Search

Generating high-quality 3D digital assets often requires expert knowledge of complex design tools. We introduce Specialized Generative Primitives, a generative framework that allows non-expert users to author high-quality 3D scenes in a seamless, lightweight, and controllable manner. Each primitive is an efficient generative model that captures the distribution of a single exemplar from the real world. With our framework, users capture a video of an environment, which we turn into a high-quality and explicit appearance model thanks to 3D Gaussian Splatting. Users then select regions of interest guided by semantically-aware features. To create a generative primitive, we adapt Generative Cellular Automata to single-exemplar training and controllable generation. We decouple the generative task from the appearance model by operating on sparse voxels and we recover a high-quality output with a subsequent sparse patch consistency step. Each primitive can be trained within 10 minutes and used to author new scenes interactively in a fully compositional manner. We showcase interactive sessions where various primitives are extracted from real-world scenes and controlled to create 3D assets and scenes in a few minutes. We also demonstrate additional capabilities of our primitives: handling various 3D representations to control generation, transferring appearances, and editing geometries.

We present improvements to the pointing accuracy of the South Pole Telescope (SPT) using machine learning. The ability of the SPT to point accurately at the sky is limited by its structural imperfections, which are impacted by the extreme weather at the South Pole. Pointing accuracy is particularly important during SPT participation in observing campaigns with the Event Horizon Telescope (EHT), which requires stricter accuracy than typical observations with the SPT. We compile a training dataset of historical observations of astronomical sources made with the SPT-3G and EHT receivers on the SPT. We train two XGBoost models to learn a mapping from current weather conditions to two telescope drive control arguments -- one which corrects for errors in azimuth and the other for errors in elevation. Our trained models achieve root mean squared errors on withheld test data of $2.14''$ in cross-elevation and $3.57''$ in elevation, well below our goal of $5''$ along each axis. We deploy our models on the telescope control system and perform further in situ test observations during the EHT observing campaign in 2024 April. Our models result in significantly improved pointing accuracy: for sources within the range of input variables where the models are best trained, average combined pointing error improved 33%, from $15.9''$ to $10.6''$. These improvements, while significant, fall shy of our ultimate goal, but they serve as a proof of concept for the development of future models. Planned upgrades to the EHT receiver on the SPT will necessitate even stricter pointing accuracy which will be achievable with our methods., Comment: 23 pages, 9 figures, to be submitted to Journal of Astronomical Instrumentation (JAI)

We investigate the evolution of active galactic nucleus jets on kiloparsec-scales due to their interaction with the clumpy interstellar medium (ISM) of the host galaxy and, subsequently, the surrounding circumgalactic environment. Hydrodynamic simulations of this jet-environment interaction are presented for a range of jet kinetic powers, peak densities of the multiphase ISM, and scale radii of the larger-scale environment -- characteristic of either a galaxy cluster or poor group. Synthetic radio images are generated by considering the combination of synchrotron radiation from the jet plasma and free-free absorption from the multiphase ISM. We find that jet propagation is slowed by interactions with a few very dense clouds in the host galaxy ISM, producing asymmetries in lobe length and brightness which persist to scales of tens of kpc for poor group environments. The classification of kiloparsec-scale jets is highly dependent on surface brightness sensitivity and resolution. Our simulations of young active sources can appear as restarted sources, showing double-double lobe morphology, high core prominence (CP > 0.1), and the expected radio spectra for both the inner- and outer-lobe components. We qualitatively reproduce the observed inverse correlation between peak frequency and source size, and find that the peak frequency of the integrated radio spectrum depends on ISM density but not the jet power. Spectral turnover in resolved young radio sources therefore provides a new probe of the ISM., Comment: 16 pages, 12 figures, 3 tables. Submitted to PASA

A remarkable characteristic of quantum computing is the potential for reliable computation despite faulty qubits. This can be achieved through quantum error correction, which is typically implemented by repeatedly applying static syndrome checks, permitting correction of logical information. Recently, the development of time-dynamic approaches to error correction has uncovered new codes and new code implementations. In this work, we experimentally demonstrate three time-dynamic implementations of the surface code, each offering a unique solution to hardware design challenges and introducing flexibility in surface code realization. First, we embed the surface code on a hexagonal lattice, reducing the necessary couplings per qubit from four to three. Second, we walk a surface code, swapping the role of data and measure qubits each round, achieving error correction with built-in removal of accumulated non-computational errors. Finally, we realize the surface code using iSWAP gates instead of the traditional CNOT, extending the set of viable gates for error correction without additional overhead. We measure the error suppression factor when scaling from distance-3 to distance-5 codes of $\Lambda_{35,\text{hex}} = 2.15(2)$, $\Lambda_{35,\text{walk}} = 1.69(6)$, and $\Lambda_{35,\text{iSWAP}} = 1.56(2)$, achieving state-of-the-art error suppression for each. With detailed error budgeting, we explore their performance trade-offs and implications for hardware design. This work demonstrates that dynamic circuit approaches satisfy the demands for fault-tolerance and opens new alternative avenues for scalable hardware design., Comment: 11 pages, 5 figures, Supplementary Information

Quantum error correction is essential for bridging the gap between the error rates of physical devices and the extremely low logical error rates required for quantum algorithms. Recent error-correction demonstrations on superconducting processors have focused primarily on the surface code, which offers a high error threshold but poses limitations for logical operations. In contrast, the color code enables much more efficient logic, although it requires more complex stabilizer measurements and decoding techniques. Measuring these stabilizers in planar architectures such as superconducting qubits is challenging, and so far, realizations of color codes have not addressed performance scaling with code size on any platform. Here, we present a comprehensive demonstration of the color code on a superconducting processor, achieving logical error suppression and performing logical operations. Scaling the code distance from three to five suppresses logical errors by a factor of $\Lambda_{3/5}$ = 1.56(4). Simulations indicate this performance is below the threshold of the color code, and furthermore that the color code may be more efficient than the surface code with modest device improvements. Using logical randomized benchmarking, we find that transversal Clifford gates add an error of only 0.0027(3), which is substantially less than the error of an idling error correction cycle. We inject magic states, a key resource for universal computation, achieving fidelities exceeding 99% with post-selection (retaining about 75% of the data). Finally, we successfully teleport logical states between distance-three color codes using lattice surgery, with teleported state fidelities between 86.5(1)% and 90.7(1)%. This work establishes the color code as a compelling research direction to realize fault-tolerant quantum computation on superconducting processors in the near future.

Large language models (LLMs) have exhibited outstanding performance in natural language processing tasks. However, these models remain susceptible to adversarial attacks in which slight input perturbations can lead to harmful or misleading outputs. A gradient-based defensive suffix generation algorithm is designed to bolster the robustness of LLMs. By appending carefully optimized defensive suffixes to input prompts, the algorithm mitigates adversarial influences while preserving the models' utility. To enhance adversarial understanding, a novel total loss function ($L_{\text{total}}$) combining defensive loss ($L_{\text{def}}$) and adversarial loss ($L_{\text{adv}}$) generates defensive suffixes more effectively. Experimental evaluations conducted on open-source LLMs such as Gemma-7B, mistral-7B, Llama2-7B, and Llama2-13B show that the proposed method reduces attack success rates (ASR) by an average of 11\% compared to models without defensive suffixes. Additionally, the perplexity score of Gemma-7B decreased from 6.57 to 3.93 when applying the defensive suffix generated by openELM-270M. Furthermore, TruthfulQA evaluations demonstrate consistent improvements with Truthfulness scores increasing by up to 10\% across tested configurations. This approach significantly enhances the security of LLMs in critical applications without requiring extensive retraining., Comment: 9 pages, 2 figures

This article investigates electron bunch density fluctuations in the 1 - 10 $\mu m$ wavelength range, focusing on their impact on coherent electron cooling (CEC) in hadron storage rings. In this study, we thoroughly compare the shot-noise model with experimental observations of optical transition radiation (OTR) generated by a relativistic electron bunch ($\gamma \approx$ 50), transiting an Aluminium metal surface. The bunch parameters are close to those proposed for a stage in an Electron-Ion Collider (EIC), where the bunch size is much larger than the OTR wavelength being measured. Here we present measurements and particle tracking results of both the low-level noise for the EIC bunch parameters and longitudinal space-charge-induced microbunching for the chicane-compressed bunch with coherent OTR enhancements up to 100 times in the various bandwidth-filtered near-infrared (NIR) OTR photodiode signals. We also discuss the corresponding limitations of the OTR method.

In this paper, we introduce the concept of graph shift-invariant space (GSIS) on an undirected finite graph, which is the linear space of graph signals being invariant under graph shifts, and we study its bandlimiting, kernel reproducing and sampling properties. Graph bandlimited spaces have been widely applied where large datasets on networks need to be handled efficiently. In this paper, we show that every GSIS is a bandlimited space, and every bandlimited space is a principal GSIS. Functions in a reproducing kernel Hilbert space with shift-invariant kernel could be learnt with significantly low computational cost. In this paper, we demonstrate that every GSIS is a reproducing kernel Hilbert space with a shift-invariant kernel. Based on the nested Krylov structure of GSISs in the spatial domain, we propose a novel sampling and reconstruction algorithm with finite steps, with its performance tested for well-localized signals on circulant graphs and flight delay dataset of the 50 busiest airports in the USA.

Optical beams carrying orbital angular momentum (OAM) have gained significant interest due to their unique properties, enhancing various communication systems and enabling applications such as the characterization of material or molecular chirality. Generating and detecting the OAM of light is thus crucial for numerous applications but poses challenges. This paper proposes a method utilizing stimulated Raman scattering to detect the magnitude of OAM. By exploring the strong Raman coupling between the pump and Stokes beams in higher-order Laguerre-Gauss modes, we demonstrate the discrimination of different optical vortex beams through nonlinear optical measurements. Numerical and experimental results support the feasibility of this approach, potentially advancing optical communication and sensing technologies.

Let $f$ be a newform of prime level $p$ with any central character $\chi\, (\bmod\, p)$, and let $g$ be a fixed cusp form or Eisenstein series for $\hbox{SL}_{2}(\mathbb{Z})$. We prove the subconvexity bound: for any $\varepsilon>0$, \begin{align*} L(1/2, \, f \otimes g) \ll p^{1/2-1/524+\varepsilon}, \end{align*} where the implied constant depends on $g$, $\varepsilon$, and the archimedean parameter of $f$. This improves upon the previously best-known result by Harcos and Michel. Our method ultimately relies on non-trivial bounds for bilinear forms in Kloosterman fractions pioneered by Duke, Friedlander, and Iwaniec, with later innovations by Bettin and Chandee., Comment: 27 pages. Comments welcome!

Phishing attacks on enterprise employees present one of the most costly and potent threats to organizations. We explore an understudied facet of enterprise phishing attacks: the email relay infrastructure behind successfully delivered phishing emails. We draw on a dataset spanning one year across thousands of enterprises, billions of emails, and over 800,000 delivered phishing attacks. Our work sheds light on the network origins of phishing emails received by real-world enterprises, differences in email traffic we observe from networks sending phishing emails, and how these characteristics change over time. Surprisingly, we find that over one-third of the phishing email in our dataset originates from highly reputable networks, including Amazon and Microsoft. Their total volume of phishing email is consistently high across multiple months in our dataset, even though the overwhelming majority of email sent by these networks is benign. In contrast, we observe that a large portion of phishing emails originate from networks where the vast majority of emails they send are phishing, but their email traffic is not consistent over time. Taken together, our results explain why no singular defense strategy, such as static blocklists (which are commonly used in email security filters deployed by organizations in our dataset), is effective at blocking enterprise phishing. Based on our offline analysis, we partnered with a large email security company to deploy a classifier that uses dynamically updated network-based features. In a production environment over a period of 4.5 months, our new detector was able to identify 3-5% more enterprise email attacks that were previously undetected by the company's existing classifiers., Comment: To appear in the proceedings of the Passive and Active Network Measurement (PAM 2025)

The growing demand for higher data rates and expanded bandwidth is driving the exploration of new frequency ranges, including the upper mid-band spectrum (6-24 GHz), which is a promising candidate for future Frequency Range 3 (FR3) applications. This paper presents ultra-wideband double-directional channel measurements in line-of-sight microcellular scenarios within the upper mid-band spectrum (6-18 GHz). Conducted in an urban street canyon environment, these measurements explore key channel characteristics such as power delay profiles, angular power spectra, path loss, delay spread, and angular spread to provide insights essential for robust communication system design. Our results reveal that path loss values for both omni-directional and best beam configurations are lower than free-space predictions due to multipath contributions from the environment. Analysis also indicates a high degree of stability in delay spread and angular spread across the entire band, with small variation between sub-bands.

Motivated by the time-reversal symmetry breaking signal in muon spin relaxation below a transition temperature without accompanying noticeable magnetic Bragg peaks in $5d^2$ Os double perovskites, a rare ferro-octupolar order was proposed to account for such hidden order. Here we study the phase transitions under a magnetic field in triangular and face-centered cubic lattices using classical Monte Carlo simulations. It is expected that higher-rank moments do not couple linearly to the magnetic field. Consequently, a field applied along the ferro-octupolar order is not anticipated to influence the ordering. However, we observe the emergence of antiferro-quadrupolar ordering mixed with the antiferro-octupolar order (AFQO) due to the field-induced bond-dependent exchange interaction. This field-linear interaction among non-Kramer doublets arises via the coupling to the excited triplet states enabled by the external field. In the triangular lattice, we uncover intriguing vortex states, which could inspire future research into $5d^2$ triangular lattice systems., Comment: 9 pages, 5 figures

As social systems become increasingly complex, legal articles are also growing more intricate, making it progressively harder for humans to identify any potential competitions among them, particularly when drafting new laws or applying existing laws. Despite this challenge, no method for detecting such competitions has been proposed so far. In this paper, we propose a new legal AI task called Legal Article Competition Detection (LACD), which aims to identify competing articles within a given law. Our novel retrieval method, CAM-Re2, outperforms existing relevant methods, reducing false positives by 20.8% and false negatives by 8.3%, while achieving a 98.2% improvement in precision@5, for the LACD task. We release our codes at https://github.com/asmath472/LACD-public., Comment: under review

Accurate precipitation forecasting is crucial for early warnings of disasters, such as floods and landslides. Traditional forecasts rely on ground-based radar systems, which are space-constrained and have high maintenance costs. Consequently, most developing countries depend on a global numerical model with low resolution, instead of operating their own radar systems. To mitigate this gap, we propose the Neural Precipitation Model (NPM), which uses global-scale geostationary satellite imagery. NPM predicts precipitation for up to six hours, with an update every hour. We take three key channels to discriminate rain clouds as input: infrared radiation (at a wavelength of 10.5 $\mu m$), upper- (6.3 $\mu m$), and lower- (7.3 $\mu m$) level water vapor channels. Additionally, NPM introduces positional encoders to capture seasonal and temporal patterns, accounting for variations in precipitation. Our experimental results demonstrate that NPM can predict rainfall in real-time with a resolution of 2 km. The code and dataset are available at https://github.com/seominseok0429/Data-driven-Precipitation-Nowcasting-Using-Satellite-Imagery., Comment: Accepted by AAAI2025

This study examines the effectiveness of traditional machine learning classifiers versus deep learning models for detecting the imagined speech using electroencephalogram data. Specifically, we evaluated conventional machine learning techniques such as CSP-SVM and LDA-SVM classifiers alongside deep learning architectures such as EEGNet, ShallowConvNet, and DeepConvNet. Machine learning classifiers exhibited significantly lower precision and recall, indicating limited feature extraction capabilities and poor generalization between imagined speech and idle states. In contrast, deep learning models, particularly EEGNet, achieved the highest accuracy of 0.7080 and an F1 score of 0.6718, demonstrating their enhanced ability in automatic feature extraction and representation learning, essential for capturing complex neurophysiological patterns. These findings highlight the limitations of conventional machine learning approaches in brain-computer interface (BCI) applications and advocate for adopting deep learning methodologies to achieve more precise and reliable classification of detecting imagined speech. This foundational research contributes to the development of imagined speech-based BCI systems.

We have conducted a comprehensive investigation into the magnetic properties of the chiral multiferroic material CoTeMoO$_6$. In contrast with the previous claim of canted antiferromagnetic order with ferromagnetic components, our investigation reveals an antiferromagnetic ground state with compensated moments, providing an interesting platform for exploring exotic material properties. Through careful measurements of magnetization under a series of applied field, we demonstrate that there exist two sequential field-induced magnetic transitions in CoTeMoO$_6$, with one occurring at $H_{c1}$=460 Oe along the a-axis, and the other at $H_{c2}$=1.16 T with the field along the b-axis. The values of $H_{c1}$ and $H_{c2}$ exhibit strong angular dependence and diverge with different rates as the applied field is rotated 90 degrees within the ab plane. This reflects the distinct nature of these transitions, which is further supported by the different critical behavior of $H_{c1}$ and $H_{c2}$, characterized by the values of $\gamma$,in the function of $H_c=H_0\times(1-\frac{T}{T_c})^n$. Furthermore, we have demonstrated that there exist structural and magnetic twin domains in CoTeMoO$_6$ that strongly affect the experimental measurement of their macroscopic properties. Intriguingly, these twin domains can be related to the orthorhombicity/chirality of the crystal structure with the space group $P2_1 2_1 2$. We further explored the magnetic and structural domains with uniaxial pressure and polarized light microscopy. Our results suggest that CoTeMoO$_6$ could be used as a unique platform for investigating the intriguing physics involving intertwined degrees of freedom. The tunability of the underlying domain distribution and its strong anisotropy could also be useful for developing functional devices and applications., Comment: 11 figures 6 tables

Cardiac Magnetic Resonance (CMR) imaging is widely used for heart modelling and digital twin computational analysis due to its ability to visualize soft tissues and capture dynamic functions. However, the anisotropic nature of CMR images, characterized by large inter-slice distances and misalignments from cardiac motion, poses significant challenges to accurate model reconstruction. These limitations result in data loss and measurement inaccuracies, hindering the capture of detailed anatomical structures. This study introduces MorphiNet, a novel network that enhances heart model reconstruction by leveraging high-resolution Computer Tomography (CT) images, unpaired with CMR images, to learn heart anatomy. MorphiNet encodes anatomical structures as gradient fields, transforming template meshes into patient-specific geometries. A multi-layer graph subdivision network refines these geometries while maintaining dense point correspondence. The proposed method achieves high anatomy fidelity, demonstrating approximately 40% higher Dice scores, half the Hausdorff distance, and around 3 mm average surface error compared to state-of-the-art methods. MorphiNet delivers superior results with greater inference efficiency. This approach represents a significant advancement in addressing the challenges of CMR-based heart model reconstruction, potentially improving digital twin computational analyses of cardiac structure and functions.

Large language models (LLMs) have recently emerged as powerful tools, finding many medical applications. LLMs' ability to coalesce vast amounts of information from many sources to generate a response-a process similar to that of a human expert-has led many to see potential in deploying LLMs for clinical use. However, medicine is a setting where accurate reasoning is paramount. Many researchers are questioning the effectiveness of multiple choice question answering (MCQA) benchmarks, frequently used to test LLMs. Researchers and clinicians alike must have complete confidence in LLMs' abilities for them to be deployed in a medical setting. To address this need for understanding, we introduce a knowledge graph (KG)-based method to evaluate the biomedical reasoning abilities of LLMs. Essentially, we map how LLMs link medical concepts in order to better understand how they reason. We test GPT-4, Llama3-70b, and PalmyraMed-70b, a specialized medical model. We enlist a panel of medical students to review a total of 60 LLM-generated graphs and compare these graphs to BIOS, a large biomedical KG. We observe GPT-4 to perform best in our human review but worst in our ground truth comparison; vice-versa with PalmyraMed, the medical model. Our work provides a means of visualizing the medical reasoning pathways of LLMs so they can be implemented in clinical settings safely and effectively., Comment: Findings paper presented at Machine Learning for Health (ML4H) symposium 2024, December 15-16, 2024, Vancouver, Canada, 19 pages

Developing excavation autonomy is challenging given the environments where excavators operate, the complexity of physical interaction and the degrees of freedom of operation of the excavator itself. Simulation is a useful tool to build parts of the autonomy without the complexity of experimentation. Traditional excavator simulators are geared towards high fidelity interactions between the joints or between the terrain but do not incorporate other challenges such as perception required for end to end autonomy. A complete simulator should be capable of supporting real time operation while providing high fidelity simulation of the excavator(s), the environment, and their interaction. In this paper we present TERA (Terrain Excavation Robot Autonomy), a simulator geared towards autonomous excavator applications based on Unity3D and AGX that provides the extensibility and scalability required to study full autonomy. It provides the ability to configure the excavator and the environment per the user requirements. We also demonstrate realistic dynamics by incorporating a time-varying model that introduces variations in the system's responses. The simulator is then evaluated with different scenarios such as track deformation, velocities on different terrains, similarity of the system with the real excavator and the overall path error to show the capabilities of the simulation., Comment: 7 pages, 9 figures

Silicon G centers show significant promise as single photon sources in a scalable silicon platform. But these color centers have large non-radiative decay and a low Debye-Waller factor, limiting their usability in quantum applications. In this work, we demonstrate bright Purcell-enhanced emission from a silicon G center by coupling it to a nanophotonic cavity. The nanobeam cavity enhances the spontaneous emission rate of a single G center by a factor of 6, corresponding to a Purcell factor greater than 31 when accounting for decay into the phonon sideband. We obtain a spontaneous emission rate of 0.97 ns, which is the fastest single photon emission rate reported in silicon. With this radiative enhancement, we achieve an order of magnitude improvement in emitter brightness compared to previously reported values. These results pave the way for scalable quantum light sources on a silicon photonic chip.

Cooperative effects such as super(sub)radiance in quantum systems arise from the interplay among quantum emitters. While bright superradiant states have been extensively studied and yielded significant insights into cooperative phenomena, subradiant states have remained less explored due to their inherently dark state nature. However, subradiance holds significant potential as valuable quantum resources that exploit long-lived and large-scale entanglement, which is a key for advancing quantum information technologies. Here, we demonstrate a long-lived subradiant state among multiple quantum emitters coupled to a directional low Q cavity. In a tailored photonic environment with balanced cavity dissipation, emitter-field coupling strength, and incoherent pumping, two coupled quantum dots exhibit a steady-state population in a subradiant state with highly negative cooperativity. As an important hallmark of a subradiant state, the system shows large photon bunching (g^((2))(0)>>2) and suppressed single-photon decay. In addition, controlling the excitation wavelength provides a useful tool for manipulating dephasing and the number of coupled emitters, which leads to significant changes in photon statistics. Our approach to inducing cavity-mediated subradiance paves the way for creating and harnessing quantum correlations in quantum emitters via a long-lived entangled quantum state, essential for quantum storage and metrology., Comment: In the manuscript, 16 pages and 4 figures. In supplementary, 6 pages and 7 figures

Magnetotransport of conventional semiconductor based double layer systems with barrier suppressed interlayer tunneling has been a rewarding subject due to the emergence of an interlayer coherent state that behaves as an excitonic superfluid. Large angle twisted bilayer graphene offers unprecedented strong interlayer Coulomb interaction, since both layer thickness and layer spacing are of atomic scale and a barrier is no more needed as the twist induced momentum mismatch suppresses tunneling. The extra valley degree of freedom also adds richness. Here we report the observation of fractional quantum Hall physics at 1/3 total filling for balanced layer population in this system. Monte Carlo simulations support that the ground state is also an excitonic superfluid but the excitons are composed of fractional rather than elementary charges. The observed phase transitions with an applied displacement field at this and other fractional fillings are also addressed with simulations. They reveal ground states with different topology and symmetry properties., Comment: accepted to Nature communications

Video inpainting (VI) is a challenging task that requires effective propagation of observable content across frames while simultaneously generating new content not present in the original video. In this study, we propose a robust and practical VI framework that leverages a large generative model for reference generation in combination with an advanced pixel propagation algorithm. Powered by a strong generative model, our method not only significantly enhances frame-level quality for object removal but also synthesizes new content in the missing areas based on user-provided text prompts. For pixel propagation, we introduce a one-shot pixel pulling method that effectively avoids error accumulation from repeated sampling while maintaining sub-pixel precision. To evaluate various VI methods in realistic scenarios, we also propose a high-quality VI benchmark, HQVI, comprising carefully generated videos using alpha matte composition. On public benchmarks and the HQVI dataset, our method demonstrates significantly higher visual quality and metric scores compared to existing solutions. Furthermore, it can process high-resolution videos exceeding 2K resolution with ease, underscoring its superiority for real-world applications., Comment: AAAI 2025

Supernovae (SNe) come in various flavors and are classified into different types based on emission and absorption lines in their spectra. SN candidates are now abundant with the advent of large systematic dynamic sky surveys like the Zwicky Transient Facility (ZTF), however, the identification bottleneck lies in their spectroscopic confirmation and classification. Fully robotic telescopes with dedicated spectrographs optimized for SN follow-up have eased the burden of data acquisition. However, the task of classifying the spectra still largely rests with the astronomers. Automating this classification step reduces human effort, and can make the SN type available sooner to the public. For this purpose, we have developed a deep-learning based program for classifying core-collapse supernovae (CCSNe) with ultra-low resolution spectra obtained with the SED-Machine IFU spectrograph on the Palomar 60-inch telescope. The program consists of hierarchical classification task layers, with each layer composed of multiple binary classifiers which are run in parallel to produce a reliable classification. The binary classifiers utilize RNN and CNN architecture and are designed to take multiple inputs, to supplement spectra with g- and r-band photometry from ZTF. On non-host-contaminated and good quality SEDM spectra ("gold" test set), CCSNscore is ~94% accurate (correct classifications) in distinguishing between hydrogen-rich (Type II) and hydrogen-poor (Type Ibc) CCSNe. With the help of light curve input, CCSNscore classifies ~83% of the gold set with high confidence (score >= 0.8 and score-error < 0.05), with ~98% accuracy., Comment: Submitted to PASP

Domain Generalization (DG) seeks to transfer knowledge from multiple source domains to unseen target domains, even in the presence of domain shifts. Achieving effective generalization typically requires a large and diverse set of labeled source data to learn robust representations that can generalize to new, unseen domains. However, obtaining such high-quality labeled data is often costly and labor-intensive, limiting the practical applicability of DG. To address this, we investigate a more practical and challenging problem: semi-supervised domain generalization (SSDG) under a label-efficient paradigm. In this paper, we propose a novel method, CAT, which leverages semi-supervised learning with limited labeled data to achieve competitive generalization performance under domain shifts. Our method addresses key limitations of previous approaches, such as reliance on fixed thresholds and sensitivity to noisy pseudo-labels. CAT combines adaptive thresholding with noisy label refinement techniques, creating a straightforward yet highly effective solution for SSDG tasks. Specifically, our approach uses flexible thresholding to generate high-quality pseudo-labels with higher class diversity while refining noisy pseudo-labels to improve their reliability. Extensive experiments across multiple benchmark datasets demonstrate the superior performance of our method, highlighting its effectiveness in achieving robust generalization under domain shift., Comment: 12 pages

Recent advancements in Large Language Models (LLMs) have been remarkable, with new models consistently surpassing their predecessors. These advancements are underpinned by extensive research on various training mechanisms. Among these, Preference Optimization has played a significant role in improving the performance of LLMs by incorporating human preferences into the training process. However, constructing preference optimization datasets is challenging and the optimization process is highly sensitive to the dataset quality. In this study, we propose a novel approach to augment Preference Optimization datasets. Additionally, we introduce a Multi-response-based Preference Optimization training method that enables the simultaneous learning of multiple responses.

Cosmic shear, galaxy clustering, and the abundance of massive halos each probe the large-scale structure of the universe in complementary ways. We present cosmological constraints from the joint analysis of the three probes, building on the latest analyses of the lensing-informed abundance of clusters identified by the South Pole Telescope (SPT) and of the auto- and cross-correlation of galaxy position and weak lensing measurements (3$\times$2pt) in the Dark Energy Survey (DES). We consider the cosmological correlation between the different tracers and we account for the systematic uncertainties that are shared between the large-scale lensing correlation functions and the small-scale lensing-based cluster mass calibration. Marginalized over the remaining $\Lambda$CDM parameters (including the sum of neutrino masses) and 52 astrophysical modeling parameters, we measure $\Omega_\mathrm{m}=0.300\pm0.017$ and $\sigma_8=0.797\pm0.026$. Compared to constraints from Planck primary CMB anisotropies, our constraints are only 15% wider with a probability to exceed of 0.22 ($1.2\sigma$) for the two-parameter difference. We further obtain $S_8\equiv\sigma_8(\Omega_\mathrm{m}/0.3)^{0.5}=0.796\pm0.013$ which is lower than the Planck measurement at the $1.6\sigma$ level. The combined SPT cluster, DES 3$\times$2pt, and Planck datasets mildly prefer a non-zero positive neutrino mass, with a 95% upper limit $\sum m_\nu<0.25~\mathrm{eV}$ on the sum of neutrino masses. Assuming a $w$CDM model, we constrain the dark energy equation of state parameter $w=-1.15^{+0.23}_{-0.17}$ and when combining with Planck primary CMB anisotropies, we recover $w=-1.20^{+0.15}_{-0.09}$, a $1.7\sigma$ difference with a cosmological constant. The precision of our results highlights the benefits of multiwavelength multiprobe cosmology., Comment: Submitted to Phys. Rev. D

Low-frequency solar radio emission is sourced in the solar corona, with sub-100 MHz radio emission largely originating from the $\sim$10$^{5}$\,$\mathrm{K}$ plasma around 2 optical radii. However, the region of emission has yet to be constrained at 35--45\,MHz due to both instrumentation limitations and the rarity of astronomical events, such as total solar eclipses, which allow for direct observational approaches. In this work, we present the results from a student-led project to commission a low-frequency radio telescope array situated in the path of totality of the 2024 total solar eclipse in an effort to probe the middle corona. The Deployable Low-Band Ionosphere and Transient Experiment (DLITE) is a low-frequency radio array comprised of four dipole antennas, optimized to observe at 35--45\,MHz, and capable of resolving the brightest radio sources in the sky. We constructed a DLITE station in Observatory Park, a dark sky park in Montville, Ohio. Results of observations during the total solar eclipse demonstrate that DLITE stations can be quickly deployed for observations and provide constraints on the radius of solar emission at our center observing frequency of 42\,MHz. In this work, we outline the construction of DLITE Ohio and the solar observation results from the total solar eclipse that transversed North America in April 2024.

Generalized linear models are a popular tool in applied statistics, with their maximum likelihood estimators enjoying asymptotic Gaussianity and efficiency. As all models are wrong, it is desirable to understand these estimators' behaviours under model misspecification. We study semiparametric multilevel generalized linear models, where only the conditional mean of the response is taken to follow a specific parametric form. Pre-existing estimators from mixed effects models and generalized estimating equations require specificaiton of a conditional covariance, which when misspecified can result in inefficient estimates of fixed effects parameters. It is nevertheless often computationally attractive to consider a restricted, finite dimensional class of estimators, as these models naturally imply. We introduce sandwich regression, that selects the estimator of minimal variance within a parametric class of estimators over all distributions in the full semiparametric model. We demonstrate numerically on simulated and real data the attractive improvements our sandwich regression approach enjoys over classical mixed effects models and generalized estimating equations.

With the advent of diffusion models, new proteins can be generated at an unprecedented rate. The \textit{motif scaffolding problem} requires steering this generative process to yield proteins with a desirable functional substructure -- a motif. While models have been trained to take the motif as conditional input, recent techniques in diffusion posterior sampling can be leveraged as zero-shot alternatives whose approximations can be corrected with sequential Monte Carlo (SMC) algorithms. In this work, we introduce a new set of guidance potentials to describe and solve scaffolding tasks by adapting SMC-aided diffusion posterior samplers with an unconditional model, Genie, acting as a prior. Against established benchmarks, we successfully scaffold several single-motif and multi-motif problems. The latter is possible by pairing reconstruction guidance with $\mathrm{SE}(3)$-invariant potentials. In the single-motif case, we find these potentials perform comparably to the conventional masking approach and that reconstruction guidance outperforms replacement methods when aided with SMC. We additionally consider a guidance potential for point symmetry constraints and produce designable internally symmetric monomers with our setup. Overall, this work highlights the capabilities and areas for improvement of zero-shot posterior samplers in motif scaffolding tasks. Code is available at: https://github.com/matsagad/mres-project

Symplectic mappings of the plane serve as key models for exploring the fundamental nature of complex behavior in nonlinear systems. Central to this exploration is the effective visualization of stability regimes, which enables the interpretation of how systems evolve under varying conditions. While the area-preserving quadratic H\'enon map has received significant theoretical attention, a comprehensive description of its mixed parameter-space dynamics remain lacking. This limitation arises from early attempts to reduce the full two-dimensional phase space to a one-dimensional projection, a simplification that resulted in the loss of important dynamical features. Consequently, there is a clear need for a more thorough understanding of the underlying qualitative aspects. This paper aims to address this gap by revisiting the foundational concepts of reversibility and associated symmetries, first explored in the early works of G.D. Birkhoff. We extend the original framework proposed by H\'enon by adding a period-doubling diagram to his isochronous diagram, which allows to represents the system's bifurcations and the groups of symmetric periodic orbits that emerge in typical bifurcations of the fixed point. A qualitative and quantitative explanation of the main features of the region of parameters with bounded motion is provided, along with the application of this technique to other symplectic mappings, including cases of multiple reversibility. Modern chaos indicators, such as the Reversibility Error Method and the Generalized Alignment Index, are employed to distinguish between various dynamical regimes in the mixed space of variables and parameters. These tools prove effective in differentiating regular and chaotic dynamics, as well as in identifying twistless orbits and their associated bifurcations.

Designing sparse directed spanners, which are subgraphs that approximately maintain distance constraints, has attracted sustained interest in TCS, especially due to their wide applicability, as well as the difficulty to obtain tight results. However, a significant drawback of the notion of spanners is that it cannot capture multiple distance-like constraints for the same demand pair. In this paper we initiate the study of directed multicriteria spanners, in which the notion of edge lengths is replaced by the notion of resource consumption vectors, where each entry corresponds to the consumption of the respective resource on that edge. The goal is to find a minimum-cost routing solution that satisfies the multiple constraints. To the best of our knowledge, we obtain the first approximation algorithms for the directed multicriteria spanners problems, under natural assumptions. Our results match the state-of-the-art approximation ratios in special cases of ours. We also establish reductions from other natural network connectivity problems to the directed multicriteria spanners problems, including Group Steiner Distances, introduced in the undirected setting by Bil\`o, Gual\`a, Leucci and Straziota (ESA 2024), and Edge-Avoiding spanners. Our reductions imply approximation algorithms for these problems and illustrate that the notion of directed multicriteria spanners is an appropriate abstraction and generalization of natural special cases from the literature. Our main technical tool is a delicate generalization of the minimum-density junction tree framework of Chekuri, Even, Gupta, and Segev (SODA 2008, TALG 2011) to the notion of minimum-density resource-constrained junction trees, which also extends ideas from Chlamt\'a\v{c}, Dinitz, Kortsarz, and Laekhanukit (SODA 2017, TALG 2020).

We analyzed the 7.92$\times 10^{11}$ cosmic-ray-induced muon events collected by the IceCube Neutrino Observatory from May 13, 2011, when the fully constructed experiment started to take data, to May 12, 2023. This dataset provides an up-to-date cosmic-ray arrival direction distribution in the Southern Hemisphere with unprecedented statistical accuracy covering more than a full period length of a solar cycle. Improvements in Monte Carlo event simulation and better handling of year-to-year differences in data processing significantly reduce systematic uncertainties below the level of statistical fluctuations compared to the previously published results. We confirm the observation of a change in the angular structure of the cosmic-ray anisotropy between 10 TeV and 1 PeV, more specifically in the 100-300 TeV energy range.

Dementia is associated with various cognitive impairments and typically manifests only after significant progression, making intervention at this stage often ineffective. To address this issue, the Prediction and Recognition of Cognitive Decline through Spontaneous Speech (PROCESS) Signal Processing Grand Challenge invites participants to focus on early-stage dementia detection. We provide a new spontaneous speech corpus for this challenge. This corpus includes answers from three prompts designed by neurologists to better capture the cognition of speakers. Our baseline models achieved an F1-score of 55.0% on the classification task and an RMSE of 2.98 on the regression task., Comment: 2 pages, no figure, conference

Understanding the internal computations of large language models (LLMs) is crucial for aligning them with human values and preventing undesirable behaviors like toxic content generation. However, mechanistic interpretability is hindered by polysemanticity -- where individual neurons respond to multiple, unrelated concepts. While Sparse Autoencoders (SAEs) have attempted to disentangle these features through sparse dictionary learning, they have compromised LLM performance due to reliance on post-hoc reconstruction loss. To address this issue, we introduce Mixture of Monosemantic Experts for Transformers (Monet) architecture, which incorporates sparse dictionary learning directly into end-to-end Mixture-of-Experts pretraining. Our novel expert decomposition method enables scaling the expert count to 262,144 per layer while total parameters scale proportionally to the square root of the number of experts. Our analyses demonstrate mutual exclusivity of knowledge across experts and showcase the parametric knowledge encapsulated within individual experts. Moreover, Monet allows knowledge manipulation over domains, languages, and toxicity mitigation without degrading general performance. Our pursuit of transparent LLMs highlights the potential of scaling expert counts to enhance mechanistic interpretability and directly resect the internal knowledge to fundamentally adjust model behavior. The source code and pretrained checkpoints are available at https://github.com/dmis-lab/Monet.

Retractions play a vital role in maintaining scientific integrity, yet systematic studies of retractions in computer science and other STEM fields remain scarce. We present WithdrarXiv, the first large-scale dataset of withdrawn papers from arXiv, containing over 14,000 papers and their associated retraction comments spanning the repository's entire history through September 2024. Through careful analysis of author comments, we develop a comprehensive taxonomy of retraction reasons, identifying 10 distinct categories ranging from critical errors to policy violations. We demonstrate a simple yet highly accurate zero-shot automatic categorization of retraction reasons, achieving a weighted average F1-score of 0.96. Additionally, we release WithdrarXiv-SciFy, an enriched version including scripts for parsed full-text PDFs, specifically designed to enable research in scientific feasibility studies, claim verification, and automated theorem proving. These findings provide valuable insights for improving scientific quality control and automated verification systems. Finally, and most importantly, we discuss ethical issues and take a number of steps to implement responsible data release while fostering open science in this area., Comment: 11 pages, 5 figures

Democratic governments comprise a subset of a population whose goal is to produce coherent decisions that solve societal challenges while respecting the will of the people they represent. New governance frameworks represent this problem as a social network rather than as a hierarchical pyramid with centralized authority. But how should this network be structured? To investigate this question, we model the set of decisions a population must make as a satisfiability problem and the structure of information flow involved in decision-making as a social hypergraph. This allows us to consider the benefits of different governance structures, from dictatorships to direct democracy. In between these extremes, we find a regime of effective governance where decision groups are formed as needed by key stakeholders to discuss and make specific decisions. This regime of effective governance allows even incoherent or polarized populations to make coherent decisions at low coordination costs. More broadly, we present not just simulation results, but a modeling framework that can be used to explore the costs and benefits of a wide range of governance strategies using bottom-up approaches and their ability to tackle decision problems that challenge standard governments.

DEtection TRansformer (DETR) has emerged as a promising architecture for object detection, offering an end-to-end prediction pipeline. In practice, however, DETR generates hundreds of predictions that far outnumber the actual number of objects present in an image. This raises the question: can we trust and use all of these predictions? Addressing this concern, we present empirical evidence highlighting how different predictions within the same image play distinct roles, resulting in varying reliability levels across those predictions. More specifically, while multiple predictions are often made for a single object, our findings show that most often one such prediction is well-calibrated, and the others are poorly calibrated. Based on these insights, we demonstrate identifying a reliable subset of DETR's predictions is crucial for accurately assessing the reliability of the model at both object and image levels. Building on this viewpoint, we first tackle the shortcomings of widely used performance and calibration metrics, such as average precision and various forms of expected calibration error. Specifically, they are inadequate for determining which subset of DETR's predictions should be trusted and utilized. In response, we present Object-level Calibration Error (OCE), which is capable of assessing the calibration quality both across different models and among various configurations within a specific model. As a final contribution, we introduce a post hoc Uncertainty Quantification (UQ) framework that predicts the accuracy of the model on a per-image basis. By contrasting the average confidence scores of positive (i.e., likely to be matched) and negative predictions determined by OCE, the framework assesses the reliability of the DETR model for each test image.

This paper presents a novel approach to designing millimeter-wave (mmWave) cellular communication systems, based on joint phase time array (JPTA) radio frequency (RF) frontend architecture. JPTA architecture comprises time-delay components appended to conventional phase shifters, which offer extra degrees of freedom to be exploited for designing frequency-selective analog beams. Hence, a mmWave device equipped with JPTA can receive and transmit signals in multiple directions in a single time slot per RF chain, one direction per frequency subband, which alleviates the traditional constraint of one analog beam per transceiver chain per time slot. The utilization of subband-specific analog beams offers a new opportunity in designing mmWave systems, allowing for enhanced cell capacity and reduced pilot overhead. To understand the practical feasibility of JPTA, a few challenges and system design considerations are discussed in relation to the performance and complexity of the JPTA systems. For example, frequency-selective beam gain losses are present for the subband analog beams, e.g., up to 1 dB losses for 2 subband cases, even with the state-of-the-art JPTA delay and phase optimization methods. Despite these side effects, system-level analysis reveals that the JPTA system is capable of improving cell capacity: the 5%-tile throughput by up to 65%.

Research on manipulating materials using light has garnered significant interest, yet examples of controlling electronic polarization in magnetic materials remain scarce. Here, we demonstrate the hysteresis of electronic polarization in the antiferromagnetic semiconductor FePS3 via light. Below the N\'eel temperature, we observe linear dichroism (i.e., optical anisotropy) without structural symmetry breaking. Light-induced net polarization aligns along the a-axis (zigzag direction) at 1.6 eV due to the dipolar polarization and along the b-axis (armchair direction) at 2.0 eV due to the combined effects of dipolar and octupolar polarizations, resulting from charge transfer from the armchair to the zigzag direction by light. Unexpected hysteresis of the electronic polarization occurs at 2.0 eV due to the octupolar polarization, in contrast to the absence of such hysteresis at 1.6 eV. We attribute this to a symmetry breaking of the light-induced phase of FePS3 involving electronic polarization within the spin lattice. This study suggests a new mechanism for generating and controlling electronic polarization in magnetic materials using light, with implications for future device applications., Comment: 34 pages, 5 figures, 13 supplementary figures

We investigate holographic complexities in the context of wedge holography, focusing specifically on black string geometry in AdS$_3$. The wedge spacetime is bounded by two end-of-the-world (EOW) branes with intrinsic Dvali-Gabadadze-Porrati (DGP) gravity. In line with this codimension-two holography, there are three equivalent perspectives: bulk perspective, brane perspective, and boundary perspective. Using both the ''Complexity=Volume'' (CV) and ''Complexity=Action'' (CA) proposals, we analyze the complexity in wedge black string geometry in the tensionless limit. By treating the branes as rigid, we find the late-time growth rates of CV and CA match exactly across bulk and brane perspectives. These results are consistent with those from JT gravity, with additional contributions from the intrinsic gravity of the branes. For fluctuating branes, we find that the late-time growth rates of CV and CA match between bulk and brane perspectives at the linear order of fluctuation. The CV results exhibit $\frac{\phi_h^2}{\phi_0}$ corrections from fluctuations, consistent with the findings in previous work. Moreover, the CA results reveal an additional constant term in the fluctuating branes case. We provide an interpretation of this in terms of gravitational edge mode effects. The distinct corrections arising from fluctuations in the CA and CV proposals suggest that the CV proposal is more sensitive to geometric details. Furthermore, we discuss these results in relation to Lloyd's bound on complexity, their general time dependence, and the effects of fluctuations., Comment: 35 pages, 3 figures

Quantum phase transitions (QPTs) are investigated in biquadratic spin-$1$ XY chain with rhombic single-ion anisotropy by using the ground state energy (GE), the bipartite entanglement entropy (BEE), and the mutual information (MI). It turns out that there are three spin nematic phases and two Tomonaga-Luttinger (TL) liquid phases with the central charge $c = 1$. The TL Liquid phases emerge roughly for biquadratic interaction strength two times stronger than the absolute value of the single-ion anisotropy. The GE and the derivatives up to the second order reveal a first-order QPT between spin nematic ferroquarupole (FQ) phases but cannot capture an evident signal of QPTs between the spin nematic phases and the TL Liquid phases as well as QPT between the two TL liquid phases. The TL liquid-to-liquid transition point features a highly degenerate state and the spin-block entanglement entropy increases logarithmically with block size. The BEE exhibits a divergent or convergent behavior identifying the TL Liquid or spin nematic FQ phases, respectively. Similarly, the MI and the spin-spin correlation are shown to decay algebraically or exponentially with increasing the lattice distance in the TL Liquid or spin nematic FQ phases, respectively. In the TL liquid phase, the exponents $\eta_I$ and $\eta_z$ of the MI and the spin-spin correlation vary with the interaction parameter of the biquadratic interaction strength and the rhombic single-ion anisotropy and satisfy the relationship of $\eta_z <\eta_I$. Such changes of characteristic behavior of the BEE, the MI and the spin-spin correlation indicate an occurrence of the Berezinskii-Kosterlitz-Thouless (BKT)-type QPT between the TL Liquid phase and the spin nematic FQ phase. The staggered spin fluctuation $\langle S^x S^y \rangle$ is shown to play a significant role for the emergence of the TL liquid phase and thus give rise to the BKT-type QPT., Comment: 14 pages, 11 figures

Modern vehicles equip dashcams that primarily collect visual evidence for traffic accidents. However, most of the video data collected by dashcams that is not related to traffic accidents is discarded without any use. In this paper, we present a use case for dashcam videos that aims to improve driving safety. By analyzing the real-time videos captured by dashcams, we can detect driving hazards and driver distractedness to alert the driver immediately. To that end, we design and implement a Distributed Edge-based dashcam Video Analytics system (DEVA), that analyzes dashcam videos using personal edge (mobile) devices in a vehicle. DEVA consolidates available in-vehicle edge devices to maintain the resource pool, distributes video frames for analysis to devices considering resource availability in each device, and dynamically adjusts frame rates of dashcams to control the overall workloads. The entire video analytics task is divided into multiple independent phases and executed in a pipelined manner to improve the overall frame processing throughput. We implement DEVA in an Android app and also develop a dashcam emulation app to be used in vehicles that are not equipped with dashcams. Experimental results using the apps and commercial smartphones show that DEVA can process real-time videos from two dashcams with frame rates of around 22~30 FPS per camera within 200 ms of latency, using three high-end devices., Comment: Submitted to Elsevier Internet of Things

3D surface reconstruction is essential across applications of virtual reality, robotics, and mobile scanning. However, RGB-based reconstruction often fails in low-texture, low-light, and low-albedo scenes. Handheld LiDARs, now common on mobile devices, aim to address these challenges by capturing depth information from time-of-flight measurements of a coarse grid of projected dots. Yet, these sparse LiDARs struggle with scene coverage on limited input views, leaving large gaps in depth information. In this work, we propose using an alternative class of "blurred" LiDAR that emits a diffuse flash, greatly improving scene coverage but introducing spatial ambiguity from mixed time-of-flight measurements across a wide field of view. To handle these ambiguities, we propose leveraging the complementary strengths of diffuse LiDAR with RGB. We introduce a Gaussian surfel-based rendering framework with a scene-adaptive loss function that dynamically balances RGB and diffuse LiDAR signals. We demonstrate that, surprisingly, diffuse LiDAR can outperform traditional sparse LiDAR, enabling robust 3D scanning with accurate color and geometry estimation in challenging environments.

In this paper, we propose a neural window decoder (NWD) for spatially coupled low-density parity-check (SC-LDPC) codes. The proposed NWD retains the conventional window decoder (WD) process but incorporates trainable neural weights. To train the weights of NWD, we introduce two novel training strategies. First, we restrict the loss function to target variable nodes (VNs) of the window, which prunes the neural network and accordingly enhances training efficiency. Second, we employ the active learning technique with a normalized loss term to prevent the training process from biasing toward specific training regions. Next, we develop a systematic method to derive non-uniform schedules for the NWD based on the training results. We introduce trainable damping factors that reflect the relative importance of check node (CN) updates. By skipping updates with less importance, we can omit $\mathbf{41\%}$ of CN updates without performance degradation compared to the conventional WD. Lastly, we address the error propagation problem inherent in SC-LDPC codes by deploying a complementary weight set, which is activated when an error is detected in the previous window. This adaptive decoding strategy effectively mitigates error propagation without requiring modifications to the code and decoder structures., Comment: 12 pages, 16 figures

Active research is currently underway to enhance the efficiency of vision transformers (ViTs). Most studies have focused solely on effective token mixers, overlooking the potential relationship with normalization. To boost diverse feature learning, we propose two components: a normalization module called multi-view normalization (MVN) and a token mixer called multi-view token mixer (MVTM). The MVN integrates three differently normalized features via batch, layer, and instance normalization using a learnable weighted sum. Each normalization method outputs a different distribution, generating distinct features. Thus, the MVN is expected to offer diverse pattern information to the token mixer, resulting in beneficial synergy. The MVTM is a convolution-based multiscale token mixer with local, intermediate, and global filters, and it incorporates stage specificity by configuring various receptive fields for the token mixer at each stage, efficiently capturing ranges of visual patterns. We propose a novel ViT model, multi-vision transformer (MVFormer), adopting the MVN and MVTM in the MetaFormer block, the generalized ViT scheme. Our MVFormer outperforms state-of-the-art convolution-based ViTs on image classification, object detection, and instance and semantic segmentation with the same or lower parameters and MACs. Particularly, MVFormer variants, MVFormer-T, S, and B achieve 83.4%, 84.3%, and 84.6% top-1 accuracy, respectively, on ImageNet-1K benchmark.

In this paper, we introduce the Actions and Objects Pathways (AOPath) for out-of-domain generalization in video question answering tasks. AOPath leverages features from a large pretrained model to enhance generalizability without the need for explicit training on the unseen domains. Inspired by human brain, AOPath dissociates the pretrained features into action and object features, and subsequently processes them through separate reasoning pathways. It utilizes a novel module which converts out-of-domain features into domain-agnostic features without introducing any trainable weights. We validate the proposed approach on the TVQA dataset, which is partitioned into multiple subsets based on genre to facilitate the assessment of generalizability. The proposed approach demonstrates 5% and 4% superior performance over conventional classifiers on out-of-domain and in-domain datasets, respectively. It also outperforms prior methods that involve training millions of parameters, whereas the proposed approach trains very few parameters.

The contribution of a time-independent toroidally-symmetric radial electric field $E_r$ is implemented in VisualStart [Comp. Phys. Comm. 275 (2022) 108305; arXiv:2111.08224], a code whose purposes include the construction of guiding center (GC) drift orbit databases for the study of plasma instabilities in tokamaks. $E_r$ is important for the thermal part of the velocity distribution and for fast particle resonances in the kHz frequency range. KSTAR, JT-60U and ITER tokamak cases are used as working examples to test our methods and discuss practical issues connected with $E_r$. Two points are worth noting: First, the GC orbit space is sampled in the magnetic midplane as before, and we find that in the presence of $E_r$, midplane-based coordinates are not only equivalent but superior to conventional constants of motion, allowing to attain high numerical accuracy and efficiency with a relatively simple mesh. Second, the periodic parallel acceleration and deceleration of GCs via the mirror force is modulated by $E_r$. Although this parallel electric acceleration averages to zero during a poloidal transit (or bounce) period, it has important consequences, one being the known shift of the trapped-passing boundary. Another consequence is that electric frequency shifts depend on the chosen reference point, so that some care is required when evaluating the $E_r$-dependence of transit frequencies for resonance analyses., Comment: 42 pages, 23 figures, submitted to Computer Physics Communications

Personalized image generation has been significantly advanced, enabling the creation of highly realistic and customized images. However, existing methods often struggle with generating images of multiple people due to occlusions and fail to accurately personalize full-body shapes. In this paper, we propose PersonaCraft, a novel approach that combines diffusion models with 3D human modeling to address these limitations. Our method effectively manages occlusions by incorporating 3D-aware pose conditioning with SMPLx-ControlNet and accurately personalizes human full-body shapes through SMPLx fitting. Additionally, PersonaCraft enables user-defined body shape adjustments, adding flexibility for individual body customization. Experimental results demonstrate the superior performance of PersonaCraft in generating high-quality, realistic images of multiple individuals while resolving occlusion issues, thus establishing a new standard for multi-person personalized image synthesis. Project page: https://gwang-kim.github.io/persona_craft, Comment: Project page: https://gwang-kim.github.io/persona_craft