Search

Vision transformers dominate image processing tasks due to their superior performance. However, the quadratic complexity of self-attention limits the scalability of these systems and their deployment on resource-constrained devices. State Space Models (SSMs) have emerged as a solution by introducing a linear recurrence mechanism, which reduces the complexity of sequence modeling from quadratic to linear. Recently, SSMs have been extended to high-resolution vision tasks. Nonetheless, the linear recurrence mechanism struggles to fully utilize matrix multiplication units on modern hardware, resulting in a computational bottleneck. We address this issue by introducing \textit{VMeanba}, a training-free compression method that eliminates the channel dimension in SSMs using mean operations. Our key observation is that the output activations of SSM blocks exhibit low variances across channels. Our \textit{VMeanba} leverages this property to optimize computation by averaging activation maps across the channel to reduce the computational overhead without compromising accuracy. Evaluations on image classification and semantic segmentation tasks demonstrate that \textit{VMeanba} achieves up to a 1.12x speedup with less than a 3\% accuracy loss. When combined with 40\% unstructured pruning, the accuracy drop remains under 3\%., Comment: Accepted by NeurIPS 2024 Machine Learning for Systems workshop

Multimodal foundation models, such as Gemini and ChatGPT, have revolutionized human-machine interactions by seamlessly integrating various forms of data. Developing a universal spoken language model that comprehends a wide range of natural language instructions is critical for bridging communication gaps and facilitating more intuitive interactions. However, the absence of a comprehensive evaluation benchmark poses a significant challenge. We present Dynamic-SUPERB Phase-2, an open and evolving benchmark for the comprehensive evaluation of instruction-based universal speech models. Building upon the first generation, this second version incorporates 125 new tasks contributed collaboratively by the global research community, expanding the benchmark to a total of 180 tasks, making it the largest benchmark for speech and audio evaluation. While the first generation of Dynamic-SUPERB was limited to classification tasks, Dynamic-SUPERB Phase-2 broadens its evaluation capabilities by introducing a wide array of novel and diverse tasks, including regression and sequence generation, across speech, music, and environmental audio. Evaluation results indicate that none of the models performed well universally. SALMONN-13B excelled in English ASR, while WavLLM demonstrated high accuracy in emotion recognition, but current models still require further innovations to handle a broader range of tasks. We will soon open-source all task data and the evaluation pipeline.

Assigning upper elementary grade teachers to teach their strongest subjects to multiple classes ("departmentalizing"), rather than teaching all subjects to a single class, ("self-contained instruction") could mean more specialized instructional expertise in the classroom or focus for teacher planning time and professional development. This study examined the experiences of 90 schools that either voluntarily switched to departmentalized instruction for up to two years or continued with self-contained teaching in 4th and 5th grade classes beginning in fall 2019. The findings show that schools that switched were able to implement the key steps needed to departmentalize instruction but struggled with aspects of the approach. Teachers' reported experiences were consistent with some of both the benefits and challenges that prior research had hypothesized. Although the pandemic significantly disrupted instruction and the study activities, school and teacher experiences during this challenging time may inform schools and districts considering adopting departmentalized instruction.

This appendix is a companion to the report "Evaluation of Departmentalized Instruction in Elementary Schools: Exploring Implementation Experiences." The appendix provides additional information about the study, which sought to explore schools' implementation of departmentalized instruction, teachers' experiences with the strategy, and how the strategy was related to key outcomes. This additional information includes details on departmentalized instruction (Section A), study design and execution (Section B), key findings presented in the report (Section C), and additional findings that are not included in the report but provide important context (Section D).

We report measurements of anisotropic triple-$q$ charge density wave (CDW) fluctuations in the transition metal dichalcogenide 1$T$-TiSe$_2$ over a large volume of reciprocal space with X-ray diffuse scattering. Above the transition temperature, $T_{\text{CDW}}$, the out-of-plane diffuse scattering is characterized by rod-like structures which indicate that the CDW fluctuations in neighboring layers are largely decoupled. In addition, the in-plane diffuse scattering is marked by ellipses which reveal that the in-plane fluctuations are anisotropic. Our analysis of the diffuse scattering line shapes and orientations suggests that the three charge density wave components contain independent phase fluctuations. At $T_{\text{CDW}}$, long range coherence is established in both the in-plane and out-of-plane directions, consistent with the large observed value of the CDW gap compared to $T_{\text{CDW}}$, and the predicted presence of a hierarchy of energy scales.

This paper proposes the ViT Token Constraint and Multi-scale Memory bank (TCMM) method to address the patch noises and feature inconsistency in unsupervised person re-identification works. Many excellent methods use ViT features to obtain pseudo labels and clustering prototypes, then train the model with contrastive learning. However, ViT processes images by performing patch embedding, which inevitably introduces noise in patches and may compromise the performance of the re-identification model. On the other hand, previous memory bank based contrastive methods may lead data inconsistency due to the limitation of batch size. Furthermore, existing pseudo label methods often discard outlier samples that are difficult to cluster. It sacrifices the potential value of outlier samples, leading to limited model diversity and robustness. This paper introduces the ViT Token Constraint to mitigate the damage caused by patch noises to the ViT architecture. The proposed Multi-scale Memory enhances the exploration of outlier samples and maintains feature consistency. Experimental results demonstrate that our system achieves state-of-the-art performance on common benchmarks. The project is available at \href{https://github.com/andy412510/TCMM}{https://github.com/andy412510/TCMM}.

Expected improvement (EI) is one of the most widely-used acquisition functions in Bayesian optimization (BO). Despite its proven success in applications for decades, important open questions remain on the theoretical convergence behaviors and rates for EI. In this paper, we contribute to the convergence theories of EI in three novel and critical area. First, we consider objective functions that are under the Gaussian process (GP) prior assumption, whereas existing works mostly focus on functions in the reproducing kernel Hilbert space (RKHS). Second, we establish the first asymptotic error bound and its corresponding rate for GP-EI with noisy observations under the GP prior assumption. Third, by investigating the exploration and exploitation of the non-convex EI function, we prove improved error bounds for both the noise-free and noisy cases. The improved noiseless bound is extended to the RKHS assumption as well.

We present Omni-RGPT, a multimodal large language model designed to facilitate region-level comprehension for both images and videos. To achieve consistent region representation across spatio-temporal dimensions, we introduce Token Mark, a set of tokens highlighting the target regions within the visual feature space. These tokens are directly embedded into spatial regions using region prompts (e.g., boxes or masks) and simultaneously incorporated into the text prompt to specify the target, establishing a direct connection between visual and text tokens. To further support robust video understanding without requiring tracklets, we introduce an auxiliary task that guides Token Mark by leveraging the consistency of the tokens, enabling stable region interpretation across the video. Additionally, we introduce a large-scale region-level video instruction dataset (RegVID-300k). Omni-RGPT achieves state-of-the-art results on image and video-based commonsense reasoning benchmarks while showing strong performance in captioning and referring expression comprehension tasks., Comment: Project page: https://miranheo.github.io/omni-rgpt/

It is now common to evaluate Large Language Models (LLMs) by having humans manually vote to evaluate model outputs, in contrast to typical benchmarks that evaluate knowledge or skill at some particular task. Chatbot Arena, the most popular benchmark of this type, ranks models by asking users to select the better response between two randomly selected models (without revealing which model was responsible for the generations). These platforms are widely trusted as a fair and accurate measure of LLM capabilities. In this paper, we show that if bot protection and other defenses are not implemented, these voting-based benchmarks are potentially vulnerable to adversarial manipulation. Specifically, we show that an attacker can alter the leaderboard (to promote their favorite model or demote competitors) at the cost of roughly a thousand votes (verified in a simulated, offline version of Chatbot Arena). Our attack consists of two steps: first, we show how an attacker can determine which model was used to generate a given reply with more than $95\%$ accuracy; and then, the attacker can use this information to consistently vote for (or against) a target model. Working with the Chatbot Arena developers, we identify, propose, and implement mitigations to improve the robustness of Chatbot Arena against adversarial manipulation, which, based on our analysis, substantially increases the cost of such attacks. Some of these defenses were present before our collaboration, such as bot protection with Cloudflare, malicious user detection, and rate limiting. Others, including reCAPTCHA and login are being integrated to strengthen the security in Chatbot Arena.

Using the full four-year SPTpol 500 deg$^2$ dataset in both the 95 GHz and 150 GHz frequency bands, we present measurements of the temperature and $E$-mode polarization of the cosmic microwave background (CMB), as well as the $E$-mode polarization auto-power spectrum ($EE$) and temperature-$E$-mode cross-power spectrum ($TE$) in the angular multipole range $50<\ell<8000$. We find the SPTpol dataset to be self-consistent, passing several internal consistency tests based on maps, frequency bands, bandpowers, and cosmological parameters. The full SPTpol dataset is well-fit by the $\Lambda CDM$ model, for which we find $H_0=70.48\pm2.16$ km s$^{-1}$ Mpc$^{-1}$ and $\Omega_m=0.271\pm0.026$, when using only the SPTpol data and a Planck-based prior on the optical depth to reionization. The $\Lambda CDM$ parameter constraints are consistent across the 95 GHz-only, 150 GHz-only, $TE$-only, and $EE$-only data splits. Between the $\ell<1000$ and $\ell>1000$ data splits, the $\Lambda CDM$ parameter constraints are borderline consistent at the $\sim2\sigma$ level. This consistency improves when including a parameter $A_L$, the degree of lensing of the CMB inferred from the smearing of acoustic peaks. When marginalized over $A_L$, the $\Lambda CDM$ parameter constraints from SPTpol are consistent with those from Planck. The power spectra presented here are the most sensitive measurements of the lensed CMB damping tail to date for roughly $\ell > 1700$ in $TE$ and $\ell > 2000$ in $EE$.

Many renewable energy resources are integrated into power systems via grid-following (GFL) inverters which rely on a phase-locked loop (PLL) for grid synchronization. During severe grid faults, GFL inverters are vulnerable to transient instability, often leading to disconnection from the grid. This paper aims to elucidate the interaction mechanisms and define the stability boundaries of systems of two inverters, including GFL, grid-forming (GFM), or grid-supporting (GSP) inverters. First, the generalized large-signal expression for the two-inverter system under various inverter combinations is derived, revealing that no energy function exists for systems containing GFL inverters. This implies that the traditional direct method cannot be applied to such systems. To overcome these challenges, a manifold method is employed to precisely determine the domain of attraction (DOA) of the system, and the transient stability margin is assessed by a new metric termed the critical clearing radius (CCR). A case study of the two-inverter system under various inverter combinations is conducted to explore large-signal interactions across different scenarios. Manifold analysis and simulation results reveal that GSP inverters using PLL for grid synchronization exhibit behavior similar to GFM inverters when the droop coefficients in the terminal voltage control loop (TVC) are sufficiently large. Compared to GFL inverters, GSP inverters incorporating a TVC significantly enhances the transient stability of other inverters. In the STATCOM case, the optimal placement of the STATCOM, realized by GSP or GFM inverters, is identified to be at the midpoint of a transmission line. All findings in this paper are validated through electromagnetic transient (EMT) simulations

The OpenUniverse2024 simulation suite is a cross-collaboration effort to produce matched simulated imaging for multiple surveys as they would observe a common simulated sky. Both the simulated data and associated tools used to produce it are intended to uniquely enable a wide range of studies to maximize the science potential of the next generation of cosmological surveys. We have produced simulated imaging for approximately 70 deg$^2$ of the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST) Wide-Fast-Deep survey and the Nancy Grace Roman Space Telescope High-Latitude Wide-Area Survey, as well as overlapping versions of the ELAIS-S1 Deep-Drilling Field for LSST and the High-Latitude Time-Domain Survey for Roman. OpenUniverse2024 includes i) an early version of the updated extragalactic model called Diffsky, which substantially improves the realism of optical and infrared photometry of objects, compared to previous versions of these models; ii) updated transient models that extend through the wavelength range probed by Roman and Rubin; and iii) improved survey, telescope, and instrument realism based on up-to-date survey plans and known properties of the instruments. It is built on a new and updated suite of simulation tools that improves the ease of consistently simulating multiple observatories viewing the same sky. The approximately 400 TB of synthetic survey imaging and simulated universe catalogs are publicly available, and we preview some scientific uses of the simulations.

Neural speech editing advancements have raised concerns about their misuse in spoofing attacks. Traditional partially edited speech corpora primarily focus on cut-and-paste edits, which, while maintaining speaker consistency, often introduce detectable discontinuities. Recent methods, like A\textsuperscript{3}T and Voicebox, improve transitions by leveraging contextual information. To foster spoofing detection research, we introduce the Speech INfilling Edit (SINE) dataset, created with Voicebox. We detailed the process of re-implementing Voicebox training and dataset creation. Subjective evaluations confirm that speech edited using this novel technique is more challenging to detect than conventional cut-and-paste methods. Despite human difficulty, experimental results demonstrate that self-supervised-based detectors can achieve remarkable performance in detection, localization, and generalization across different edit methods. The dataset and related models will be made publicly available., Comment: SLT 2024

High-performance DUV spectroscopy drives advancements in biomedical research, clinical diagnosis, and material science. Existing DUV resonant nanostructures face instability and photoluminescent noise challenges. We propose robust Si metasurfaces leveraging polaritonic resonances, a unique property driven by interband transitions, for enhanced nanophotonic sensing. Our polaritonic Kerker-type void metasurface enables double-resonance Raman scattering to analyze 2D semiconductors, improves biomolecule autofluorescence, and offers superior stability. This scalable platform unlocks versatile applications in interdisciplinary DUV spectroscopy and emerging nanomaterials research., Comment: in press, the DOI will be DOI: 10.1002/adfm.202420439

Studies have indicated that personality is related to achievement, and several personality assessment models have been developed. However, most are either questionnaires or based on marker systems, which entails limitations. We proposed a physiological signal based model, thereby ensuring the objectivity of the data and preventing unreliable responses. Thirty participants were recruited from the Department of Electrical Engineering of Yuan Ze University in Taiwan. Wearable sensors were used to collect physiological signals as the participants watched and summarized a video. They then completed a personality questionnaire based on the big five factor markers system. The results were used to construct a personality prediction model, which revealed that galvanic skin response and heart rate variance were key factors predicting extroversion; heart rate variance also predicted agreeableness and conscientiousness. The results of this experiment can elucidate students personality traits, which can help educators select the appropriate pedagogical methods.

Past researches show that personality trait is a strong predictor for ones academic performance. Today, mature and verified marker systems for assessing personality traits already exist. However, marker systems-based assessing methods have their own limitations. For example, dishonest responses cannot be avoided. In this research, the goal is to develop a method that can overcome the limitations. The proposed method will rely on physiological signals for the assessment. Thirty participants have participated in this experiment. Based on the statistical results, we found that there are correlations between students personality traits and their physiological signal change when learning via videos. Specifically, we found that participants degree of extraversion, agreeableness, conscientiousness, and openness to experiences are correlated with the variance of heart rates, the variance of GSR values, and the skewness of voice frequencies, etc.

Recent advancements in object-centric text-to-3D generation have shown impressive results. However, generating complex 3D scenes remains an open challenge due to the intricate relations between objects. Moreover, existing methods are largely based on score distillation sampling (SDS), which constrains the ability to manipulate multiobjects with specific interactions. Addressing these critical yet underexplored issues, we present a novel framework of Scene Graph and Layout Guided 3D Scene Generation (GraLa3D). Given a text prompt describing a complex 3D scene, GraLa3D utilizes LLM to model the scene using a scene graph representation with layout bounding box information. GraLa3D uniquely constructs the scene graph with single-object nodes and composite super-nodes. In addition to constraining 3D generation within the desirable layout, a major contribution lies in the modeling of interactions between objects in a super-node, while alleviating appearance leakage across objects within such nodes. Our experiments confirm that GraLa3D overcomes the above limitations and generates complex 3D scenes closely aligned with text prompts., Comment: 13 pages, 12 figures

We discuss the landscape of flavor physics at the Circular Electron-Positron Collider (CEPC), based on the nominal luminosity outlined in its Technical Design Report. The CEPC is designed to operate in multiple modes to address a variety of tasks. At the $Z$ pole, the expected production of 4 Tera $Z$ bosons will provide unique and highly precise measurements of $Z$ boson couplings, while the substantial number of boosted heavy-flavored quarks and leptons produced in clean $Z$ decays will facilitate investigations into their flavor physics with unprecedented precision. We investigate the prospects of measuring various physics benchmarks and discuss their implications for particle theories and phenomenological models. Our studies indicate that, with its highlighted advantages and anticipated excellent detector performance, the CEPC can explore beauty and $\tau$ physics in ways that are superior to or complementary with the Belle II and Large-Hadron-Collider-beauty experiments, potentially enabling the detection of new physics at energy scales of 10 TeV and above. This potential also extends to the observation of yet-to-be-discovered rare and exotic processes, as well as testing fundamental principles such as lepton flavor universality, lepton and baryon number conservation, etc., making the CEPC a vibrant platform for flavor physics research. The $WW$ threshold scan, Higgs-factory operation and top-pair productions of the CEPC further enhance its merits in this regard, especially for measuring the Cabibbo-Kobayashi-Maskawa matrix elements, and Flavor-Changing-Neutral-Current physics of Higgs boson and top quarks. We outline the requirements for detector performance and considerations for future development to achieve the anticipated scientific goals.

Bayesian optimization (BO) with Gaussian process (GP) surrogate models is a powerful black-box optimization method. Acquisition functions are a critical part of a BO algorithm as they determine how the new samples are selected. Some of the most widely used acquisition functions include upper confidence bound (UCB) and Thompson sampling (TS). The convergence analysis of BO algorithms has focused on the cumulative regret under both the Bayesian and frequentist settings for the objective. In this paper, we establish new pointwise bounds on the prediction error of GP under the frequentist setting with Gaussian noise. Consequently, we prove improved convergence rates of cumulative regret bound for both GP-UCB and GP-TS. Of note, the new prediction error bound under Gaussian noise can be applied to general BO algorithms and convergence analysis, e.g., the asymptotic convergence of expected improvement (EI) with noise.

Climate change poses an extreme threat to biodiversity, making it imperative to efficiently model the geographical range of different species. The availability of large-scale remote sensing images and environmental data has facilitated the use of machine learning in Species Distribution Models (SDMs), which aim to predict the presence of a species at any given location. Traditional SDMs, reliant on expert observation, are labor-intensive, but advancements in remote sensing and citizen science data have facilitated machine learning approaches to SDM development. However, these models often struggle with leveraging spatial relationships between different inputs -- for instance, learning how climate data should inform the data present in satellite imagery -- without upsampling or distorting the original inputs. Additionally, location information and ecological characteristics at a location play a crucial role in predicting species distribution models, but these aspects have not yet been incorporated into state-of-the-art approaches. In this work, we introduce MiTREE: a multi-input Vision-Transformer-based model with an ecoregion encoder. MiTREE computes spatial cross-modal relationships without upsampling as well as integrates location and ecological context. We evaluate our model on the SatBird Summer and Winter datasets, the goal of which is to predict bird species encounter rates, and we find that our approach improves upon state-of-the-art baselines., Comment: 11 pages, GeoAI Workshop and SIGSPATIAL 2024

Artifact detection algorithms are crucial to correcting the output generated by diffusion models. However, because of the variety of artifact forms, existing methods require substantial annotated data for training. This requirement limits their scalability and efficiency, which restricts their wide application. This paper shows that the similarity of denoised images between consecutive time steps during the sampling process is related to the severity of artifacts in images generated by diffusion models. Building on this observation, we introduce the concept of Similarity Trajectory to characterize the sampling process and its correlation with the image artifacts presented. Using an annotated data set of 680 images, which is only 0.1% of the amount of data used in the prior work, we trained a classifier on these trajectories to predict the presence of artifacts in images. By performing 10-fold validation testing on the balanced annotated data set, the classifier can achieve an accuracy of 72.35%, highlighting the connection between the Similarity Trajectory and the occurrence of artifacts. This approach enables differentiation between artifact-exhibiting and natural-looking images using limited training data.

Deep Neural Networks (DNNs) have revolutionized artificial intelligence, achieving impressive results on diverse data types, including images, videos, and texts. However, DNNs still lag behind Gradient Boosting Decision Trees (GBDT) on tabular data, a format extensively utilized across various domains. In this paper, we propose DOFEN, short for \textbf{D}eep \textbf{O}blivious \textbf{F}orest \textbf{EN}semble, a novel DNN architecture inspired by oblivious decision trees. DOFEN constructs relaxed oblivious decision trees (rODTs) by randomly combining conditions for each column and further enhances performance with a two-level rODT forest ensembling process. By employing this approach, DOFEN achieves state-of-the-art results among DNNs and further narrows the gap between DNNs and tree-based models on the well-recognized benchmark: Tabular Benchmark \citep{grinsztajn2022tree}, which includes 73 total datasets spanning a wide array of domains. The code of DOFEN is available at: \url{https://github.com/Sinopac-Digital-Technology-Division/DOFEN}., Comment: NeurIPS 2024 (poster); (v2: modify and rearrange sections, propose multihead extension of DOFEN, include new results on tabular benchmark and other benchmarks)

Disease progression models are widely used to inform the diagnosis and treatment of many progressive diseases. However, a significant limitation of existing models is that they do not account for health disparities that can bias the observed data. To address this, we develop an interpretable Bayesian disease progression model that captures three key health disparities: certain patient populations may (1) start receiving care only when their disease is more severe, (2) experience faster disease progression even while receiving care, or (3) receive follow-up care less frequently conditional on disease severity. We show theoretically and empirically that failing to account for disparities produces biased estimates of severity (underestimating severity for disadvantaged groups, for example). On a dataset of heart failure patients, we show that our model can identify groups that face each type of health disparity, and that accounting for these disparities meaningfully shifts which patients are considered high-risk.

We describe the global structure of a particle model with dark matter called G2HDM, which incorporates a dark sector represented by $SU(2)_{H} \times U(1)_{X}\;$. The gauge group of such model is $\tilde{G}=U(1)_{Y} \times SU(2)_{L} \times SU(3)_{C} \times SU(2)_{H} \times U(1)_{X} \;$, with an ambiguity that it may actually be $G= \tilde{G} / {\Gamma} \;$, where $\Gamma$ is a subgroup of its center. We also describe how the electric and magnetic charges depend on $\Gamma$ and the periodicity of theta angles for any choice of $G$. Finally, we describe the minimal charges that arise after electroweak symmetry breaking for any choice of $\Gamma$., Comment: 21 pages, 13 figures

We examine Friedmann-Lema\^itre-Robertson-Walker cosmology, incorporating quantum gravitational corrections through the functional renormalization group flow of the effective action for gravity. We solve the Einstein equation with quantum improved coupling perturbatively including the case with non-vanishing classical cosmological constant (CC) which was overlooked in the literatures. We discuss what is the suitable identification of the momentum cutoff $k$ with time scale, and find that the choice of the Hubble parameter is suitable for vanishing CC but not so for non-vanishing CC. We suggest suitable identification in this case. The energy-scale dependent running coupling breaks the time translation symmetry and then introduces a new physical scale., Comment: 16 pages

Although facial landmark detection (FLD) has gained significant progress, existing FLD methods still suffer from performance drops on partially non-visible faces, such as faces with occlusions or under extreme lighting conditions or poses. To address this issue, we introduce ORFormer, a novel transformer-based method that can detect non-visible regions and recover their missing features from visible parts. Specifically, ORFormer associates each image patch token with one additional learnable token called the messenger token. The messenger token aggregates features from all but its patch. This way, the consensus between a patch and other patches can be assessed by referring to the similarity between its regular and messenger embeddings, enabling non-visible region identification. Our method then recovers occluded patches with features aggregated by the messenger tokens. Leveraging the recovered features, ORFormer compiles high-quality heatmaps for the downstream FLD task. Extensive experiments show that our method generates heatmaps resilient to partial occlusions. By integrating the resultant heatmaps into existing FLD methods, our method performs favorably against the state of the arts on challenging datasets such as WFLW and COFW., Comment: WACV 2025 Project Link: https://ben0919.github.io/ORFormer/

With increasing revelations of academic fraud, detecting forged experimental images in the biomedical field has become a public concern. The challenge lies in the fact that copy-move targets can include background tissue, small foreground objects, or both, which may be out of the training domain and subject to unseen attacks, rendering standard object-detection-based approaches less effective. To address this, we reformulate the problem of detecting biomedical copy-move forgery regions as an intra-image co-saliency detection task and propose CMSeg-Net, a copy-move forgery segmentation network capable of identifying unseen duplicated areas. Built on a multi-resolution encoder-decoder architecture, CMSeg-Net incorporates self-correlation and correlation-assisted spatial-attention modules to detect intra-image regional similarities within feature tensors at each observation scale. This design helps distinguish even small copy-move targets in complex microscopic images from other similar objects. Furthermore, we created a copy-move forgery dataset of optical microscopic images, named FakeParaEgg, using open data from the ICIP 2022 Challenge to support CMSeg-Net's development and verify its performance. Extensive experiments demonstrate that our approach outperforms previous state-of-the-art methods on the FakeParaEgg dataset and other open copy-move detection datasets, including CASIA-CMFD, CoMoFoD, and CMF. The FakeParaEgg dataset, our source code, and the CMF dataset with our manually defined segmentation ground truths available at ``https://github.com/YoursEver/FakeParaEgg''., Comment: submitted to IEEE SPL

We study conditions under which transformers using soft attention can simulate hard attention, that is, effectively focus all attention on a subset of positions. First, we examine several variants of linear temporal logic, whose formulas have been previously been shown to be computable using hard attention transformers. We demonstrate how soft attention transformers can compute formulas of these logics using unbounded positional embeddings or temperature scaling. Second, we demonstrate how temperature scaling allows softmax transformers to simulate a large subclass of average-hard attention transformers, those that have what we call the uniform-tieless property.

The Hydrogen Intensity and Real-time Analysis eXperiment (HIRAX) is an upcoming radio interferometric telescope designed to constrain dark energy through the 21cm intensity mapping of Baryon Acoustic Oscillations (BAO). Instrumental systematics must be controlled and carefully characterized to measure the 21cm power spectrum with fidelity and achieve high-precision constraints on the cosmological parameters. The chromaticity of the primary beam is one such complicated systematic, which can leak the power of spectrally smooth foregrounds beyond the ideal horizon limits due to the complex spatial and spectral structures of the sidelobes and the mainlobe. This paper studies the chromaticity of the HIRAX Stokes I primary beam and its effects on accurate measurements of the 21cm power spectrum. To investigate the effect of chromaticity in the 21cm power spectrum, we present a physically motivated beam modeling technique, which uses a flexible basis derived from traditional optics that can account for higher-order radial and azimuthal structures in the primary beam. We investigate the impact of imperfect knowledge of the mainlobe and sidelobes chromaticity in the power spectrum space by subtracting a simple foreground model in simulated snapshot visibilities to recover the H$\textsc{i}$ power spectrum. Additionally, we find that modeling up to the octupolar azimuthal order feature (fourth-order angular variation) in the primary beam is sufficient to reduce the leakage outside the wedge with minimal bias., Comment: 20 pages, 12 figures. Prepared for submitting in the Astrophysical Journal (ApJ)

With the growing adoption and capabilities of vision-language models (VLMs) comes the need for benchmarks that capture authentic user-VLM interactions. In response, we create VisionArena, a dataset of 230K real-world conversations between users and VLMs. Collected from Chatbot Arena - an open-source platform where users interact with VLMs and submit preference votes - VisionArena spans 73K unique users, 45 VLMs, and 138 languages. Our dataset contains three subsets: VisionArena-Chat, 200k single and multi-turn conversations between a user and a VLM; VisionArena-Battle, 30K conversations comparing two anonymous VLMs with user preference votes; and VisionArena-Bench, an automatic benchmark of 500 diverse user prompts that efficiently approximate the live Chatbot Arena model rankings. Additionally, we highlight the types of question asked by users, the influence of response style on preference, and areas where models often fail. We find open-ended tasks like captioning and humor are highly style-dependent, and current VLMs struggle with spatial reasoning and planning tasks. Lastly, we show finetuning the same base model on VisionArena-Chat outperforms Llava-Instruct-158K, with a 17-point gain on MMMU and a 46-point gain on the WildVision benchmark. Dataset at https://huggingface.co/lmarena-ai

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

We provide a taxonomy of a dozen enabling network architectures, protocols, and technologies that will define the evolution from 5G to 6G. These technologies span the network protocol stack, different target deployment environments, and various perceived levels of technical maturity. We outline four areas of societal focus that will be impacted by these technologies, and overview several research directions that hold the potential to address the problems in these important focus areas., Comment: This paper has been accepted for publication in the IEEE Communications Magazine. Portions were released online as a report titled 6G Roadmap: A Global Taxonomy in November 2023

We propose an efficient radiance field rendering algorithm that incorporates a rasterization process on sparse voxels without neural networks or 3D Gaussians. There are two key contributions coupled with the proposed system. The first is to render sparse voxels in the correct depth order along pixel rays by using dynamic Morton ordering. This avoids the well-known popping artifact found in Gaussian splatting. Second, we adaptively fit sparse voxels to different levels of detail within scenes, faithfully reproducing scene details while achieving high rendering frame rates. Our method improves the previous neural-free voxel grid representation by over 4db PSNR and more than 10x rendering FPS speedup, achieving state-of-the-art comparable novel-view synthesis results. Additionally, our neural-free sparse voxels are seamlessly compatible with grid-based 3D processing algorithms. We achieve promising mesh reconstruction accuracy by integrating TSDF-Fusion and Marching Cubes into our sparse grid system., Comment: Code release in progress

The CO-to-H$_2$ conversion factor ($\alpha_\mathrm{CO}$) is expected to vary with dust abundance and grain size distribution through the efficiency of shielding gas from CO-dissociation radiation. We present a comprehensive analysis of $\alpha_\mathrm{CO}$ and grain size distribution for nearby galaxies, using the PAH fraction ($q_\mathrm{PAH}$) as an observable proxy of grain size distribution. We adopt the resolved observations at 2-kpc resolution in 42 nearby galaxies, where $\alpha_\mathrm{CO}$ is derived from measured metallicity and surface densities of dust and HI assuming a fixed dust-to-metals ratio. We use an analytical model for the evolution of H$_2$ and CO, in which the evolution of grain size distribution is controlled by the dense gas fraction ($\eta$). We find that the observed level of $q_\mathrm{PAH}$ is consistent with the diffuse-gas-dominated model ($\eta=0.2$) where dust shattering is more efficient. Meanwhile, the slight decreasing trend of observed $q_\mathrm{PAH}$ with metallicity is more consistent with high-$\eta$ predictions, likely due to the more efficient loss of PAHs by coagulation. We discuss how grain size distribution (indicated by $q_\mathrm{PAH}$) and metallicity impact $\alpha_\mathrm{CO}$; we however did not obtain conclusive evidence that the grain size distribution affects $\alpha_\mathrm{CO}$. Observations and model predictions show similar anti-correlation between $\alpha_\mathrm{CO}$ and 12+log(O/H). Meanwhile, there is a considerable difference in how resolved $\alpha_\mathrm{CO}$ behaves with $q_\mathrm{PAH}$. The observed $\alpha_\mathrm{CO}$ has a positive correlation with $q_\mathrm{PAH}$, while the model-predicted $\alpha_\mathrm{CO}$ does not have a definite correlation with $q_\mathrm{PAH}$. This difference is likely due to the limitation of one-zone treatment in the model., Comment: 12 pages, 7 figures, accepted for publication in MNRAS

The recent surge in high-quality visual instruction tuning samples from closed-source vision-language models (VLMs) such as GPT-4V has accelerated the release of open-source VLMs across various model sizes. However, scaling VLMs to improve performance using larger models brings significant computational challenges, especially for deployment on resource-constrained devices like mobile platforms and robots. To address this, we propose VLsI: Verbalized Layers-to-Interactions, a new VLM family in 2B and 7B model sizes, which prioritizes efficiency without compromising accuracy. VLsI leverages a unique, layer-wise distillation process, introducing intermediate "verbalizers" that map features from each layer to natural language space, allowing smaller VLMs to flexibly align with the reasoning processes of larger VLMs. This approach mitigates the training instability often encountered in output imitation and goes beyond typical final-layer tuning by aligning the small VLMs' layer-wise progression with that of the large ones. We validate VLsI across ten challenging vision-language benchmarks, achieving notable performance gains (11.0% for 2B and 17.4% for 7B) over GPT-4V without the need for model scaling, merging, or architectural changes., Comment: Project page: https://byungkwanlee.github.io/VLsI-page/

In this work, we perform on-chip quantum random number generation (QRNG) that uses a novel differential amplifier configuration for conjugate homodyne detection. Leveraging separate integrated photonics and integrated analog circuit platforms, we present an alternative method for QRNG. This approach exploits the observable $\hat{\text{Z}}$, derived from the sum of squared conjugate quadrature distributions which we compare to the traditional single quadrature approach. Utilizing this method, we report a shot noise clearance (SNC) of 25.6 dB and a common mode rejection ratio (CMRR) of 69 dB for our homodyne detection system. We used a variety of design tools to model and predict performance and compare results with our measurements. The realization of our QRNG system consists of a 90{\deg} optical hybrid, a dual differential transimpedance amplifier (TIA), and a field-programmable gate array (FPGA) used for the real-time post-processing to produce a uniform random bitstream. The randomness extraction is implemented using a Toeplitz hashing algorithm and is validated by the National Institute of Standards and Technology (NIST) randomness test suites., Comment: 8 pages + appendix. 10 figures + 2 tables

Weak supervision combines the advantages of training on real data with the ability to exploit signal properties. However, training a neural network using weak supervision often requires an excessive amount of signal data, which severely limits its practical applicability. In this study, we propose addressing this limitation through data augmentation, increasing the training data's size and diversity. Specifically, we focus on physics-inspired data augmentation methods, such as $p_{\text{T}}$ smearing and jet rotation. Our results demonstrate that data augmentation can significantly enhance the performance of weak supervision, enabling neural networks to learn efficiently from substantially less data.

The rapid advancement of generative models has introduced serious risks, including deepfake techniques for facial synthesis and editing. Traditional approaches rely on training classifiers and enhancing generalizability through various feature extraction techniques. Meanwhile, training-free detection methods address issues like limited data and overfitting by directly leveraging statistical properties from vision foundation models to distinguish between real and fake images. The current leading training-free approach, RIGID, utilizes DINOv2 sensitivity to perturbations in image space for detecting fake images, with fake image embeddings exhibiting greater sensitivity than those of real images. This observation prompts us to investigate how detection performance varies across model backbones, perturbation types, and datasets. Our experiments reveal that detection performance is closely linked to model robustness, with self-supervised (SSL) models providing more reliable representations. While Gaussian noise effectively detects general objects, it performs worse on facial images, whereas Gaussian blur is more effective due to potential frequency artifacts. To further improve detection, we introduce Contrastive Blur, which enhances performance on facial images, and MINDER (MINimum distance DetEctoR), which addresses noise type bias, balancing performance across domains. Beyond performance gains, our work offers valuable insights for both the generative and detection communities, contributing to a deeper understanding of model robustness property utilized for deepfake detection.

Neutrino nucleus elastic scattering ({\nu}Ael) with reactor neutrinos is an interaction under full quantum-mechanical coherence. It has not yet been experimentally observed. We present new results on the studies of {\nu}Ael cross section with an electro-cooled p-type point-contact germanium detector at the Kuo-Sheng Reactor Neutrino laboratory. A total of (242)357 kg-days of Reactor ON(OFF) data at a detector threshold of 200 eVee in electron equivalent unit are analyzed. The Lindhard model parametrized by a single variable k which characterizes the quenching function was used. Limits at 90% confidence level are derived on the ratio {\rho} relative to standard model (SM) cross section of {\rho}<4.7 at the predicted value of k=0.162, while k<0.285 at the SM-value of {\rho}=1. Prospects on future positive measurements are discussed., Comment: 6 pages, 4 figures

Despite the significant strides made by generative AI in just a few short years, its future progress is constrained by the challenge of building modular and robust systems. This capability has been a cornerstone of past technological revolutions, which relied on combining components to create increasingly sophisticated and reliable systems. Cars, airplanes, computers, and software consist of components-such as engines, wheels, CPUs, and libraries-that can be assembled, debugged, and replaced. A key tool for building such reliable and modular systems is specification: the precise description of the expected behavior, inputs, and outputs of each component. However, the generality of LLMs and the inherent ambiguity of natural language make defining specifications for LLM-based components (e.g., agents) both a challenging and urgent problem. In this paper, we discuss the progress the field has made so far-through advances like structured outputs, process supervision, and test-time compute-and outline several future directions for research to enable the development of modular and reliable LLM-based systems through improved specifications.

Modern laser plasma accelerators (LPAs) often require meter-scale plasma waveguides to propagate a high-intensity drive laser pulse. Tapering the longitudinal gas density profile in meter-scale gas jets could allow for single stage laser plasma acceleration well beyond 10 GeV with current petawatt-class laser systems. Via simulation and interferometry measurements, we show density control by longitudinally adjusting the throat width and jet angle. Density profiles appropriate for tapering were calculated analytically and via particle-in-cell (PIC) simulations, and were matched experimentally. These simulations show that tapering can increase electron beam energy using 19 J laser energy from $\sim$9 GeV to $>$12 GeV in a 30 cm plasma, and the accelerated charge by an order of magnitude.

Tightly packed granular particles under shear often exhibit intriguing intermittencies, specifically, sudden stress drops that we refer to as quaking. To probe the nature of this phenomenon, we prototype a circular shear cell that is capable of imposing a uniform and unlimited shear strain under quasi-static cyclic driving. Spherical PDMS(polydimethylsiloxane) particles, immersed in fluid, are driven in a fixed total volume at a wide range of shear rates, with particle trajectories captured in 3D space via refraction-index-matched fluorescent tomography. Statistics on the magnitude of fluctuating displacements of individual particles shows distinct dependence on the shear rate. Particle motions are smooth at high shear rates. At intermediate shear rates, quaking emerges with clusters of particles exhibiting relatively large displacements. At low shear rates, a cluster can span the entire system. and the cluster exhibits substructures in view of localized particle movements. Overall, we have confirmed that the quaking phenomena in the current setup are consistent with our previous work [Phys. Rev. Lett.,126, 128001 (2021)], and that the dimensionless shear rate that we have proposed is indeed a good parameter for unifying the transitions observed in different experimental geometries.

Recent developments in All-in-One (AiO) RGB image restoration and prompt learning have enabled the representation of distinct degradations through prompts, allowing degraded images to be effectively addressed by a single restoration model. However, this paradigm faces significant challenges when transferring to hyperspectral image (HSI) restoration tasks due to: 1) the domain gap between RGB and HSI features and difference on their structures, 2) information loss in visual prompts under severe composite degradations, and 3) difficulties in capturing HSI-specific degradation representations through text prompts. To address these challenges, we propose PromptHSI, the first universal AiO HSI restoration framework. By leveraging the frequency-aware feature modulation based on characteristics of HSI degradations, we decompose text prompts into intensity and bias controllers to effectively guide the restoration process while avoiding domain gaps. Our unified architecture excels at both fine-grained recovery and global information restoration tasks. Experimental results demonstrate superior performance under various degradation combinations, indicating great potential for practical remote sensing applications. The source code and dataset will be publicly released., Comment: 11 pages, 8 figures

A charge density wave (CDW) is a phase of matter characterized by a periodic modulation of the valence electron density accompanied by a distortion of the lattice structure. The microscopic details of CDW formation are closely tied to the dynamic charge susceptibility, $\chi(q,\omega)$, which describes the behavior of electronic collective modes. Despite decades of extensive study, the behavior of $\chi(q,\omega)$ in the vicinity of a CDW transition has never been measured with high energy resolution ($\sim$meV). Here, we investigate the canonical CDW transition in ErTe$_3$ using momentum-resolved electron energy loss spectroscopy (M-EELS), a technique uniquely sensitive to valence band charge excitations. Unlike phonons in these materials, which undergo conventional softening due to the Kohn anomaly at the CDW wavevector, the electronic excitations display purely relaxational dynamics that are well described by a diffusive model. The diffusivity peaks around 250 K, just below the critical temperature. Additionally, we report, for the first time, a divergence in the real part of $\chi(q,\omega)$ in the static limit ($\omega \rightarrow 0$), a phenomenon predicted to characterize CDWs since the 1970s. These results highlight the importance of energy- and momentum-resolved measurements of electronic susceptibility and demonstrate the power of M-EELS as a versatile probe of charge dynamics in materials.

Here, we present the outcomes from the second Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistry, which engaged participants across global hybrid locations, resulting in 34 team submissions. The submissions spanned seven key application areas and demonstrated the diverse utility of LLMs for applications in (1) molecular and material property prediction; (2) molecular and material design; (3) automation and novel interfaces; (4) scientific communication and education; (5) research data management and automation; (6) hypothesis generation and evaluation; and (7) knowledge extraction and reasoning from scientific literature. Each team submission is presented in a summary table with links to the code and as brief papers in the appendix. Beyond team results, we discuss the hackathon event and its hybrid format, which included physical hubs in Toronto, Montreal, San Francisco, Berlin, Lausanne, and Tokyo, alongside a global online hub to enable local and virtual collaboration. Overall, the event highlighted significant improvements in LLM capabilities since the previous year's hackathon, suggesting continued expansion of LLMs for applications in materials science and chemistry research. These outcomes demonstrate the dual utility of LLMs as both multipurpose models for diverse machine learning tasks and platforms for rapid prototyping custom applications in scientific research., Comment: Updating author information, the submission remains largely unchanged. 98 pages total

Record linkage integrates diverse data sources by identifying records that refer to the same entity. In the context of mineral site records, accurate record linkage is crucial for identifying and mapping mineral deposits. Properly linking records that refer to the same mineral deposit helps define the spatial coverage of mineral areas, benefiting resource identification and site data archiving. Mineral site record linkage falls under the spatial record linkage category since the records contain information about the physical locations and non-spatial attributes in a tabular format. The task is particularly challenging due to the heterogeneity and vast scale of the data. While prior research employs pre-trained discriminative language models (PLMs) on spatial entity linkage, they often require substantial amounts of curated ground-truth data for fine-tuning. Gathering and creating ground truth data is both time-consuming and costly. Therefore, such approaches are not always feasible in real-world scenarios where gold-standard data are unavailable. Although large generative language models (LLMs) have shown promising results in various natural language processing tasks, including record linkage, their high inference time and resource demand present challenges. We propose a method that leverages an LLM to generate training data and fine-tune a PLM to address the training data gap while preserving the efficiency of PLMs. Our approach achieves over 45\% improvement in F1 score for record linkage compared to traditional PLM-based methods using ground truth data while reducing the inference time by nearly 18 times compared to relying on LLMs. Additionally, we offer an automated pipeline that eliminates the need for human intervention, highlighting this approach's potential to overcome record linkage challenges., Comment: 11 pages, 10 figures

The strange metal is a peculiar phase of matter in which the electron scattering rate, $\tau^{-1} \sim k_B T/\hbar$, which determines the electrical resistance, is universal across a wide family of materials and determined only by fundamental constants. In 1989, theorists hypothesized that this universality would manifest as scale-invariant behavior in the dynamic charge susceptibility, $\chi''(q,\omega)$. Here, we present momentum-resolved inelastic electron scattering measurements of the strange metal Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$ showing that the susceptibility has the scale-invariant form $\chi''(q,\omega) = T^{-\nu} f(\omega/T)$, with exponent $\nu = 0.93$. We find the response is consistent with conformal invariance, meaning the dynamics may be thought of as occurring on a circle of radius $1/T$ in imaginary time, characterized by conformal dimension $\Delta = 0.05$. Our study indicates that the strange metal is a universal phenomenon whose properties are not determined by microscopic properties of a particular material., Comment: 14 pages, 4 figures + supplementary data

We present a machine-learning interatomic potential (MLIP) framework, which substantially accelerates the prediction of lattice thermal conductivity for both particle-like and wave-like thermal transport, including three-phonon and four-phonon scattering processes, and achieves speedups of five orders of magnitude compared to the conventional DFT calculations. We illustrate our approach through an in-depth study of Mg2GeSe4 as an exemplar thermoelectric material. Four phonon scattering is found to reduce lattice thermal conductivity by 22.5% at 300K and 26.7% at 900 K. The particle-like contribution to lattice thermal conductivity decreases, while the wave-like component increases with increasing temperature. The maximum figure of merit zT at 900K is found to be 0.49 for the n-type and 0.45 for the p-type Mg2GeSe4, respectively. Our analysis reveals that a substantial contribution to the Seebeck coefficient is driven by multi-band features in the valence band. Our study gives insight into the promising thermoelectric performance of Mg2GeSe4 and provides an efficient and accurate scheme for developing rational materials discovery strategies and practical applications of thermoelectric materials.

The ability to concentrate sound energy with a tunable focal point is essential for a wide range of acoustic applications, offering precise control over the location and intensity of sound pressure maxima. However, conventional acoustic metalenses are typically passive, with fixed focal positions, limiting their versatility. A significant obstacle in achieving tunable sound wave focusing lies in the complexity of precise and programmable adjustments, which often require intricate mechanical or electronic systems. In this study, we present a theoretical and experimental investigation of a reconfigurable acoustic metalens based on a bistable origami design. The metalens comprises eight flexible origami units, each capable of switching between two stable equilibrium states, enabling local modulation of sound waves through two distinct reflection phases. The metalens can be locked into specific symmetric or asymmetric configurations by manually tailoring the origami units to settle either of the two states. Each configuration generates a unique phase profile, focusing sound energy at a specific point. This concept allows the focal spot to be dynamically reconfigured both on and off-axis. Furthermore, the approach introduces a simple yet effective mechanism for tuning sound energy concentration, offering a solution for flexible acoustic manipulation.