Search

Merging models becomes a fundamental procedure in some applications that consider model efficiency and robustness. The training randomness or Non-I.I.D. data poses a huge challenge for averaging-based model fusion. Previous research efforts focus on element-wise regularization or neural permutations to enhance model averaging while overlooking weight scope variations among models, which can significantly affect merging effectiveness. In this paper, we reveal variations in weight scope under different training conditions, shedding light on its influence on model merging. Fortunately, the parameters in each layer basically follow the Gaussian distribution, which inspires a novel and simple regularization approach named Weight Scope Alignment (WSA). It contains two key components: 1) leveraging a target weight scope to guide the model training process for ensuring weight scope matching in the subsequent model merging. 2) fusing the weight scope of two or more models into a unified one for multi-stage model fusion. We extend the WSA regularization to two different scenarios, including Mode Connectivity and Federated Learning. Abundant experimental studies validate the effectiveness of our approach.

We present CasualGaze, a novel eye-gaze-based target selection technique to support natural and casual eye-gaze input. Unlike existing solutions that require users to keep the eye-gaze center on the target actively, CasualGaze allows users to glance at the target object to complete the selection simply. To understand casual gaze behavior, we studied the spatial distribution of casual gaze for different layouts and user behavior in a simulated real-world environment. Results revealed the impacts of object parameters, the speed and randomness features of casual gaze, and special gaze behavior patterns in "blurred areas". Based on the results, we devised CasualGaze algorithms, employing a bivariate Gaussian distribution model along with temporal compensation and voting algorithms for robust target prediction. Usability evaluation study showed significant improvements in recognition and selection speed for CasualGaze compared with two baseline techniques. Subjective ratings and comments further supported the preference for CasualGaze regarding efficiency, accuracy, and stability.

A smartwatch worn continuously on the wrist has the potential to perceive rich interactive gestures and natural behaviors of the user. Unfortunately, the current interaction functionality of smartwatches is primarily limited by the small touch screen. This paper proposes SonarWatch, a novel sensing technique that uses the acoustic field generated by the transceiver on the opposite sides of the watch to detect the presence of nearby objects and their shapes. This enables a range of gesture interactions and natural behavior perception. We designed an algorithm combining IMU and acoustic fields to identify these actions and optimize power consumption. We tested the performance of SonarWatch in different noise environments, achieving an overall accuracy of 93.7%. Its power consumption is close to that of physiological sensors. SonarWatch can achieve the above capabilities by utilizing the existing built-in sensors, making it a technology with solid practical value.

A long-standing challenge in end-user programming (EUP) is to trade off between natural user expression and the complexity of programming tasks. As large language models (LLMs) are empowered to handle semantic inference and natural language understanding, it remains under-explored how such capabilities can facilitate end-users to configure complex automation more naturally and easily. We propose AwareAuto, an EUP system that standardizes user expression and finishes two-step inference with the LLMs to achieve automation generation. AwareAuto allows contextual, multi-modality, and flexible user expression to configure complex automation tasks (e.g., dynamic parameters, multiple conditional branches, and temporal constraints), which are non-manageable in traditional EUP solutions. By studying realistic, complex rules data, AwareAuto gains 91.7% accuracy in matching user intentions and feasibility. We introduced user interaction to ensure system controllability and usability. We discuss the opportunities and challenges of incorporating LLMs in end-user programming techniques and grounding complex smart home contexts.

In order to explore how the ubiquitous stochastic low-frequency (SLF) variability of O-type stars is related to various stellar characteristics, we compiled a sample of 150 O-type stars observed via ground-based spectroscopic surveys, alongside photometric data obtained from the Transiting Exoplanet Survey Satellite (TESS). We analyzed 298 light curves obtained from TESS sectors 1-65 for the stars in our sample. Leveraging the spectroscopic parameters, we used Bonnsai to determine masses, radii, fractional main sequence ages, and mass-loss rates for stars of our sample. Subsequently, we identified possible correlations between the fitted parameters of SLF variability and stellar properties. Our analysis unveiled four significant correlations between the amplitude and stellar parameters, including mass, radius, fractional main sequence ages, and mass-loss rate. For stars with $M \gtrsim 30 M_{\odot}$, we observed a decrease in characteristic frequency and steepness with increasing radius. Finally, we compared various physical processes that may account for the SLF variability with our results. The observed SLF variability may arise from the combined effects of FeCZ and IGWs, with IGWs potentially more dominant in the early stages of stellar evolution, and the contribution of FeCZ becoming more significant as stars evolve. Meanwhile, our results indicate that the SLF variability of O-type stars bears certain signatures of the line-driven wind instability and granulation., Comment: 40 pages, 12 figures,accepted by Astrophysical Journal Supplement Series

Diffusion Models have emerged as powerful generative models for high-quality image synthesis, with many subsequent image editing techniques based on them. However, the ease of text-based image editing introduces significant risks, such as malicious editing for scams or intellectual property infringement. Previous works have attempted to safeguard images from diffusion-based editing by adding imperceptible perturbations. These methods are costly and specifically target prevalent Latent Diffusion Models (LDMs), while Pixel-domain Diffusion Models (PDMs) remain largely unexplored and robust against such attacks. Our work addresses this gap by proposing a novel attacking framework with a feature representation attack loss that exploits vulnerabilities in denoising UNets and a latent optimization strategy to enhance the naturalness of protected images. Extensive experiments demonstrate the effectiveness of our approach in attacking dominant PDM-based editing methods (e.g., SDEdit) while maintaining reasonable protection fidelity and robustness against common defense methods. Additionally, our framework is extensible to LDMs, achieving comparable performance to existing approaches.

Visual State Space Model (VSS) has demonstrated remarkable performance in various computer vision tasks. However, in the process of development, backdoor attacks have brought severe challenges to security. Such attacks cause an infected model to predict target labels when a specific trigger is activated, while the model behaves normally on benign samples. In this paper, we conduct systematic experiments to comprehend on robustness of VSS through the lens of backdoor attacks, specifically how the state space model (SSM) mechanism affects robustness. We first investigate the vulnerability of VSS to different backdoor triggers and reveal that the SSM mechanism, which captures contextual information within patches, makes the VSS model more susceptible to backdoor triggers compared to models without SSM. Furthermore, we analyze the sensitivity of the VSS model to patch processing techniques and discover that these triggers are effectively disrupted. Based on these observations, we consider an effective backdoor for the VSS model that recurs in each patch to resist patch perturbations. Extensive experiments across three datasets and various backdoor attacks reveal that the VSS model performs comparably to Transformers (ViTs) but is less robust than the Gated CNNs, which comprise only stacked Gated CNN blocks without SSM., Comment: 11 pages, 9 figures, minor revise, under review

Quantum computing holds immense potential for addressing a myriad of intricate challenges, which is significantly amplified when scaled to thousands of qubits. However, a major challenge lies in developing an efficient and scalable quantum control system. To address this, we propose a novel Hierarchical MicroArchitecture (HiMA) designed to facilitate qubit scaling and exploit quantum process-level parallelism. This microarchitecture is based on three core elements: (i) discrete qubit-level drive and readout, (ii) a process-based hierarchical trigger mechanism, and (iii) multiprocessing with a staggered triggering technique to enable efficient quantum process-level parallelism. We implement HiMA as a control system for a 72-qubit tunable superconducting quantum processing unit, serving a public quantum cloud computing platform, which is capable of expanding to 6144 qubits through three-layer cascading. In our benchmarking tests, HiMA achieves up to a 4.89x speedup under a 5-process parallel configuration. Consequently, to the best of our knowledge, we have achieved the highest CLOPS (Circuit Layer Operations Per Second), reaching up to 43,680, across all publicly available platforms.

Fast, collimated jets are ubiquitous features of young stellar objects (YSOs). They are generally thought to be powered by disk accretion, but the details are debated. Through 2D (axisymmetric) MHD simulations, we find that a fast ($>100$~km/s) collimated bipolar jet is continuously driven along the north and south poles of the circumstellar disk that is initially magnetized by a large-scale open poloidal field and contains a thermally ionized inner magnetically active zone surrounded by a dead zone. The fast jet is primarily driven magneto-centrifugally by the release of the gravitational binding energy of the so-called ``avalanche accretion streams" near the boundary of an evacuated poloidal field-dominated polar region and a thick disk atmosphere raised by a toroidal magnetic field. Specifically, the fast outflow is driven along the upper (open) branch of the highly pinched poloidal field lines threading the (strongly magnetically braked) accretion streams where the density is relatively low so that the lightly loaded material can be accelerated magneto-centrifugally along the open field line to a high speed. The highly pinched poloidal magnetic fields threading the avalanche accretion streams tend to reconnect, enabling mass to accrete to the center without dragging along the poloidal magnetic flux with it. The reconnection provides a potential heating source for producing chondrules and calcium- and aluminum-rich inclusions (CAIs).

Recent advancements in generative AI, particularly Latent Diffusion Models (LDMs), have revolutionized image synthesis and manipulation. However, these generative techniques raises concerns about data misappropriation and intellectual property infringement. Adversarial attacks on machine learning models have been extensively studied, and a well-established body of research has extended these techniques as a benign metric to prevent the underlying misuse of generative AI. Current approaches to safeguarding images from manipulation by LDMs are limited by their reliance on model-specific knowledge and their inability to significantly degrade semantic quality of generated images. In response to these shortcomings, we propose the Posterior Collapse Attack (PCA) based on the observation that VAEs suffer from posterior collapse during training. Our method minimizes dependence on the white-box information of target models to get rid of the implicit reliance on model-specific knowledge. By accessing merely a small amount of LDM parameters, in specific merely the VAE encoder of LDMs, our method causes a substantial semantic collapse in generation quality, particularly in perceptual consistency, and demonstrates strong transferability across various model architectures. Experimental results show that PCA achieves superior perturbation effects on image generation of LDMs with lower runtime and VRAM. Our method outperforms existing techniques, offering a more robust and generalizable solution that is helpful in alleviating the socio-technical challenges posed by the rapidly evolving landscape of generative AI., Comment: 21 pages, 7 figures, 10 tables

Metasurfaces, with their capability to control all possible dimensions of light, have become integral to quantum optical applications, including quantum state generation, operation, and tomography. In this work, we utilize a metasurface to generate polarization-hologram hybrid entanglement between a signal-idler photon pair to construct a quantum hologram. The properties of the quantum hologram can be revealed by collapsing the polarization degree of freedom of the idler photon, inducing interference between two holographic states of the signal photon, as a meaningful and selective erasure of the holographic content. In contrary, interference disappears when the idler photon is detected without observing polarization. This process can be further interpreted as a quantum holographic eraser, where the erasing action is visualized with erased contents in holograms. Our construction of polarization-hologram hybrid entangled state with metasurfaces will be useful for quantum communication with enhanced robustness, anti-counterfeiting applications through the additional quantum degrees of freedom, and as an emerging platform for exploring fundamental quantum concepts for entanglement and non-locality., Comment: 12 pages, 6 figures

Chirality is a property of a gapped phase of matter in two spatial dimensions that can be manifested through non-zero thermal or electrical Hall conductance. In this paper, we prove two no-go theorems that forbid such chirality for a quantum state in a finite dimensional local Hilbert space with strict area law entanglement entropies. As a crucial ingredient in the proofs, we introduce a new quantum information-theoretic primitive called instantaneous modular flow, which has many other potential applications., Comment: 5+9 pages, 4 figures

GQ Lup B is one of the few substellar companions with a detected cicumplanetary disk, or CPD. Observations of the CPD suggest the presence of a cavity, possibly formed by an exo-satellite. Using the Keck Planet Imager and Characterizer (KPIC), a high contrast imaging suite that feeds a high resolution spectrograph (1.9-2.5 microns, R$\sim$35,000), we present the first dedicated radial velocity (RV) observations around a high-contrast, directly imaged substellar companion, GQ Lup B, to search for exo-satellites. Over 11 epochs, we find a best and median RV error of 400-1000 m/s, most likely limited by systematic fringing in the spectra due to transmissive optics within KPIC. With this RV precision, KPIC is sensitive to exomoons 0.6-2.8% the mass of GQ Lup B ($\sim 30 M_{\text{Jup}}$) at separations between the Roche limit and $65 R_{\text{Jup}}$, or the extent of the cavity inferred within the CPD detected around GQ Lup B. Using simulations of HISPEC, a high resolution infrared spectrograph planned to debut at W.M. Keck Observatory in 2026, we estimate future exomoon sensitivity to increase by over an order of magnitude, providing sensitivity to less massive satellites potentially formed within the CPD itself. Additionally, we run simulations to estimate the amount of material that different masses of satellites could clear in a CPD to create the observed cavity. We find satellite-to-planet mass ratios of $q > 2 \times 10^{-4}$ can create observable cavities and report a maximum cavity size of $\sim 51 \, R_{\text{Jup}}$ carved from a satellite., Comment: 15 pages, 5 figures

Over the past decade, there has been extensive work in developing integrated silicon photonics (SiPh) gratings for the optical addressing of trapped ion qubits in the ion trap quantum computing community. However, when viewing beam profiles from infrared (IR) cameras, it is often difficult to determine the corresponding heights where the beam profiles are located. In this work, we developed transformer models to recognize the corresponding height categories of beam profiles of light from SiPh gratings. The model is trained using two techniques: (1) input patches, and (2) input sequence. For model trained with input patches, the model achieved recognition accuracy of 0.938. Meanwhile, model trained with input sequence shows lower accuracy of 0.895. However, when repeating the model-training 150 cycles, model trained with input patches shows inconsistent accuracy ranges between 0.445 to 0.959, while model trained with input sequence exhibit higher accuracy values between 0.789 to 0.936. The obtained outcomes can be expanded to various applications, including auto-focusing of light beam and auto-adjustment of z-axis stage to acquire desired beam profiles.

The Keck Planet Imager and Characterizer (KPIC) combines high contrast imaging with high resolution spectroscopy (R$\sim$35,000 in K band) to study directly imaged exoplanets and brown dwarfs in unprecedented detail. KPIC aims to spectrally characterize substellar companions through measurements of planetary radial velocities, spins, and atmospheric composition. Currently, the dominant source of systematic noise for KPIC is fringing, or oscillations in the spectrum as a function of wavelength. The fringing signal can dominate residuals by up to 10% of the continuum for high S/N exposures, preventing accurate wavelength calibration, retrieval of atmospheric parameters, and detection of planets with flux ratios less than 1% of the host star. To combat contamination from fringing, we first identify its three unique sources and adopt a physically informed model of Fabry-P\'{e}rot cavities to apply to post-processed data. We find this strategy can effectively model the fringing in observations of A0V/F0V stars, reducing the residual systematics caused by fringing by a factor of 2. Next, we wedge two of the transmissive optics internal to KPIC to eliminate two sources of fringing and confirm the third source as the entrance window to the spectrograph. Finally, we apply our previous model of the Fabry-P\'{e}rot cavity to new data taken with the wedged optics to reduce the amplitude of the residuals by a factor of 10., Comment: 13 pages, 6 figures

Endorectal ultrasound (ERUS) is an important imaging modality that provides high reliability for diagnosing the depth and boundary of invasion in colorectal cancer. However, the lack of a large-scale ERUS dataset with high-quality annotations hinders the development of automatic ultrasound diagnostics. In this paper, we collected and annotated the first benchmark dataset that covers diverse ERUS scenarios, i.e. colorectal cancer segmentation, detection, and infiltration depth staging. Our ERUS-10K dataset comprises 77 videos and 10,000 high-resolution annotated frames. Based on this dataset, we further introduce a benchmark model for colorectal cancer segmentation, named the Adaptive Sparse-context TRansformer (ASTR). ASTR is designed based on three considerations: scanning mode discrepancy, temporal information, and low computational complexity. For generalizing to different scanning modes, the adaptive scanning-mode augmentation is proposed to convert between raw sector images and linear scan ones. For mining temporal information, the sparse-context transformer is incorporated to integrate inter-frame local and global features. For reducing computational complexity, the sparse-context block is introduced to extract contextual features from auxiliary frames. Finally, on the benchmark dataset, the proposed ASTR model achieves a 77.6% Dice score in rectal cancer segmentation, largely outperforming previous state-of-the-art methods.

In this paper, we derive a thermodynamically consistent non-isothermal diffuse interface model for phase transition and interface evolution involving heat transfer. This model is constructed by integrating concepts from classical irreversible thermodynamics with the energetic variational approach. By making specific assumptions about the kinematics of the temperature, we derive a non-isothermal Allen-Cahn equation. Through both asymptotic analysis and numerical simulations, we demonstrate that in the sharp interface limit, the non-isothermal Allen-Cahn equation converges to a two-phase Stefan-like problem, under a certain scaling of the melting/freezing energy. In this regime, the motion of the liquid-solid interface and the temperature interface coincide and are governed by the mean curvature, at least for a short time. The result provides a justification for the classical Stefan problem within a certain physical regime., Comment: 23 pages, 5 figures

Whole slide images (WSIs) are gigapixel-scale digital images of H\&E-stained tissue samples widely used in pathology. The substantial size and complexity of WSIs pose unique analytical challenges. Multiple Instance Learning (MIL) has emerged as a powerful approach for addressing these challenges, particularly in cancer classification and detection. This survey provides a comprehensive overview of the challenges and methodologies associated with applying MIL to WSI analysis, including attention mechanisms, pseudo-labeling, transformers, pooling functions, and graph neural networks. Additionally, it explores the potential of MIL in discovering cancer cell morphology, constructing interpretable machine learning models, and quantifying cancer grading. By summarizing the current challenges, methodologies, and potential applications of MIL in WSI analysis, this survey aims to inform researchers about the state of the field and inspire future research directions.

The ferroelectric domain structure of a freestanding BiFeO$_3$ film was visualized by ptychographic dichroic imaging with linearly polarized X-rays at the O K-edge around 530 eV. The dichroic contrast is maximized at the energy of the hybridization of the O 2p state and the Fe 3d orbitals, which is split by the octahedral crystal field of the perovskite structure. The thus obtained microscopy images compliment the ptychographic imaging of the antiferromagnetic contribution at the Fe L$_3$-edge. The approach is extendible to the separation of different ferroic contributions in other multiferroic oxides.

In this paper, we study user association and wireless bandwidth allocation for a hierarchical federated learning system that consists of mobile users, edge servers, and a cloud server. To minimize the length of a global round in hierarchical federated learning with equal bandwidth allocation, we formulate a combinatorial optimization problem. We design the twin sorting dynamic programming (TSDP) algorithm that obtains a globally optimal solution in polynomial time when there are two edge servers. In addition, we put forward the TSDP-assisted algorithm for user association when there are three or more edge servers. Furthermore, given a user association matrix, we formulate and solve a convex optimization problem for optimal wireless bandwidth allocation. Simulation results show that the proposed approach outperforms a number of alternative schemes., Comment: 14 pages

NLP-assisted solutions have gained considerable traction to support qualitative data analysis. However, there does not exist a unified evaluation framework that can account for the many different settings in which qualitative researchers may employ them. In this paper, we take a first step in this direction by proposing an evaluation framework to study the way in which different tools may result in different outcomes depending on the collaboration strategy employed. Specifically, we study the impact of synchronous vs. asynchronous collaboration using two different NLP-assisted qualitative research tools and present a comprehensive analysis of significant differences in the consistency, cohesiveness, and correctness of their outputs.

Multiterminal Josephson junctions are a promising platform to host synthetic topological phases of matter and Floquet states. However, the energy scales governing topological protection in these devices are on the order of the spacing between Andreev bound states. Recent theories suggest that similar phenomena may instead be explored in circuits composed of two-terminal Josephson junctions, allowing for the topological protection to be controlled by the comparatively large Josephson energy. Here, we explore a Josephson circuit, in which three superconducting electrodes are connected through Josephson junctions to a common superconducting island. We demonstrate the dynamic generation of multiplet resonances, which have previously been observed in multiterminal Josephson junctions. The multiplets are found to be robust to elevated temperatures and are confirmed by exhibiting the expected Shapiro step quantization under a microwave drive. We also find an unexpected novel supercurrent, which couples a pair of contacts that are both voltage-biased with respect to the common superconducting island. We show that this supercurrent results from synchronization of the phase dynamics and pose the question whether it should also carry a topological contribution.

Sequential recommendation systems fundamentally rely on users' historical interaction sequences, which are often contaminated by noisy interactions. Identifying these noisy interactions accurately without additional information is particularly difficult due to the lack of explicit supervisory signals to denote noise. Large Language Models (LLMs), equipped with extensive open knowledge and semantic reasoning abilities, present a promising avenue to bridge this information gap. However, employing LLMs for denoising in sequential recommendation introduces notable challenges: 1) Direct application of pretrained LLMs may not be competent for the denoising task, frequently generating nonsensical responses; 2) Even after fine-tuning, the reliability of LLM outputs remains questionable, especially given the complexity of the task and th inherent hallucinatory issue of LLMs. To tackle these challenges, we propose LLM4DSR, a tailored approach for denoising sequential recommendation using LLMs. We constructed a self-supervised fine-tuning task to activate LLMs' capabilities to identify noisy items and suggest replacements. Furthermore, we developed an uncertainty estimation module that ensures only high-confidence responses are utilized for sequence corrections. Remarkably, LLM4DSR is model-agnostic, allowing the corrected sequences to be flexibly applied across various recommendation models. Extensive experiments validate the superiority of LLM4DSR over existing methods across three datasets and three recommendation backbones.

Recent research on knowledge distillation has increasingly focused on logit distillation because of its simplicity, effectiveness, and versatility in model compression. In this paper, we introduce Refined Logit Distillation (RLD) to address the limitations of current logit distillation methods. Our approach is motivated by the observation that even high-performing teacher models can make incorrect predictions, creating a conflict between the standard distillation loss and the cross-entropy loss. This conflict can undermine the consistency of the student model's learning objectives. Previous attempts to use labels to empirically correct teacher predictions may undermine the class correlation. In contrast, our RLD employs labeling information to dynamically refine teacher logits. In this way, our method can effectively eliminate misleading information from the teacher while preserving crucial class correlations, thus enhancing the value and efficiency of distilled knowledge. Experimental results on CIFAR-100 and ImageNet demonstrate its superiority over existing methods. The code is provided at \text{https://github.com/zju-SWJ/RLD}., Comment: 11 pages, 7 figures

Due to the high noise caused by solar background radiation, the existing satellite-based free-space quantum key distribution (QKD) experiments are mainly carried out at night, hindering the establishment of a practical all-day real-time global-scale quantum network. Given that the 3-5 {\mu}m mid-infrared (MIR) band has extremely low solar background radiation and strong scattering resistance, it is one of the ideal bands for free-space quantum communication. Here, firstly, we report on the preparation of a high-quality MIR (3370 nm) and telecom band (1555 nm) two-color polarization-entangled photon source, then we use this source to realize a principle QKD based on free-space and fiber hybrid channels in a laboratory. The theoretical analysis clearly shows that a long-distance QKD over 500 km of free-space and 96 km of fiber hybrid channels can be reached simultaneously. This work represents a significant step toward developing all-day global-scale quantum communication networks., Comment: 24 pages, 9 figures

Fine-tuning large language models (LLMs) with high parameter efficiency for downstream tasks has become a new paradigm. Low-Rank Adaptation (LoRA) significantly reduces the number of trainable parameters for fine-tuning. Although it has demonstrated commendable performance, updating parameters within a single scale may not be the optimal choice for complex downstream tasks.In this paper, we extend the LoRA to multiple scales, dubbed as LoRA$^2$. We first combine orthogonal projection theory to train a set of LoRAs in two mutually orthogonal planes. Then, we improve the importance score algorithm, which reduce parameter sensitivity score calculations by approximately 98.5\%. By pruning singular values with lower importance scores, thereby enhancing adaptability to various downstream tasks. Extensive experiments are conducted on two widely used pre-trained models to validate the effectiveness of LoRA$^2$. Results show that it significantly reduces the number of trainable parameters to just 0.72\% compared to full fine-tuning, while still delivering highly impressive performance. Even when the parameters are further reduced to 0.17M, it still achieves comparable results to the baseline with 8 times more parameters. Our code is available here: https://anonymous.4open.science/r/LoRA-2-5B4C

Elucidating gene function is one of the ultimate goals of plant science. Despite this, only ~15% of all genes in the model plant Arabidopsis thaliana have comprehensively experimentally verified functions. While bioinformatical gene function prediction approaches can guide biologists in their experimental efforts, neither the performance of the gene function prediction methods nor the number of experimental characterisation of genes has increased dramatically in recent years. In this review, we will discuss the status quo and the trajectory of gene function elucidation and outline the recent advances in gene function prediction approaches. We will then discuss how recent artificial intelligence advances in large language models and knowledge graphs can be leveraged to accelerate gene function predictions and keep us updated with scientific literature.

Web-based Large Language Model (LLM) services have been widely adopted and have become an integral part of our Internet experience. Third-party plugins enhance the functionalities of LLM by enabling access to real-world data and services. However, the privacy consequences associated with these services and their third-party plugins are not well understood. Sensitive prompt data are stored, processed, and shared by cloud-based LLM providers and third-party plugins. In this paper, we propose Casper, a prompt sanitization technique that aims to protect user privacy by detecting and removing sensitive information from user inputs before sending them to LLM services. Casper runs entirely on the user's device as a browser extension and does not require any changes to the online LLM services. At the core of Casper is a three-layered sanitization mechanism consisting of a rule-based filter, a Machine Learning (ML)-based named entity recognizer, and a browser-based local LLM topic identifier. We evaluate Casper on a dataset of 4000 synthesized prompts and show that it can effectively filter out Personal Identifiable Information (PII) and privacy-sensitive topics with high accuracy, at 98.5% and 89.9%, respectively.

Quantum walk (QW) utilizes its internal quantum states to decide the displacement, thereby introducing single-particle entanglement between the internal and positional degrees of freedom. By simulating three variants of QW with the conventional, symmetric, and split-step translation operators with or without classical randomness in the coin operator, we show the entanglement is robust against both time- and spatially- dependent randomness, which can cause localization transitions of QW. We propose a classical quantity call overlap, which literally measures the overlap between the probability distributions of the internal states, as a proxy of entanglement. The overlap captures the inverse behavior of the entanglement entropy in most cases, which can be explained by analyzing the structure of the total wave function of the walker. We test the limitation of the classical proxy by constructing a special case with high population imbalance between the internal states to blind the overlap. Possible implications and experimental measurements are also discussed., Comment: 13 pages, 10 figures

Meta-learning has been extensively applied in the domains of few-shot learning and fast adaptation, achieving remarkable performance. While Meta-learning methods like Model-Agnostic Meta-Learning (MAML) and its variants provide a good set of initial parameters for the model, the model still tends to learn shortcut features, which leads to poor generalization. In this paper, we propose the formal conception of "learn to learn more precisely", which aims to make the model learn precise target knowledge from data and reduce the effect of noisy knowledge, such as background and noise. To achieve this target, we proposed a simple and effective meta-learning framework named Meta Self-Distillation(MSD) to maximize the consistency of learned knowledge, enhancing the models' ability to learn precise target knowledge. In the inner loop, MSD uses different augmented views of the same support data to update the model respectively. Then in the outer loop, MSD utilizes the same query data to optimize the consistency of learned knowledge, enhancing the model's ability to learn more precisely. Our experiment demonstrates that MSD exhibits remarkable performance in few-shot classification tasks in both standard and augmented scenarios, effectively boosting the accuracy and consistency of knowledge learned by the model., Comment: 10pages,4 figures, meta learning

We consider a novel active learning problem motivated by the need of learning machine learning models for health monitoring in wireless body area network (WBAN). Due to the limited resources at body sensors, collecting each unlabeled sample in WBAN incurs a nontrivial cost. Moreover, training health monitoring models typically requires labels indicating the patient's health state that need to be generated by healthcare professionals, which cannot be obtained at the same pace as data collection. These challenges make our problem fundamentally different from classical active learning, where unlabeled samples are free and labels can be queried in real time. To handle these challenges, we propose a two-phased active learning method, consisting of an online phase where a coreset construction algorithm is proposed to select a subset of unlabeled samples based on their noisy predictions, and an offline phase where the selected samples are labeled to train the target model. The samples selected by our algorithm are proved to yield a guaranteed error in approximating the full dataset in evaluating the loss function. Our evaluation based on real health monitoring data and our own experimentation demonstrates that our solution can drastically save the data curation cost without sacrificing the quality of the target model.

Geometrically local quantum codes, comprised of qubits and checks embedded in $\mathbb{R}^D$ with local check operators, have been a subject of significant interest. A key challenge is identifying the optimal code construction that maximizes both dimension and distance. Recent advancements have produced several constructions, but these either depend on specific good quantum low-density parity-check (qLDPC) codes or are limited to three dimensions. In this work, we introduce a construction that can transform any good qLDPC code into an optimal geometrically local quantum code. Our approach hinges on a novel procedure that extracts a two-dimensional structure from an arbitrary three-term chain complex. We expect that this procedure will find broader applications in areas such as weight reduction and the geometric realization of chain complexes., Comment: 25 pages, 15 figures

Recently, pre-trained model and efficient parameter tuning have achieved remarkable success in natural language processing and high-level computer vision with the aid of masked modeling and prompt tuning. In low-level computer vision, however, there have been limited investigations on pre-trained models and even efficient fine-tuning strategy has not yet been explored despite its importance and benefit in various real-world tasks such as alleviating memory inflation issue when integrating new tasks on AI edge devices. Here, we propose a novel efficient parameter tuning approach dubbed contribution-based low-rank adaptation (CoLoRA) for multiple image restorations along with effective pre-training method with random order degradations (PROD). Unlike prior arts that tune all network parameters, our CoLoRA effectively fine-tunes small amount of parameters by leveraging LoRA (low-rank adaptation) for each new vision task with our contribution-based method to adaptively determine layer by layer capacity for that task to yield comparable performance to full tuning. Furthermore, our PROD strategy allows to extend the capability of pre-trained models with improved performance as well as robustness to bridge synthetic pre-training and real-world fine-tuning. Our CoLoRA with PROD has demonstrated its superior performance in various image restoration tasks across diverse degradation types on both synthetic and real-world datasets for known and novel tasks., Comment: 33 pages, 15 figures, for homepage see this url : https://janeyeon.github.io/colora/

This work presents a Bayesian inference study for relativistic heavy-ion collisions in the Beam Energy Scan program at the Relativistic Heavy-Ion Collider. The theoretical model simulates event-by-event (3+1)D collision dynamics using hydrodynamics and hadronic transport theory. We analyze the model's 20-dimensional posterior distributions obtained using three model emulators with different accuracy and demonstrate the essential role of training an accurate model emulator in the Bayesian analysis. Our analysis provides robust constraints on the Quark-Gluon Plasma's transport properties and various aspects of (3+1)D relativistic nuclear dynamics. By running full model simulations with 100 parameter sets sampled from the posterior distribution, we make predictions for $p_{\rm T}$-differential observables and estimate their systematic theory uncertainty. A sensitivity analysis is performed to elucidate how individual experimental observables respond to different model parameters, providing useful physics insights into the phenomenological model for heavy-ion collisions., Comment: 18 pages, 9 figures

In this work we are going to establish H\"older continuity of harmonic maps from an open set $\Omega$ in an ${\rm RCD}(K,N)$ space valued into a ${\rm CAT}(\kappa)$ space, with the constraint that the image of $\Omega$ via the map is contained in a sufficiently small ball in the target. Building on top of this regularity and assuming a local Lipschitz regularity of the map, we establish a weak version of the Bochner-Eells-Sampson inequality in such a non-smooth setting. Finally we study the boundary regularity of such maps., Comment: Comments are welcome!

Dynamical heterogeneity in which transitory local fluctuations occur in the conformation and dynamics of constituent particles, is essential for evolution of supercooled liquids into the glass state. Yet its microscopic spatiotemporal phenomenology has remained unobservable in virtually all supercooled glass forming liquids. Recent theoretical advances predict that corresponding dynamical heterogeneity could also occur in supercooled magnetic monopole fluids. Motivated thus, we searched for dynamical heterogeneity when entering the supercooled monopole fluid of Dy2Ti2O7. By measuring microsecond-resolved spontaneous magnetization noise M(t,T) at temperatures between 15 mK

This paper studies the structure of graphs with given tree-width and excluding a fixed complete bipartite subgraph, which generalises the bounded degree setting. We give a new structural description of such graphs in terms of so-called quasi-tree-partitions. We demonstrate the utility of this result through applications to (fractional) clustered colouring. Further generalisations of these structural and colouring results are presented.

Projected wave functions offer a means for incorporating local correlation effects in gapless electronic phases of matter like metals. Although such wave functions can be readily specified formally, it is challenging to compute their associated physical observables. Tensor network approaches offer a modern numerical method for this task. In this work, we develop and demonstrate a constructive tensor-network approach for obtaining physical quantities, like fermion two-point functions and density-density correlation functions, for one-dimensional projected Fermi sea states. We benchmark our method against exact analytical results for spin-1/2 electrons subjected to the Gutzwiller projection, and then present results on spinless fermions with nearest-neighbor repulsion. For a state with two pairs of Fermi points, we reveal a correlation-tuning of the characteristic wave vector of the charge density modulation in the system., Comment: 17 pages, 10 figures; version 2: Typos corrected and references updated

Studying and controlling quantum many-body interactions is fundamentally important for quantum science and related emerging technologies. Optically addressable solid-state spins offer a promising platform for exploring various quantum many-body phenomena due to their scalability to a large Hilbert space. However, it is often challenging to probe many-body dynamics in solid-state spin systems due to large on-site disorder and undesired coupling to the environment. Here, we investigate an optically addressable solid-state spin system comprising a strongly interacting ensemble of millions of ytterbium-171 ions in a crystal. Notably, this platform features a clock transition that gives rise to pure long-range spin-exchange interactions, termed the dipolar XY model. Leveraging this unique feature, we investigate quantum thermalization by varying the relative ratio of interaction strength to disorder, dynamically engineering the XY model into other many-body Hamiltonian models, and realizing a time-crystalline phase of matter through periodic driving. Our findings indicate that an ensemble of rare-earth ions serves as a versatile testbed for many-body physics and offers valuable insights for advancing quantum technologies.

Young, self-luminous super-Jovian companions discovered by direct imaging provide a challenging test of planet formation and evolution theories. By spectroscopically characterizing the atmospheric compositions of these super-Jupiters, we can constrain their formation histories. Here we present studies of the recently discovered HIP 99770 b, a 16 MJup high-contrast companion on a 17 au orbit, using the fiber-fed high-resolution spectrograph KPIC (R~35,000) on the Keck II telescope. Our K-band observations led to detections of H2O and CO in the atmosphere of HIP 99770 b. We carried out free retrieval analyses using petitRADTRANS to measure its chemical abundances, including the metallicity and C/O ratio, projected rotation velocity (vsini), and radial velocity (RV). We found that the companion's atmosphere has C/O=0.55(-0.04/+0.06) and [M/H]=0.26(-0.23/+0.24) (1{\sigma} confidence intervals), values consistent with those of the Sun and with a companion formation via gravitational instability or core accretion. The projected rotation velocity < 7.8 km/s is small relative to other directly imaged companions with similar masses and ages. This may imply a near pole-on orientation or effective magnetic braking by a circumplanetary disk. In addition, we added the companion-to-primary relative RV measurement to the orbital fitting and obtained updated constraints on orbital parameters. Detailed characterization of super-Jovian companions within 20 au like HIP 99770 b is critical for understanding the formation histories of this population., Comment: 18 pages, 5 figures, accepted to AJ

Let $\{\rho_{\ell}:\mathrm{Gal}_K\to\mathrm{GL}_n(\mathbb{Q}_{\ell})\}_{\ell}$ be a semisimple compatible system of $\ell$-adic representations of a number field $K$ that is arising from geometry. Let $\textbf{G}_{\ell}\subset\mathrm{GL}_{n,\mathbb{Q}_{\ell}}$ and $\widehat{\underline{G_{\ell}}}\subset\mathrm{GL}_{n,\mathbb{F}_\ell}$ be respectively the algebraic monodromy group and full algebraic envelope of $\rho_{\ell}$. We prove that there is a natural isomorphism between the component groups $\pi_0(\textbf{G}_{\ell}) \simeq \pi_0(\widehat{\underline{G_\ell}})$ for all sufficiently large $\ell$., Comment: 9 pages

Non-Bayesian social learning enables multiple agents to conduct networked signal and information processing through observing environmental signals and information aggregating. Traditional non-Bayesian social learning models only consider single signals, limiting their applications in scenarios where multiple viewpoints of information are available. In this work, we exploit, in the information aggregation step, the independently learned results from observations taken from multiple viewpoints and propose a novel non-Bayesian social learning model for scenarios with multiview observations. We prove the convergence of the model under traditional assumptions and provide convergence conditions for the algorithm in the presence of misleading signals. Through theoretical analyses and numerical experiments, we validate the strong reliability and robustness of the proposed algorithm, showcasing its potential for real-world applications.

This paper investigates the capabilities and effectiveness of backward sensing centered on reconfigurable intelligent surfaces (RISs). We demonstrate that the direction of arrival (DoA) estimation of incident waves in the far-field regime can be accomplished using a single RIS by leveraging configurational diversity. Furthermore, we identify that the spatial diversity achieved through deploying multiple RISs enables accurate localization of multiple power sources. Physically accurate and mathematically concise models are introduced to characterize forward signal aggregations via RISs. By employing linearized approximations inherent in the far-field region, the measurement process for various configurations can be expressed as a system of linear equations. The mathematical essence of backward sensing lies in solving this system. A theoretical framework for determining key performance indicators is established through condition number analysis of the sensing operators. In the context of localization using multiple RISs, we examine relationships among the rank of sensing operators, the size of the region of interest (RoI), and the number of elements and measurements. For DoA estimations, we provide an upper bound for the relative error of the least squares reconstruction algorithm. These quantitative analyses offer essential insights for system design and optimization. Numerical experiments validate our findings. To demonstrate the practicality of our proposed RIS-centric sensing approach, we develop a proof-of-concept prototype using universal software radio peripherals (USRP) and employ a magnitude-only reconstruction algorithm tailored for this system. To our knowledge, this represents the first trial of its kind., Comment: 13 pages

Principal component analysis (PCA) is a widely used dimension reduction method, but its performance is known to be non-robust to outliers. Recently, product-PCA (PPCA) has been shown to possess the efficiency-loss free ordering-robustness property: (i) in the absence of outliers, PPCA and PCA share the same asymptotic distributions; (ii), in the presence of outliers, PPCA is more ordering-robust than PCA in estimating the leading eigenspace. PPCA is thus different from the conventional robust PCA methods, and may deserve further investigations. In this article, we study the high-dimensional statistical properties of the PPCA eigenvalues via the techniques of random matrix theory. In particular, we derive the critical value for being distant spiked eigenvalues, the limiting values of the sample spiked eigenvalues, and the limiting spectral distribution of PPCA. Similar to the case of PCA, the explicit forms of the asymptotic properties of PPCA become available under the special case of the simple spiked model. These results enable us to more clearly understand the superiorities of PPCA in comparison with PCA. Numerical studies are conducted to verify our results., Comment: 20 pages, 3 figures, 1 table

Electroencephalography (EEG) is a critical tool in neuroscience and clinical practice for monitoring and analyzing brain activity. Traditional neural network models, such as EEGNet, have achieved considerable success in decoding EEG signals but often struggle with the complexity and high dimensionality of the data. Recent advances in quantum computing present new opportunities to enhance machine learning models through quantum machine learning (QML) techniques. In this paper, we introduce Quantum-EEGNet (QEEGNet), a novel hybrid neural network that integrates quantum computing with the classical EEGNet architecture to improve EEG encoding and analysis, as a forward-looking approach, acknowledging that the results might not always surpass traditional methods but it shows its potential. QEEGNet incorporates quantum layers within the neural network, allowing it to capture more intricate patterns in EEG data and potentially offering computational advantages. We evaluate QEEGNet on a benchmark EEG dataset, BCI Competition IV 2a, demonstrating that it consistently outperforms traditional EEGNet on most of the subjects and other robustness to noise. Our results highlight the significant potential of quantum-enhanced neural networks in EEG analysis, suggesting new directions for both research and practical applications in the field., Comment: 7 pages, 4 figures

The facility adaptive optics of the Subaru Telescope AO188 recently received some long-awaited upgrades: a new 3224-actuator deformable mirror (DM) from ALPAO (hence the name change to AO3000 or AO3k), an upgraded GPU-based real-time computer, a visible nonlinear curvature wavefront sensor and a near-infrared wavefront sensor (NIR WFS), closing the loop at up to 2~kHz. The wavefront sensors were added in 2023, while the DM will be installed at the beginning of 2024. With these new features, AO3k will provide extreme-AO level of correction to all the instruments on the IR Nasmyth platform: The NIR-MIR camera and spectrograph IRCS, the high-resolution Doppler spectrograph IRD, and the high-contrast instrument SCExAO. AO3k will also support laser tomography (LTAO), delivering high Strehl ratio imaging with large sky coverage. The high Strehl will especially benefit SCExAO for high-contrast imaging, both in infrared and visible. The second stage extreme AO will no longer have to chase large residual atmospheric turbulence, and will focus on truly high-contrast techniques to create and stabilize dark holes, as well as coherent differential imaging techniques. We will finally be able to leverage the several high performance coronagraphs tested in SCExAO, even in the visible. AO3k will answer crucial questions as a precursor for future adaptive optics systems for ELTs, especially as a technology demonstrator for the HCI Planetary Systems Imager on the Thirty Meter Telescope. A lot of questions are still unanswered on the on-sky behavior of high actuator counts DMs, NIR wavefront sensing, the effect of rolling shutters or persistence. We present here the first on-sky results of AO3k, before the system gets fully offered to the observers in the second half of 2024. These results give us some insight on the great scientific results we hope to achieve in the future., Comment: 15 pages, 11 figures, proceedings of the SPIE Astronomical Telescopes and Instrumentation 2024 (Yokohama, Japan)

Quantum private query (QPQ) is the quantum version for symmetrically private retrieval. However, the user privacy in QPQ is generally guarded in the non-realtime and cheat sensitive way. That is, the dishonest database holder's cheating to elicit user privacy can only be discovered after the protocol is finished (when the user finds some errors in the retrieved database item). Such delayed detection may cause very unpleasant results for the user in real-life applications. Current efforts to protect user privacy in realtime in existing QPQ protocols mainly use two techniques, i.e., adding an honesty checking on the database or allowing the user to reorder the qubits. We reexamine these two kinds of QPQ protocols and find neither of them can work well. We give concrete cheating strategies for both participants and show that honesty checking of inner participant should be dealt more carefully in for example the choosing of checking qubits. We hope such discussion can supply new concerns when detection of dishonest participant is considered in quantum multi-party secure computations.

The legal landscape encompasses a wide array of lawsuit types, presenting lawyers with challenges in delivering timely and accurate information to clients, particularly concerning critical aspects like potential imprisonment duration or financial repercussions. Compounded by the scarcity of legal experts, there's an urgent need to enhance the efficiency of traditional legal workflows. Recent advances in deep learning, especially Large Language Models (LLMs), offer promising solutions to this challenge. Leveraging LLMs' mathematical reasoning capabilities, we propose a novel approach integrating LLM-based methodologies with specially designed prompts to address precision requirements in legal Artificial Intelligence (LegalAI) applications. The proposed work seeks to bridge the gap between traditional legal practices and modern technological advancements, paving the way for a more accessible, efficient, and equitable legal system. To validate this method, we introduce a curated dataset tailored to precision-oriented LegalAI tasks, serving as a benchmark for evaluating LLM-based approaches. Extensive experimentation confirms the efficacy of our methodology in generating accurate numerical estimates within the legal domain, emphasizing the role of LLMs in streamlining legal processes and meeting the evolving demands of LegalAI., Comment: The paper has been accepted by the 33rd ACM International Conference on Information and Knowledge Management (CIKM) in 2024

In this paper, we present an enhanced semi-analytic method for calculating quasinormal excitation factors in the eikonal regime, specifically for Schwarzschild black holes. To achieve improved accuracy in our quasinormal mode computations, we extend the Dolan and Ottewill inverse multipolar expansion technique and incorporate higher-order corrections from the WKB method of Iyer and Will. Our approach is carried out to a higher order than previous methods, thereby reducing the relative error, particularly for lower multipolar numbers. We validate our results by comparing them with those obtained using the Mano, Suzuki, and Takasugi method, demonstrating excellent agreement. A key advantage of our method is its ability to extract quasinormal excitation factors, which are crucial for accurately modeling gravitational wave signals from binary black hole mergers. This advancement provides a useful tool for future gravitational wave studies, enabling better quantification of quasinormal mode excitations and more precise identification of individual modes during black hole ringdowns., Comment: 32 pages, 4 figures, 3 tables, and 2 appendices

Amid the proliferation of forged images, notably the tsunami of deepfake content, extensive research has been conducted on using artificial intelligence (AI) to identify forged content in the face of continuing advancements in counterfeiting technologies. We have investigated the use of AI to provide the original authentic image after deepfake detection, which we believe is a reliable and persuasive solution. We call this "image-based automated fact verification," a name that originated from a text-based fact-checking system used by journalists. We have developed a two-phase open framework that integrates detection and retrieval components. Additionally, inspired by a dataset proposed by Meta Fundamental AI Research, we further constructed a large-scale dataset that is specifically designed for this task. This dataset simulates real-world conditions and includes both content-preserving and content-aware manipulations that present a range of difficulty levels and have potential for ongoing research. This multi-task dataset is fully annotated, enabling it to be utilized for sub-tasks within the forgery identification and fact retrieval domains. This paper makes two main contributions: (1) We introduce a new task, "image-based automated fact verification," and present a novel two-phase open framework combining "forgery identification" and "fact retrieval." (2) We present a large-scale dataset tailored for this new task that features various hand-crafted image edits and machine learning-driven manipulations, with extensive annotations suitable for various sub-tasks. Extensive experimental results validate its practicality for fact verification research and clarify its difficulty levels for various sub-tasks., Comment: Pages 1-13 are the main body of the paper, and pages 14-16 are the supplementary material