creativecommons.org/licenses/by/4.0/ | The increasing demand for reliable, high-capacity communication during large-scale outdoor events poses significant challenges for traditional Terrestrial Networks (TNs), which often struggle to provide consistent coverage in high-density environments. This paper presents a novel 6G radio network planning framework that integrates Non-Terrestrial Networks (NTNs) with Reconfigurable Intelligent Surfaces (RISs) to deliver ubiquitous coverage and enhanced network capacity. Our framework overcomes the limitations of conventional deployable base stations by leveraging NTN architectures, including Low Earth Orbit (LEO) satellites and passive RIS platforms seamlessly integrated with Beyond 5G (B5G) TNs. By incorporating advanced B5G technologies such as Massive Multiple Input Multiple Output (mMIMO) and beamforming, and by optimizing spectrum utilization across the C, S, and Ka bands, we implement a rigorous interference management strategy based on a dynamic SINR model. Comprehensive calculations and simulations validate the proposed framework, demonstrating significant improvements in connectivity, reliability, and cost-efficiency in crowded scenarios. This integration strategy represents a promising solution for meeting the evolving demands of future 6G networks. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:08:27 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Farré",
"Valdemar",
""
],
[
"Estrada",
"Juan",
""
],
[
"Vega",
"David",
""
],
[
"Urquiza-Aguiar",
"Luis F",
""
],
[
"Peralvo",
"Juan A. Vásquez",
""
],
[
"Chatzinotas",
"Symeon",
""
]
] |
2506.10901 | Vytautas Paskunas | Vytautas Pa\v{s}k\=unas and Julian Quast | Deformations of pseudocharacters and Mazur's finiteness condition | 16 pages | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | We show that deformation rings $R^{\mathrm{ps}}$ of $G$-pseudocharacters of a profinite group $\Gamma$ are noetherian, when $\Gamma$ satisfies Mazur's finiteness condition. The proof proceeds by reduction to the case when $\Gamma$ is finitely generated, where the result was previously established by the second author. This enables us to extend our work on moduli spaces of $R^{\mathrm{ps}}$-condensed representations of a finitely generated profinite group $\Gamma$, to the groups satisfying Mazur's finiteness condition. We also show that the functor from rigid analytic spaces over $\mathbb{Q}_p$ to sets, which associates to a rigid space $Y$ the set of continuous $\mathcal{O}(Y)$-valued $G$-pseudocharacters of $\Gamma$ is representable by a quasi-Stein rigid analytic space, and we study its general properties. We expect these results to be useful, when studying global Galois representations. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:08:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Paškūnas",
"Vytautas",
""
],
[
"Quast",
"Julian",
""
]
] |
2506.10902 | Stuart White | Christopher Schafhauser, Aaron Tikuisis and Stuart White | Nuclear C*-algebras: 99 problems | 98 pages | null | null | null | math.OA | http://creativecommons.org/licenses/by/4.0/ | We present a collection of questions related to the structure and classification of nuclear C*-algebras. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:09:42 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Schafhauser",
"Christopher",
""
],
[
"Tikuisis",
"Aaron",
""
],
[
"White",
"Stuart",
""
]
] |
2506.10903 | Lan Zhang | Lan Zhang, Marco Valentino, Andre Freitas | Beyond Gold Standards: Epistemic Ensemble of LLM Judges for Formal Mathematical Reasoning | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Autoformalization plays a crucial role in formal mathematical reasoning by enabling the automatic translation of natural language statements into formal languages. While recent advances using large language models (LLMs) have shown promising results, methods for automatically evaluating autoformalization remain underexplored. As one moves to more complex domains (e.g., advanced mathematics), human evaluation requires significant time and domain expertise, especially as the complexity of the underlying statements and background knowledge increases. LLM-as-a-judge presents a promising approach for automating such evaluation. However, existing methods typically employ coarse-grained and generic evaluation criteria, which limit their effectiveness for advanced formal mathematical reasoning, where quality hinges on nuanced, multi-granular dimensions. In this work, we take a step toward addressing this gap by introducing a systematic, automatic method to evaluate autoformalization tasks. The proposed method is based on an epistemically and formally grounded ensemble (EFG) of LLM judges, defined on criteria encompassing logical preservation (LP), mathematical consistency (MC), formal validity (FV), and formal quality (FQ), resulting in a transparent assessment that accounts for different contributing factors. We validate the proposed framework to serve as a proxy for autoformalization assessment within the domain of formal mathematics. Overall, our experiments demonstrate that the EFG ensemble of LLM judges is a suitable emerging proxy for evaluation, more strongly correlating with human assessments than a coarse-grained model, especially when assessing formal qualities. These findings suggest that LLM-as-judges, especially when guided by a well-defined set of atomic properties, could offer a scalable, interpretable, and reliable support for evaluating formal mathematical reasoning. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:09:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Lan",
""
],
[
"Valentino",
"Marco",
""
],
[
"Freitas",
"Andre",
""
]
] |
2506.10904 | Xing Fan | Xing Fan and Lan Cheng | Nuclear Electric Quadrupole Moment-Induced Parity Doubling in Molecules for Symmetry-Violation Searches | 13 pages, 3 figures | null | null | null | physics.atom-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Searches for the nuclear magnetic quadrupole moment (MQM) and nuclear Schiff moment (NSM) have high discovery potential for violations of time ($T$) and parity ($P$) reversal symmetries beyond the Standard Model. Molecules containing heavy nuclei are typically used to enhance the sensitivity to MQMs and NSMs due to their strong internal electric fields and potential octupole deformation. To extract these effects in the laboratory frame, a bias electric field is required to polarize the molecule by mixing states of opposite parity (parity doublets). Typical heavy nuclei that are sensitive to symmetry-violation also possess large nuclear electric quadrupole moments (EQMs) when its nuclear spin is $I\geq1$. We show that EQMs can significantly modify the energy splitting between parity doublet states and thus change the required polarizing electric field. As a result, the EQM-induced energy splitting must be taken into account in designing such experiments. We provide qualitative estimates of parity doubling from EQMs and supporting \textit{ab initio} calculations, along with implications for candidate molecules in symmetry-violation searches. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:10:06 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Fan",
"Xing",
""
],
[
"Cheng",
"Lan",
""
]
] |
2506.10905 | Fernando Garcia | Fernando A. Garcia, Sushma Kumari, Juan Schmidt, Cris Adriano, Aashish Sapkota, Paul C. Canfield, Rebecca Flint and Raquel A. Ribeiro | Physical properties of $R$Co$_{2}$Al$_{8}$ ($R=$ La, Ce, Pr, Nd and Sm) single crystals: an emerging structure-type for anisotropic Kondo lattice studies | 13 pages, 8 figures, submitted to Physical Review B | null | null | null | cond-mat.str-el cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Systematic investigations of rare-earth ($R$) based intermetallic materials is a leading strategy to reveal the underlying mechanisms governing a range of physical phenomena, such as the formation of a Kondo lattice and competing electronic and magnetic anisotropies. In this work, the magnetic, thermal and transport properties of $R$Co$_{2}$Al$_{8}$ ($R=$ La, Ce, Pr, Nd and Sm) single crystals are presented. LaCo$_{2}$Al$_{8}$ is characterized as a Pauli paramagnet and transport measurements, with the current along and perpendicular to the orthorhombic $c$-axis ($\rho_{c}$ and $\rho_{ab}$, respectively), reveal a clear electronic anisotropy, with $\rho_{ab }\approx(4-7)\rho_{c }$ at $300$ K. We show that CeCo$_{2}$Al$_{8}$ is a Kondo-lattice for which the Kondo coherence temperature $T_{\text{K}}^{*}$, deduced from broad maximums in $\rho_{c}$ and $\rho_{ab}$ at $\approx$ 68 and 46 K, respectively, is also anisotropic. This finding is related to a possible underlying anisotropy of the Kondo coupling in CeCo$_{2}$Al$_{8}$. The Pr and Nd-based materials present strong easy-axis anisotropy ($c$-axis) and antiferromagnetic (AFM) orders below $T=4.84$ K and $T=8.1$ K, respectively. Metamagnetic transitions from this AFM to a spin-polarized paramagnetic phase state are investigated by isothermal magnetization measurements. The Sm-based compound is also an easy-axis AFM with a transition at $T=21.6$ K. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:13:11 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Garcia",
"Fernando A.",
""
],
[
"Kumari",
"Sushma",
""
],
[
"Schmidt",
"Juan",
""
],
[
"Adriano",
"Cris",
""
],
[
"Sapkota",
"Aashish",
""
],
[
"Canfield",
"Paul C.",
""
],
[
"Flint",
"Rebecca",
""
],
[
"Ribeiro",
"Raquel A.",
""
]
] |
2506.10906 | George Wilkie | George J. Wilkie | Analytic model for neutral penetration and plasma fueling | 13 pages, 6 figures | null | null | null | physics.plasm-ph physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | Neutral atoms recycled from wall interaction interact with confined plasma, thereby refueling it, most strongly in the region closest to the wall. This occurs near the X-point in diverted configurations, or else near the wall itself in limited configurations. A progression of analytic models are developed for neutral density in the vicinity of a planar or linear source in an ionizing domain. First-principles neutral transport simulations with DEGAS2 are used throughout to test the validity and limits of the model when using equivalent sources. The model is further generalized for strong plasma gradients or the inclusion of charge exchange. An important part of the problem of neutral fueling from recycling is thereby isolated and solved with a closed-form analytic model. A key finding is that charge exchange with the confined plasma can be significantly simplified with a reasonable sacrifice of accuracy by treating it as a loss. The several assumptions inherent to the model (and the simulations to which it is compared) can be adapted according to the particular behavior of neutrals in the divertor and the manner in which they cross the separatrix. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:14:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wilkie",
"George J.",
""
]
] |
2506.10907 | Nathan Berkovits | Nathan Berkovits (ICTP-SAIFR/IFT-UNESP, Sao Paulo) | Topological A-Model for $AdS_5\times S^5$ Superstring and the Maldacena Conjecture | 12 pages harvmac tex | null | null | null | hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | A topological A-model constructed from $CP^{3|4}$ supertwistor variables is proposed for the $AdS_5\times S^5$ superstring. At zero $AdS$ radius, free N=4 d=4 super-Yang-Mills amplitudes are reproduced by topological amplitudes of the corresponding gauged linear sigma model where the closed superstring vertex operator for a trace of $k$ super-Yang-Mills fields is described by a boundary state with U(1) worldsheet instanton number $k$. After turning on a Fayet-Iliopoulis term in the sigma model with coefficient $R^2 =\sqrt{g^2 N}$, the topological amplitudes are claimed to reproduce the 't Hooft expansion of perturbative super-Yang-Mills amplitudes. Finally, this topological A-model is related to the usual $AdS_5\times S^5$ superstring in the pure spinor formalism. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:15:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Berkovits",
"Nathan",
"",
"ICTP-SAIFR/IFT-UNESP, Sao Paulo"
]
] |
2506.10908 | Tijana Zrnic | Emmanuel J. Cand\`es, Andrew Ilyas, Tijana Zrnic | Probably Approximately Correct Labels | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Obtaining high-quality labeled datasets is often costly, requiring either extensive human annotation or expensive experiments. We propose a method that supplements such "expert" labels with AI predictions from pre-trained models to construct labeled datasets more cost-effectively. Our approach results in probably approximately correct labels: with high probability, the overall labeling error is small. This solution enables rigorous yet efficient dataset curation using modern AI models. We demonstrate the benefits of the methodology through text annotation with large language models, image labeling with pre-trained vision models, and protein folding analysis with AlphaFold. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:16:26 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Candès",
"Emmanuel J.",
""
],
[
"Ilyas",
"Andrew",
""
],
[
"Zrnic",
"Tijana",
""
]
] |
2506.10909 | Vin de Silva | Vin de Silva, Chad Giusti, Vladimir Itskov, Michael Robinson, Radmila Sazdanovic, Nikolas Schonsheck, Melvin Vaupel, Iris H. R. Yoon | Dowker's theorem for higher-order relations | 36 pages, 14 figures | null | null | null | math.AT | http://creativecommons.org/licenses/by/4.0/ | Given a relation $R \subseteq I \times J$ between two sets, Dowker's Theorem (1952) states that the homology groups of two associated simplicial complexes, now known as Dowker complexes, are isomorphic. In its modern form, the full result asserts a functorial homotopy equivalence between the two Dowker complexes. What can be said about relations defined on three or more sets? We present a simple generalization to multiway relations of the form $R \subseteq I_1 \times I_2 \times \cdots \times I_m$. The theorem asserts functorial homotopy equivalences between $m$ multiway Dowker complexes and a variant of the rectangle complex of Brun and Salbu from their recent short proof of Dowker's Theorem. Our proof uses Smale's homotopy mapping theorem and factors through a cellular Dowker lemma that expresses the main idea in more general form. To make the geometry more transparent, we work with a class of spaces called prod-complexes then transfer the results to simplicial complexes through a simplexification process. We conclude with a detailed study of ternary relations, identifying seven functorially defined homotopy types and twelve natural transformations between them. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:19:38 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"de Silva",
"Vin",
""
],
[
"Giusti",
"Chad",
""
],
[
"Itskov",
"Vladimir",
""
],
[
"Robinson",
"Michael",
""
],
[
"Sazdanovic",
"Radmila",
""
],
[
"Schonsheck",
"Nikolas",
""
],
[
"Vaupel",
"Melvin",
""
],
[
"Yoon",
"Iris H. R.",
""
]
] |
2506.10910 | Teven Le Scao | Mistral-AI: Abhinav Rastogi, Albert Q. Jiang, Andy Lo, Gabrielle Berrada, Guillaume Lample, Jason Rute, Joep Barmentlo, Karmesh Yadav, Kartik Khandelwal, Khyathi Raghavi Chandu, L\'eonard Blier, Lucile Saulnier, Matthieu Dinot, Maxime Darrin, Neha Gupta, Roman Soletskyi, Sagar Vaze, Teven Le Scao, Yihan Wang, Adam Yang, Alexander H. Liu, Alexandre Sablayrolles, Am\'elie H\'eliou, Am\'elie Martin, Andy Ehrenberg, Anmol Agarwal, Antoine Roux, Arthur Darcet, Arthur Mensch, Baptiste Bout, Baptiste Rozi\`ere, Baudouin De Monicault, Chris Bamford, Christian Wallenwein, Christophe Renaudin, Cl\'emence Lanfranchi, Darius Dabert, Devon Mizelle, Diego de las Casas, Elliot Chane-Sane, Emilien Fugier, Emma Bou Hanna, Gauthier Delerce, Gauthier Guinet, Georgii Novikov, Guillaume Martin, Himanshu Jaju, Jan Ludziejewski, Jean-Hadrien Chabran, Jean-Malo Delignon, Joachim Studnia, Jonas Amar, Josselin Somerville Roberts, Julien Denize, Karan Saxena, Kush Jain, Lingxiao Zhao, Louis Martin, Luyu Gao, L\'elio Renard Lavaud, Marie Pellat, Mathilde Guillaumin, Mathis Felardos, Maximilian Augustin, Micka\"el Seznec, Nikhil Raghuraman, Olivier Duchenne, Patricia Wang, Patrick von Platen, Patryk Saffer, Paul Jacob, Paul Wambergue, Paula Kurylowicz, Pavankumar Reddy Muddireddy, Philom\`ene Chagniot, Pierre Stock, Pravesh Agrawal, Romain Sauvestre, R\'emi Delacourt, Sanchit Gandhi, Sandeep Subramanian, Shashwat Dalal, Siddharth Gandhi, Soham Ghosh, Srijan Mishra, Sumukh Aithal, Szymon Antoniak, Thibault Schueller, Thibaut Lavril, Thomas Robert, Thomas Wang, Timoth\'ee Lacroix, Valeriia Nemychnikova, Victor Paltz, Virgile Richard, Wen-Ding Li, William Marshall, Xuanyu Zhang, Yunhao Tang | Magistral | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce Magistral, Mistral's first reasoning model and our own scalable reinforcement learning (RL) pipeline. Instead of relying on existing implementations and RL traces distilled from prior models, we follow a ground up approach, relying solely on our own models and infrastructure. Notably, we demonstrate a stack that enabled us to explore the limits of pure RL training of LLMs, present a simple method to force the reasoning language of the model, and show that RL on text data alone maintains most of the initial checkpoint's capabilities. We find that RL on text maintains or improves multimodal understanding, instruction following and function calling. We present Magistral Medium, trained for reasoning on top of Mistral Medium 3 with RL alone, and we open-source Magistral Small (Apache 2.0) which further includes cold-start data from Magistral Medium. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:22:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Mistral-AI",
"",
""
],
[
":",
"",
""
],
[
"Rastogi",
"Abhinav",
""
],
[
"Jiang",
"Albert Q.",
""
],
[
"Lo",
"Andy",
""
],
[
"Berrada",
"Gabrielle",
""
],
[
"Lample",
"Guillaume",
""
],
[
"Rute",
"Jason",
""
],
[
"Barmentlo",
"Joep",
""
],
[
"Yadav",
"Karmesh",
""
],
[
"Khandelwal",
"Kartik",
""
],
[
"Chandu",
"Khyathi Raghavi",
""
],
[
"Blier",
"Léonard",
""
],
[
"Saulnier",
"Lucile",
""
],
[
"Dinot",
"Matthieu",
""
],
[
"Darrin",
"Maxime",
""
],
[
"Gupta",
"Neha",
""
],
[
"Soletskyi",
"Roman",
""
],
[
"Vaze",
"Sagar",
""
],
[
"Scao",
"Teven Le",
""
],
[
"Wang",
"Yihan",
""
],
[
"Yang",
"Adam",
""
],
[
"Liu",
"Alexander H.",
""
],
[
"Sablayrolles",
"Alexandre",
""
],
[
"Héliou",
"Amélie",
""
],
[
"Martin",
"Amélie",
""
],
[
"Ehrenberg",
"Andy",
""
],
[
"Agarwal",
"Anmol",
""
],
[
"Roux",
"Antoine",
""
],
[
"Darcet",
"Arthur",
""
],
[
"Mensch",
"Arthur",
""
],
[
"Bout",
"Baptiste",
""
],
[
"Rozière",
"Baptiste",
""
],
[
"De Monicault",
"Baudouin",
""
],
[
"Bamford",
"Chris",
""
],
[
"Wallenwein",
"Christian",
""
],
[
"Renaudin",
"Christophe",
""
],
[
"Lanfranchi",
"Clémence",
""
],
[
"Dabert",
"Darius",
""
],
[
"Mizelle",
"Devon",
""
],
[
"Casas",
"Diego de las",
""
],
[
"Chane-Sane",
"Elliot",
""
],
[
"Fugier",
"Emilien",
""
],
[
"Hanna",
"Emma Bou",
""
],
[
"Delerce",
"Gauthier",
""
],
[
"Guinet",
"Gauthier",
""
],
[
"Novikov",
"Georgii",
""
],
[
"Martin",
"Guillaume",
""
],
[
"Jaju",
"Himanshu",
""
],
[
"Ludziejewski",
"Jan",
""
],
[
"Chabran",
"Jean-Hadrien",
""
],
[
"Delignon",
"Jean-Malo",
""
],
[
"Studnia",
"Joachim",
""
],
[
"Amar",
"Jonas",
""
],
[
"Roberts",
"Josselin Somerville",
""
],
[
"Denize",
"Julien",
""
],
[
"Saxena",
"Karan",
""
],
[
"Jain",
"Kush",
""
],
[
"Zhao",
"Lingxiao",
""
],
[
"Martin",
"Louis",
""
],
[
"Gao",
"Luyu",
""
],
[
"Lavaud",
"Lélio Renard",
""
],
[
"Pellat",
"Marie",
""
],
[
"Guillaumin",
"Mathilde",
""
],
[
"Felardos",
"Mathis",
""
],
[
"Augustin",
"Maximilian",
""
],
[
"Seznec",
"Mickaël",
""
],
[
"Raghuraman",
"Nikhil",
""
],
[
"Duchenne",
"Olivier",
""
],
[
"Wang",
"Patricia",
""
],
[
"von Platen",
"Patrick",
""
],
[
"Saffer",
"Patryk",
""
],
[
"Jacob",
"Paul",
""
],
[
"Wambergue",
"Paul",
""
],
[
"Kurylowicz",
"Paula",
""
],
[
"Muddireddy",
"Pavankumar Reddy",
""
],
[
"Chagniot",
"Philomène",
""
],
[
"Stock",
"Pierre",
""
],
[
"Agrawal",
"Pravesh",
""
],
[
"Sauvestre",
"Romain",
""
],
[
"Delacourt",
"Rémi",
""
],
[
"Gandhi",
"Sanchit",
""
],
[
"Subramanian",
"Sandeep",
""
],
[
"Dalal",
"Shashwat",
""
],
[
"Gandhi",
"Siddharth",
""
],
[
"Ghosh",
"Soham",
""
],
[
"Mishra",
"Srijan",
""
],
[
"Aithal",
"Sumukh",
""
],
[
"Antoniak",
"Szymon",
""
],
[
"Schueller",
"Thibault",
""
],
[
"Lavril",
"Thibaut",
""
],
[
"Robert",
"Thomas",
""
],
[
"Wang",
"Thomas",
""
],
[
"Lacroix",
"Timothée",
""
],
[
"Nemychnikova",
"Valeriia",
""
],
[
"Paltz",
"Victor",
""
],
[
"Richard",
"Virgile",
""
],
[
"Li",
"Wen-Ding",
""
],
[
"Marshall",
"William",
""
],
[
"Zhang",
"Xuanyu",
""
],
[
"Tang",
"Yunhao",
""
]
] |
2506.10911 | Oguzhan Ersoy | Jari Kolehmainen, Nikolay Blagoev, John Donaghy, O\u{g}uzhan Ersoy and Christopher Nies | NoLoCo: No-all-reduce Low Communication Training Method for Large Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Training large language models is generally done via optimization methods on clusters containing tens of thousands of accelerators, communicating over a high-bandwidth interconnect. Scaling up these clusters is expensive and can become impractical, imposing limits on the size of models that can be trained. Several recent studies have proposed training methods that are less communication intensive, avoiding the need for a highly connected compute cluster. These state-of-the-art low communication training methods still employ a synchronization step for model parameters, which, when performed over all model replicas, can become costly on a low-bandwidth network.
In this work, we propose a novel optimization method, NoLoCo, that does not explicitly synchronize all model parameters during training and, as a result, does not require any collective communication. NoLoCo implicitly synchronizes model weights via a novel variant of the Nesterov momentum optimizer by partially averaging model weights with a randomly selected other one. We provide both a theoretical convergence analysis for our proposed optimizer as well as empirical results from language model training.
We benchmark NoLoCo on a wide range of accelerator counts and model sizes, between 125M to 6.8B parameters. Our method requires significantly less communication overhead than fully sharded data parallel training or even widely used low communication training method, DiLoCo. The synchronization step itself is estimated to be one magnitude faster than the all-reduce used in DiLoCo for few hundred accelerators training over the internet. We also do not have any global blocking communication that reduces accelerator idling time. Compared to DiLoCo, we also observe up to $4\%$ faster convergence rate with wide range of model sizes and accelerator counts. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:23:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kolehmainen",
"Jari",
""
],
[
"Blagoev",
"Nikolay",
""
],
[
"Donaghy",
"John",
""
],
[
"Ersoy",
"Oğuzhan",
""
],
[
"Nies",
"Christopher",
""
]
] |
2506.10912 | Fei Lin | Fei Lin, Ziyang Gong, Cong Wang, Yonglin Tian, Tengchao Zhang, Xue Yang, Gen Luo, Fei-Yue Wang | Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification? | null | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Toxicity remains a leading cause of early-stage drug development failure. Despite advances in molecular design and property prediction, the task of molecular toxicity repair - generating structurally valid molecular alternatives with reduced toxicity - has not yet been systematically defined or benchmarked. To fill this gap, we introduce ToxiMol, the first benchmark task for general-purpose Multimodal Large Language Models (MLLMs) focused on molecular toxicity repair. We construct a standardized dataset covering 11 primary tasks and 560 representative toxic molecules spanning diverse mechanisms and granularities. We design a prompt annotation pipeline with mechanism-aware and task-adaptive capabilities, informed by expert toxicological knowledge. In parallel, we propose an automated evaluation framework, ToxiEval, which integrates toxicity endpoint prediction, synthetic accessibility, drug-likeness, and structural similarity into a high-throughput evaluation chain for repair success. We systematically assess nearly 30 mainstream general-purpose MLLMs and design multiple ablation studies to analyze key factors such as evaluation criteria, candidate diversity, and failure attribution. Experimental results show that although current MLLMs still face significant challenges on this task, they begin to demonstrate promising capabilities in toxicity understanding, semantic constraint adherence, and structure-aware molecule editing. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:25:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lin",
"Fei",
""
],
[
"Gong",
"Ziyang",
""
],
[
"Wang",
"Cong",
""
],
[
"Tian",
"Yonglin",
""
],
[
"Zhang",
"Tengchao",
""
],
[
"Yang",
"Xue",
""
],
[
"Luo",
"Gen",
""
],
[
"Wang",
"Fei-Yue",
""
]
] |
2506.10913 | Ashley Samuelson | Ashley Samuelson, Andrew K. Hirsch, Ethan Cecchetti | Choreographic Quick Changes: First-Class Location (Set) Polymorphism | In submission to OOPSLA 2025 | null | null | null | cs.PL | http://creativecommons.org/licenses/by/4.0/ | Choreographic programming is a promising new paradigm for programming concurrent systems where a developer writes a single centralized program that compiles to individual programs for each node. Existing choreographic languages, however, lack critical features integral to modern systems, like the ability of one node to dynamically compute who should perform a computation and send that decision to others. This work addresses this gap with $\lambda_{QC}$, the first typed choreographic language with \emph{first class process names} and polymorphism over both types and (sets of) locations. $\lambda_{QC}$ also improves expressive power over previous work by supporting algebraic and recursive data types as well as multiply-located values. We formalize and mechanically verify our results in Rocq, including the standard choreographic guarantee of deadlock freedom. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:27:48 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Samuelson",
"Ashley",
""
],
[
"Hirsch",
"Andrew K.",
""
],
[
"Cecchetti",
"Ethan",
""
]
] |
2506.10914 | Yuchen Ma | Yuchen Ma, Dennis Frauen, Emil Javurek, Stefan Feuerriegel | Foundation Models for Causal Inference via Prior-Data Fitted Networks | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Prior-data fitted networks (PFNs) have recently been proposed as a promising way to train tabular foundation models. PFNs are transformers that are pre-trained on synthetic data generated from a prespecified prior distribution and that enable Bayesian inference through in-context learning. In this paper, we introduce CausalFM, a comprehensive framework for training PFN-based foundation models in various causal inference settings. First, we formalize the construction of Bayesian priors for causal inference based on structural causal models (SCMs) in a principled way and derive necessary criteria for the validity of such priors. Building on this, we propose a novel family of prior distributions using causality-inspired Bayesian neural networks that enable CausalFM to perform Bayesian causal inference in various settings, including back-door, front-door, and instrumental variable adjustment. Finally, we instantiate CausalFM and explicitly train a foundation model for estimating conditional average treatment effects (CATEs) using back-door adjustment. We show that CausalFM performs competitively for CATE estimation using various synthetic and semi-synthetic benchmarks. In sum, our framework can be used as a general recipe to train foundation models for various causal inference settings. In contrast to the current state-of-the-art in causal inference, CausalFM offers a novel paradigm with the potential to fundamentally change how practitioners perform causal inference in medicine, economics, and other disciplines. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:29:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ma",
"Yuchen",
""
],
[
"Frauen",
"Dennis",
""
],
[
"Javurek",
"Emil",
""
],
[
"Feuerriegel",
"Stefan",
""
]
] |
2506.10915 | Gengwei Zhang | Jiancheng Huang, Gengwei Zhang, Zequn Jie, Siyu Jiao, Yinlong Qian, Ling Chen, Yunchao Wei, Lin Ma | M4V: Multi-Modal Mamba for Text-to-Video Generation | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-to-video generation has significantly enriched content creation and holds the potential to evolve into powerful world simulators. However, modeling the vast spatiotemporal space remains computationally demanding, particularly when employing Transformers, which incur quadratic complexity in sequence processing and thus limit practical applications. Recent advancements in linear-time sequence modeling, particularly the Mamba architecture, offer a more efficient alternative. Nevertheless, its plain design limits its direct applicability to multi-modal and spatiotemporal video generation tasks. To address these challenges, we introduce M4V, a Multi-Modal Mamba framework for text-to-video generation. Specifically, we propose a multi-modal diffusion Mamba (MM-DiM) block that enables seamless integration of multi-modal information and spatiotemporal modeling through a multi-modal token re-composition design. As a result, the Mamba blocks in M4V reduce FLOPs by 45% compared to the attention-based alternative when generating videos at 768$\times$1280 resolution. Additionally, to mitigate the visual quality degradation in long-context autoregressive generation processes, we introduce a reward learning strategy that further enhances per-frame visual realism. Extensive experiments on text-to-video benchmarks demonstrate M4V's ability to produce high-quality videos while significantly lowering computational costs. Code and models will be publicly available at https://huangjch526.github.io/M4V_project. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:29:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Jiancheng",
""
],
[
"Zhang",
"Gengwei",
""
],
[
"Jie",
"Zequn",
""
],
[
"Jiao",
"Siyu",
""
],
[
"Qian",
"Yinlong",
""
],
[
"Chen",
"Ling",
""
],
[
"Wei",
"Yunchao",
""
],
[
"Ma",
"Lin",
""
]
] |
2506.10916 | Meredith VandeHaar | Meredith VandeHaar, M. Clinch, I. Yilmaz, M.A. Rahman, Y. Xiao, F. Dogany, H.M. Alazab, A. Nassar, Z. Akkus, B. Dangott | Semi-Automated Quality Assurance in Digital Pathology: Tile Classification Approach | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Quality assurance is a critical but underexplored area in digital pathology, where even minor artifacts can have significant effects. Artifacts have been shown to negatively impact the performance of AI diagnostic models. In current practice, trained staff manually review digitized images prior to release of these slides to pathologists which are then used to render a diagnosis. Conventional image processing approaches, provide a foundation for detecting artifacts on digital pathology slides. However, current tools do not leverage deep learning, which has the potential to improve detection accuracy and scalability. Despite these advancements, methods for quality assurance in digital pathology remain limited, presenting a gap for innovation.
We propose an AI algorithm designed to screen digital pathology slides by analyzing tiles and categorizing them into one of 10 predefined artifact types or as background. This algorithm identifies and localizes artifacts, creating a map that highlights regions of interest. By directing human operators to specific tiles affected by artifacts, the algorithm minimizes the time and effort required to manually review entire slides for quality issues.
From internal archives and The Cancer Genome Atlas, 133 whole slide images were selected and 10 artifacts were annotated using an internally developed software ZAPP (Mayo Clinic, Jacksonville, FL). Ablation study of multiple models at different tile sizes and magnification was performed. InceptionResNet was selected. Single artifact models were trained and tested, followed by a limited multiple instance model with artifacts that performed well together (chatter, fold, and pen). From the results of this study we suggest a hybrid design for artifact screening composed of both single artifact binary models as well as multiple instance models to optimize detection of each artifact. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:30:34 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"VandeHaar",
"Meredith",
""
],
[
"Clinch",
"M.",
""
],
[
"Yilmaz",
"I.",
""
],
[
"Rahman",
"M. A.",
""
],
[
"Xiao",
"Y.",
""
],
[
"Dogany",
"F.",
""
],
[
"Alazab",
"H. M.",
""
],
[
"Nassar",
"A.",
""
],
[
"Akkus",
"Z.",
""
],
[
"Dangott",
"B.",
""
]
] |
2506.10917 | Allic Sivaramakrishnan | Mathew W. Bub and Allic Sivaramakrishnan | Correlation functions of von Neumann entropy | 17 pages | null | null | null | hep-th cond-mat.stat-mech cond-mat.str-el quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this note, we study two-point correlation functions of modular Hamiltonians. We show that in general quantum systems, these correlators obey properties similar to those of von Neumann entropy and capacity of entanglement, both of which are special cases of these correlators. Then we specialize to two spacelike-separated spherical subregions in conformal field theories. We present direct computations of the vacuum two-point function that confirm its equivalence to the stress-tensor conformal block. We explore the two-point function in various kinematic regimes, including imaginary time separation between subsystems. The material presented in this note may be useful for further studying modular Hamiltonian correlators in generic quantum systems and in conformal field theories, including those with holographic duals. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:30:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Bub",
"Mathew W.",
""
],
[
"Sivaramakrishnan",
"Allic",
""
]
] |
2506.10918 | Morris Yau | Morris Yau, Sharut Gupta, Valerie Engelmayer, Kazuki Irie, Stefanie Jegelka, Jacob Andreas | Sequential-Parallel Duality in Prefix Scannable Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Modern neural sequence models are designed to meet the dual mandate of parallelizable training and fast sequential inference. Recent developments have given rise to various models, such as Gated Linear Attention (GLA) and Mamba, that achieve such ``sequential-parallel duality.'' This raises a natural question: can we characterize the full class of neural sequence models that support near-constant-time parallel evaluation and linear-time, constant-space sequential inference? We begin by describing a broad class of such models -- state space models -- as those whose state updates can be computed using the classic parallel prefix scan algorithm with a custom associative aggregation operator. We then define a more general class, Prefix-Scannable Models (PSMs), by relaxing the state aggregation operator to allow arbitrary (potentially non-associative) functions such as softmax attention. This generalization unifies many existing architectures, including element-wise RNNs (e.g., Mamba) and linear transformers (e.g., GLA, Mamba2, mLSTM), while also introducing new models with softmax-like operators that achieve O(1) amortized compute per token and log(N) memory for sequence length N. We empirically evaluate such models on illustrative small-scale language modeling and canonical synthetic tasks, including state tracking and associative recall. Empirically, we find that PSMs retain the expressivity of transformer-based architectures while matching the inference efficiency of state space models -- in some cases exhibiting better length generalization than either. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:32:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yau",
"Morris",
""
],
[
"Gupta",
"Sharut",
""
],
[
"Engelmayer",
"Valerie",
""
],
[
"Irie",
"Kazuki",
""
],
[
"Jegelka",
"Stefanie",
""
],
[
"Andreas",
"Jacob",
""
]
] |
2506.10919 | Adam Shaw | Adam L. Shaw, Anna Soper, Danial Shadmany, Aishwarya Kumar, Lukas Palm, Da-Yeon Koh, Vassilios Kaxiras, Lavanya Taneja, Matt Jaffe, David I. Schuster, Jonathan Simon | A cavity array microscope for parallel single-atom interfacing | A.L.S., A.S., and D.S. contributed equally | null | null | null | quant-ph cond-mat.quant-gas physics.atom-ph physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neutral atom arrays and optical cavity QED systems have developed in parallel as central pillars of modern experimental quantum science. While each platform has demonstrated exceptional capabilities-such as high-fidelity quantum logic in atom arrays, and strong light-matter coupling in cavities-their combination holds promise for realizing fast and non-destructive atom measurement, building large-scale quantum networks, and engineering hybrid atom-photon Hamiltonians. However, to date, experiments integrating the two platforms have been limited to interfacing the entire atom array with one global cavity mode, a configuration that constrains addressability, parallelism, and scalability. Here we introduce the cavity array microscope, an experimental platform where each individual atom is strongly coupled to its own individual cavity across a two-dimensional array of over 40 modes. Our approach requires no nanophotonic elements, and instead uses a new free-space cavity geometry with intra-cavity lenses to realize above-unity peak cooperativity with micron-scale mode waists and spacings, compatible with typical atom array length scales while keeping atoms far from dielectric surfaces. We achieve homogeneous atom-cavity coupling, and show fast, non-destructive, parallel readout on millisecond timescales, including cavity-resolved readout into a fiber array as a proof-of-principle for future networking applications. This platform is species-agnostic and scalable, and we expect key metrics to further improve in a next-generation realization anticipated to be compatible with glass-cell-based experiments. Our work unlocks, for the first time, the regime of many-cavity QED, and opens an unexplored frontier of large-scale quantum networking with atom arrays. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:32:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Shaw",
"Adam L.",
""
],
[
"Soper",
"Anna",
""
],
[
"Shadmany",
"Danial",
""
],
[
"Kumar",
"Aishwarya",
""
],
[
"Palm",
"Lukas",
""
],
[
"Koh",
"Da-Yeon",
""
],
[
"Kaxiras",
"Vassilios",
""
],
[
"Taneja",
"Lavanya",
""
],
[
"Jaffe",
"Matt",
""
],
[
"Schuster",
"David I.",
""
],
[
"Simon",
"Jonathan",
""
]
] |
2506.10920 | Or Shafran | Or Shafran, Atticus Geiger, Mor Geva | Decomposing MLP Activations into Interpretable Features via Semi-Nonnegative Matrix Factorization | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | A central goal for mechanistic interpretability has been to identify the right units of analysis in large language models (LLMs) that causally explain their outputs. While early work focused on individual neurons, evidence that neurons often encode multiple concepts has motivated a shift toward analyzing directions in activation space. A key question is how to find directions that capture interpretable features in an unsupervised manner. Current methods rely on dictionary learning with sparse autoencoders (SAEs), commonly trained over residual stream activations to learn directions from scratch. However, SAEs often struggle in causal evaluations and lack intrinsic interpretability, as their learning is not explicitly tied to the computations of the model. Here, we tackle these limitations by directly decomposing MLP activations with semi-nonnegative matrix factorization (SNMF), such that the learned features are (a) sparse linear combinations of co-activated neurons, and (b) mapped to their activating inputs, making them directly interpretable. Experiments on Llama 3.1, Gemma 2 and GPT-2 show that SNMF derived features outperform SAEs and a strong supervised baseline (difference-in-means) on causal steering, while aligning with human-interpretable concepts. Further analysis reveals that specific neuron combinations are reused across semantically-related features, exposing a hierarchical structure in the MLP's activation space. Together, these results position SNMF as a simple and effective tool for identifying interpretable features and dissecting concept representations in LLMs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:33:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Shafran",
"Or",
""
],
[
"Geiger",
"Atticus",
""
],
[
"Geva",
"Mor",
""
]
] |
2506.10921 | Luca Colagrande | Luca Colagrande, Lorenzo Leone, Maximilian Coco, Andrei Deaconeasa, Luca Benini | Towards Zero-Stall Matrix Multiplication on Energy-Efficient RISC-V Clusters for Machine Learning Acceleration | 7 pages, 5 figures, 2 tables. Accepted at ISLPED 2025 | null | null | null | cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The growing computational demands of machine learning (ML) workloads have driven the design of ML accelerators aiming at an optimal tradeoff between efficiency and flexibility. A widely explored architecture for flexible ML accelerators is based on clusters of lightweight instruction processors sharing multi-banked L1 memory, augmented with specialized instruction extensions for key ML-related computations, such as matrix multiplication (matmul). However, instruction extensions should be coupled with microarchitectural optimizations that remove inefficiencies due to control flow (loop handling) and memory access, without drastically increasing processor complexity. Moving from a state-of-the-art (SoA) ML accelerator cluster based on RISC-V processors, we propose a low-overhead optimized microarchitecture that eliminates these inefficiencies almost entirely while retaining programmability. We introduce "zero-overhead loop nests" to remove control overheads, and a "zero-conflict memory subsystem", leveraging a novel double-buffering-aware interconnect, to eliminate bank conflicts in L1 memory. With these enhancements, we attain near-ideal utilizations between 96.1% and 99.4%, achieving 11% performance and 8% energy efficiency improvements over the baseline SoA RISC-V cluster. We demonstrate comparable utilizations and performance to a specialized SoA accelerator, with only 12% difference in energy efficiency, while providing a fully-programmable general-purpose solution supporting a significantly wider range of workloads. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:34:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Colagrande",
"Luca",
""
],
[
"Leone",
"Lorenzo",
""
],
[
"Coco",
"Maximilian",
""
],
[
"Deaconeasa",
"Andrei",
""
],
[
"Benini",
"Luca",
""
]
] |
2506.10922 | Adam Karvonen | Adam Karvonen, Samuel Marks | Robustly Improving LLM Fairness in Realistic Settings via Interpretability | null | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are increasingly deployed in high-stakes hiring applications, making decisions that directly impact people's careers and livelihoods. While prior studies suggest simple anti-bias prompts can eliminate demographic biases in controlled evaluations, we find these mitigations fail when realistic contextual details are introduced. We address these failures through internal bias mitigation: by identifying and neutralizing sensitive attribute directions within model activations, we achieve robust bias reduction across all tested scenarios. Across leading commercial (GPT-4o, Claude 4 Sonnet, Gemini 2.5 Flash) and open-source models (Gemma-2 27B, Gemma-3, Mistral-24B), we find that adding realistic context such as company names, culture descriptions from public careers pages, and selective hiring constraints (e.g.,``only accept candidates in the top 10\%") induces significant racial and gender biases (up to 12\% differences in interview rates). When these biases emerge, they consistently favor Black over White candidates and female over male candidates across all tested models and scenarios. Moreover, models can infer demographics and become biased from subtle cues like college affiliations, with these biases remaining invisible even when inspecting the model's chain-of-thought reasoning. To address these limitations, our internal bias mitigation identifies race and gender-correlated directions and applies affine concept editing at inference time. Despite using directions from a simple synthetic dataset, the intervention generalizes robustly, consistently reducing bias to very low levels (typically under 1\%, always below 2.5\%) while largely maintaining model performance. Our findings suggest that practitioners deploying LLMs for hiring should adopt more realistic evaluation methodologies and consider internal mitigation strategies for equitable outcomes. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:34:38 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Karvonen",
"Adam",
""
],
[
"Marks",
"Samuel",
""
]
] |
2506.10923 | Xili Yi | Xili Yi, Nima Fazeli | Vib2Move: In-Hand Object Reconfiguration via Fingertip Micro-Vibrations | 11 pages, 12 figures | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce Vib2Move, a novel approach for in-hand object reconfiguration that uses fingertip micro-vibrations and gravity to precisely reposition planar objects. Our framework comprises three key innovations. First, we design a vibration-based actuator that dynamically modulates the effective finger-object friction coefficient, effectively emulating changes in gripping force. Second, we derive a sliding motion model for objects clamped in a parallel gripper with two symmetric, variable-friction contact patches. Third, we propose a motion planner that coordinates end-effector finger trajectories and fingertip vibrations to achieve the desired object pose. In real-world trials, Vib2Move consistently yields final positioning errors below 6 mm, demonstrating reliable, high-precision manipulation across a variety of planar objects. For more results and information, please visit https://vib2move.github.io. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:34:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yi",
"Xili",
""
],
[
"Fazeli",
"Nima",
""
]
] |
2506.10924 | Huy Nguyen Quang | Quang Huy Nguyen, Phuong Cuc Hoang, Van Chien Le, and Thi Thanh Mai Ta | A space-time interface-fitted method for moving-subdomain distributed control problems with energy regularization | null | null | null | null | math.OC cs.NA math.NA | http://creativecommons.org/licenses/by-sa/4.0/ | This paper investigates a space-time interface-fitted approximation of a moving-interface optimal control problem with energy regularization. We reformulate the optimality conditions into a variational problem involving both the state and adjoint. This problem is shown to be equivalent to our optimal control problem. Based on fully unstructured, space-time interface-fitted meshes, we propose and analyze a Petrov-Galerkin approximation of the problem. An optimal error estimate with respect to a discrete norm is established under a specific regularity assumption on the state and adjoint. Several numerical results are presented to corroborate our theoretical results. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:35:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Nguyen",
"Quang Huy",
""
],
[
"Hoang",
"Phuong Cuc",
""
],
[
"Le",
"Van Chien",
""
],
[
"Ta",
"Thi Thanh Mai",
""
]
] |
2506.10925 | Eduardo Baena | Eduardo Baena, Paolo Testolina, Michele Polese, Sergi Aliaga, Andrew Benincasa, Dimitrios Koutsonikolas, Josep Jornet, Tommaso Melodia | Agentic Semantic Control for Autonomous Wireless Space Networks: Extending Space-O-RAN with MCP-Driven Distributed Intelligence | Lunar Surface Innovation Consortium 2025 Spring Meeting, May 20-22 | null | null | null | cs.NI cs.AI cs.HC cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Lunar surface operations impose stringent requirements on wireless communication systems, including autonomy, robustness to disruption, and the ability to adapt to environmental and mission-driven context. While Space-O-RAN provides a distributed orchestration model aligned with 3GPP standards, its decision logic is limited to static policies and lacks semantic integration. We propose a novel extension incorporating a semantic agentic layer enabled by the Model Context Protocol (MCP) and Agent-to-Agent (A2A) communication protocols, allowing context-aware decision making across real-time, near-real-time, and non-real-time control layers. Distributed cognitive agents deployed in rovers, landers, and lunar base stations implement wireless-aware coordination strategies, including delay-adaptive reasoning and bandwidth-aware semantic compression, while interacting with multiple MCP servers to reason over telemetry, locomotion planning, and mission constraints. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:35:36 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Baena",
"Eduardo",
""
],
[
"Testolina",
"Paolo",
""
],
[
"Polese",
"Michele",
""
],
[
"Aliaga",
"Sergi",
""
],
[
"Benincasa",
"Andrew",
""
],
[
"Koutsonikolas",
"Dimitrios",
""
],
[
"Jornet",
"Josep",
""
],
[
"Melodia",
"Tommaso",
""
]
] |
2506.10926 | Ivan Christov | SungGyu Chun, Ivan C. Christov, Jie Feng | Experimental investigation of the flow rate--pressure drop relation of a viscoelastic Boger fluid in a deformable channel | 7 pages, 4 figures | null | null | null | physics.flu-dyn cond-mat.soft | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the steady flow-induced deformation between an incompressible non-Newtonian fluid and a three-dimensional (3D) deformable channel. Specifically, we provide a comprehensive experimental--theoretical framework for such flows of constant-viscosity viscoelastic (i.e., Boger) fluids, which allows us to quantify the influence of the fluid's viscoelasticity on the flow rate--pressure drop ($q-\Delta p$) relation. For a flow-rate-controlled regime and weakly viscoelastic flow, we find excellent agreement between the theoretical prediction and experimental results, specifically providing an experimental demonstration of how both wall compliance and fluid viscoelasticity decrease the pressure drop. To definitively demonstrate the coupled effect of wall compliance and fluid viscoelasticity (change of cross-sectional area and curving of streamlines), we perform experiments in an equivalent rigid channel, showing that both the Boger fluid and a viscosity-matched Newtonian fluid have the same pressure drop. Thus, the experimentally verified theory for the $q-\Delta p$ relation of weakly viscoelastic flows of Boger fluids in compliant channels can now be used for microfluidic system design involving complex fluids. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:35:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chun",
"SungGyu",
""
],
[
"Christov",
"Ivan C.",
""
],
[
"Feng",
"Jie",
""
]
] |
2506.10927 | Teis Arets | T. T. J. E. Arets, G. Perugia, M. Houben, W.A. IJsselsteijn | The Role of Generative AI in Facilitating Social Interactions: A Scoping Review | Preprint version of a manuscript submitted to ACM Transactions on Computer-Human Interaction (TOCHI), under review. 39 pages, 4 figures | null | null | null | cs.HC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reduced social connectedness increasingly poses a threat to mental health, life expectancy, and general well-being. Generative AI (GAI) technologies, such as large language models (LLMs) and image generation tools, are increasingly integrated into applications aimed at enhancing human social experiences. Despite their growing presence, little is known about how these technologies influence social interactions. This scoping review investigates how GAI-based applications are currently designed to facilitate social interaction, what forms of social engagement they target, and which design and evaluation methodologies designers use to create and evaluate them. Through an analysis of 30 studies published since 2020, we identify key trends in application domains including storytelling, socio-emotional skills training, reminiscence, collaborative learning, music making, and general conversation. We highlight the role of participatory and co-design approaches in fostering both effective technology use and social engagement, while also examining socio-ethical concerns such as cultural bias and accessibility. This review underscores the potential of GAI to support dynamic and personalized interactions, but calls for greater attention to equitable design practices and inclusive evaluation strategies. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:37:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Arets",
"T. T. J. E.",
""
],
[
"Perugia",
"G.",
""
],
[
"Houben",
"M.",
""
],
[
"IJsselsteijn",
"W. A.",
""
]
] |
2506.10928 | Yi Chung | Yi Chung | Two coincidences are a clue: Probing a GeV-scale dark QCD sector | 7 pages, 3 figures | null | null | null | hep-ph astro-ph.CO astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The similarity between dark matter and baryon energy densities suggests the existence of a dark sector analogous to the visible QCD sector. In addition, small-scale structure issues can be addressed by dark matter self-interactions with a cross section comparable to that of QCD. Both observations point to a GeV-scale dark QCD sector. In this letter, guided by these two coincidences, we investigate how a chiral dark QCD model with a MeV-scale dark photon can resolve the core-cusp problem. A potential third coincidence, associated with the latest measurement of $N_{\rm eff}$ being lower than the Standard Model prediction, is also discussed. Current constraints leave a finite window in the parameter space, which can be tested by future experiments such as Gamma Factory. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:37:33 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chung",
"Yi",
""
]
] |
2506.10929 | Fabio Demaria | Fabio Demaria | On feature selection in double-imbalanced data settings: a Random Forest approach | Working paper | null | null | null | stat.ME cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Feature selection is a critical step in high-dimensional classification tasks, particularly under challenging conditions of double imbalance, namely settings characterized by both class imbalance in the response variable and dimensional asymmetry in the data $(n \gg p)$. In such scenarios, traditional feature selection methods applied to Random Forests (RF) often yield unstable or misleading importance rankings. This paper proposes a novel thresholding scheme for feature selection based on minimal depth, which exploits the tree topology to assess variable relevance. Extensive experiments on simulated and real-world datasets demonstrate that the proposed approach produces more parsimonious and accurate subsets of variables compared to conventional minimal depth-based selection. The method provides a practical and interpretable solution for variable selection in RF under double imbalance conditions. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:37:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Demaria",
"Fabio",
""
]
] |
2506.10930 | Thanathai Lertpetchpun | Thanathai Lertpetchpun, Tiantian Feng, Dani Byrd, Shrikanth Narayanan | Developing a High-performance Framework for Speech Emotion Recognition in Naturalistic Conditions Challenge for Emotional Attribute Prediction | null | Accepeted to Interspeech2025 | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech emotion recognition (SER) in naturalistic conditions presents a significant challenge for the speech processing community. Challenges include disagreement in labeling among annotators and imbalanced data distributions. This paper presents a reproducible framework that achieves superior (top 1) performance in the Emotion Recognition in Naturalistic Conditions Challenge (IS25-SER Challenge) - Task 2, evaluated on the MSP-Podcast dataset. Our system is designed to tackle the aforementioned challenges through multimodal learning, multi-task learning, and imbalanced data handling. Specifically, our best system is trained by adding text embeddings, predicting gender, and including ``Other'' (O) and ``No Agreement'' (X) samples in the training set. Our system's results secured both first and second places in the IS25-SER Challenge, and the top performance was achieved by a simple two-system ensemble. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:38:06 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lertpetchpun",
"Thanathai",
""
],
[
"Feng",
"Tiantian",
""
],
[
"Byrd",
"Dani",
""
],
[
"Narayanan",
"Shrikanth",
""
]
] |
2506.10931 | Konstantina Koliogeorgi Koliogeorgi | Melina Soysal, Konstantina Koliogeorgi, Can Firtina, Nika Mansouri Ghiasi, Rakesh Nadig, Haiyu Mayo, Geraldo F. Oliveira, Yu Liang, Klea Zambaku, Mohammad Sadrosadati, Onur Mutlu | MARS: Processing-In-Memory Acceleration of Raw Signal Genome Analysis Inside the Storage Subsystem | null | null | 10.1145/3721145.3730428 | null | cs.AR q-bio.GN | http://creativecommons.org/licenses/by/4.0/ | Raw signal genome analysis (RSGA) has emerged as a promising approach to enable real-time genome analysis by directly analyzing raw electrical signals. However, rapid advancements in sequencing technologies make it increasingly difficult for software-based RSGA to match the throughput of raw signal generation. This paper demonstrates that while hardware acceleration techniques can significantly accelerate RSGA, the high volume of genomic data shifts the performance and energy bottleneck from computation to I/O data movement. As sequencing throughput increases, I/O overhead becomes the main contributor to both runtime and energy consumption. Therefore, there is a need to design a high-performance, energy-efficient system for RSGA that can both alleviate the data movement bottleneck and provide large acceleration capabilities. We propose MARS, a storage-centric system that leverages the heterogeneous resources within modern storage systems (e.g., storage-internal DRAM, storage controller, flash chips) alongside their large storage capacity to tackle both data movement and computational overheads of RSGA in an area-efficient and low-cost manner. MARS accelerates RSGA through a novel hardware/software co-design approach. First, MARS modifies the RSGA pipeline via two filtering mechanisms and a quantization scheme, reducing hardware demands and optimizing for in-storage execution. Second, MARS accelerates the RSGA steps directly within the storage by leveraging both Processing-Near-Memory and Processing-Using-Memory paradigms. Third, MARS orchestrates the execution of all steps to fully exploit in-storage parallelism and minimize data movement. Our evaluation shows that MARS outperforms basecalling-based software and hardware-accelerated state-of-the-art read mapping pipelines by 93x and 40x, on average across different datasets, while reducing their energy consumption by 427x and 72x. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:38:12 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Soysal",
"Melina",
""
],
[
"Koliogeorgi",
"Konstantina",
""
],
[
"Firtina",
"Can",
""
],
[
"Ghiasi",
"Nika Mansouri",
""
],
[
"Nadig",
"Rakesh",
""
],
[
"Mayo",
"Haiyu",
""
],
[
"Oliveira",
"Geraldo F.",
""
],
[
"Liang",
"Yu",
""
],
[
"Zambaku",
"Klea",
""
],
[
"Sadrosadati",
"Mohammad",
""
],
[
"Mutlu",
"Onur",
""
]
] |
2506.10932 | Jiaying Liu | Jiaying Lizzy Liu, Yan Zhang | Video-Mediated Emotion Disclosure: A Study of Mental Health Vlogging by People with Schizophrenia on YouTube | 10 pages | ASIS&T 2025 | null | null | cs.HC cs.CY cs.MM | http://creativecommons.org/licenses/by/4.0/ | Individuals with schizophrenia frequently experience intense emotions and often turn to vlogging as a medium for emotional expression. While previous research has predominantly focused on text based disclosure, little is known about how individuals construct narratives around emotions and emotional experiences in video blogs. Our study addresses this gap by analyzing 200 YouTube videos created by individuals with schizophrenia. Drawing on media research and self presentation theories, we developed a visual analysis framework to disentangle these videos. Our analysis revealed diverse practices of emotion disclosure through both verbal and visual channels, highlighting the dynamic interplay between these modes of expression. We found that the deliberate construction of visual elements, including environmental settings and specific aesthetic choices, appears to foster more supportive and engaged viewer responses. These findings underscore the need for future large scale quantitative research examining how visual features shape video mediated communication on social media platforms. Such investigations would inform the development of care centered video sharing platforms that better support individuals managing illness experiences. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:39:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liu",
"Jiaying Lizzy",
""
],
[
"Zhang",
"Yan",
""
]
] |
2506.10933 | Zhenhong Li | Ziwen Wang, Yue Zhang, Zhiqiang Zhang, Sheng Quan Xie, Alexander Lanzon, William P. Heath, Zhenhong Li | Instance-Based Transfer Learning with Similarity-Aware Subject Selection for Cross-Subject SSVEP-Based BCIs | IEEE Journal of Biomedical and Health Informatics | null | 10.1109/JBHI.2025.3577813 | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) can achieve high recognition accuracy with sufficient training data. Transfer learning presents a promising solution to alleviate data requirements for the target subject by leveraging data from source subjects; however, effectively addressing individual variability among both target and source subjects remains a challenge. This paper proposes a novel transfer learning framework, termed instance-based task-related component analysis (iTRCA), which leverages knowledge from source subjects while considering their individual contributions. iTRCA extracts two types of features: (1) the subject-general feature, capturing shared information between source and target subjects in a common latent space, and (2) the subject-specific feature, preserving the unique characteristics of the target subject. To mitigate negative transfer, we further design an enhanced framework, subject selection-based iTRCA (SS-iTRCA), which integrates a similarity-based subject selection strategy to identify appropriate source subjects for transfer based on their task-related components (TRCs). Comparative evaluations on the Benchmark, BETA, and a self-collected dataset demonstrate the effectiveness of the proposed iTRCA and SS-iTRCA frameworks. This study provides a potential solution for developing high-performance SSVEP-based BCIs with reduced target subject data. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 17:40:20 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Ziwen",
""
],
[
"Zhang",
"Yue",
""
],
[
"Zhang",
"Zhiqiang",
""
],
[
"Xie",
"Sheng Quan",
""
],
[
"Lanzon",
"Alexander",
""
],
[
"Heath",
"William P.",
""
],
[
"Li",
"Zhenhong",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.