Search

Extended Reality (XR) provides a more intuitive interaction method for teleoperating robots compared to traditional 2D controls. Recent studies have laid the groundwork for usable teleoperation with XR, but it fails in tasks requiring rapid motion and precise manipulations due to the large delay between user motion and agent feedback. In this work, we profile the end-to-end latency in a state-of-the-art XR teleoperation system and propose our idea to optimize the latency by implementing a duo-virtual spaces design and localizing the agent and objects in the user-side virtual space, while calibrating with periodic ground-truth poses from the agent-side virtual space., Comment: 3 pages, 3 figures

We report the first detection of the halo ellipticities of galaxy clusters by applying the halo-shear-shear correlations (HSSC), without the necessity of major axis determination. We use the Fourier\_Quad shear catalog based on the Hyper Suprime-Cam Survey and the group catalog from the DESI Legacy Surveys for the measurement of group/cluster lensing and HSSC. Our analysis includes the off-centering effects. We obtain the average projected ellipticity of dark matter halos with mass $13.5 < {\rm log} (M_G h/ M_\odot) < 14.5$ within 1.3 virial radius to be $0.48^{+0.12}_{-0.19}$. We divide the sample into two groups based on mass and redshift, and we find that halos with higher mass tend to exhibit increased ellipticity. We also reveal that high-richness halos have larger ellipticities, confirming the physical picture from numerical simulation that high-richiness halos have a dynamical youth and more active mass accretion phase., Comment: 9 figures, 4 tables. Submitted to APJ

Conventional geometry-based SLAM systems lack dense 3D reconstruction capabilities since their data association usually relies on feature correspondences. Additionally, learning-based SLAM systems often fall short in terms of real-time performance and accuracy. Balancing real-time performance with dense 3D reconstruction capabilities is a challenging problem. In this paper, we propose a real-time RGB-D SLAM system that incorporates a novel view synthesis technique, 3D Gaussian Splatting, for 3D scene representation and pose estimation. This technique leverages the real-time rendering performance of 3D Gaussian Splatting with rasterization and allows for differentiable optimization in real time through CUDA implementation. We also enable mesh reconstruction from 3D Gaussians for explicit dense 3D reconstruction. To estimate accurate camera poses, we utilize a rotation-translation decoupled strategy with inverse optimization. This involves iteratively updating both in several iterations through gradient-based optimization. This process includes differentiably rendering RGB, depth, and silhouette maps and updating the camera parameters to minimize a combined loss of photometric loss, depth geometry loss, and visibility loss, given the existing 3D Gaussian map. However, 3D Gaussian Splatting (3DGS) struggles to accurately represent surfaces due to the multi-view inconsistency of 3D Gaussians, which can lead to reduced accuracy in both camera pose estimation and scene reconstruction. To address this, we utilize depth priors as additional regularization to enforce geometric constraints, thereby improving the accuracy of both pose estimation and 3D reconstruction. We also provide extensive experimental results on public benchmark datasets to demonstrate the effectiveness of our proposed methods in terms of pose accuracy, geometric accuracy, and rendering performance.

The transport properties (electrical resistivity, thermal conductivity, and viscosity) of iron-hydrogen alloy are of great significance in the stability and evolution of planetary magnetic fields. Here, we investigate the thermal transport properties of iron doped with varying hydrogen content as functions of pressure (P) and temperature (T) for the top and bottom of Earth's outer core and beyond, corresponding to pressures of about 130 to 300 GPa and temperatures of 4000 to 7000 K. Using first-principles density functional theory molecular dynamic simulations (FPMD), we verify that crystalline FeH$_x$ is superionic with H diffusing freely. We find a low frequency viscosity of 10-11 mPa$\cdot$s for liquid Fe-H alloys at Earth's outer core conditions by the linear response Green-Kubo formula. Using the KKR method within density functional theory (DFT) plus Dynamical mean-field Theory (DMFT), we find saturation of electrical resistivity with increasing temperatures in liquid iron at outer core conditions. The effect of H on electrical and thermal transport we find is small, so that the exact H content of the core is not needed. The primary effect of H is on the equation of state, decreasing the density at constant P and T. We find the Lorenz number is smaller than the ideal value, and obtain for X(H)= 0.20, or 0.45 wt% H , thermal conductivity $\kappa$ of $\sim$105 and $\sim$190 $Wm^{-1}K^{-1}$, respectively, at conditions near the core-mantle and inner-outer core boundary.

Existing Transformer-based RGBT trackers achieve remarkable performance benefits by leveraging self-attention to extract uni-modal features and cross-attention to enhance multi-modal feature interaction and template-search correlation computation. Nevertheless, the independent search-template correlation calculations ignore the consistency between branches, which can result in ambiguous and inappropriate correlation weights. It not only limits the intra-modal feature representation, but also harms the robustness of cross-attention for multi-modal feature interaction and search-template correlation computation. To address these issues, we propose a novel approach called Cross-modulated Attention Transformer (CAFormer), which performs intra-modality self-correlation, inter-modality feature interaction, and search-template correlation computation in a unified attention model, for RGBT tracking. In particular, we first independently generate correlation maps for each modality and feed them into the designed Correlation Modulated Enhancement module, modulating inaccurate correlation weights by seeking the consensus between modalities. Such kind of design unifies self-attention and cross-attention schemes, which not only alleviates inaccurate attention weight computation in self-attention but also eliminates redundant computation introduced by extra cross-attention scheme. In addition, we propose a collaborative token elimination strategy to further improve tracking inference efficiency and accuracy. Extensive experiments on five public RGBT tracking benchmarks show the outstanding performance of the proposed CAFormer against state-of-the-art methods., Comment: 10 pages, 5 figures

We focus on the problem of fusing two or more heterogeneous large language models (LLMs) to facilitate their complementary strengths. One of the challenges on model fusion is high computational load, i.e. to fine-tune or to align vocabularies via combinatorial optimization. To this end, we propose \emph{Cool-Fusion}, a simple yet effective approach that fuses the knowledge of heterogeneous source LLMs to leverage their complementary strengths. \emph{Cool-Fusion} is the first method that does not require any type of training like the ensemble approaches. But unlike ensemble methods, it is applicable to any set of source LLMs that have different vocabularies. The basic idea is to have each source LLM individually generate tokens until the tokens can be decoded into a text segment that ends at word boundaries common to all source LLMs. Then, the source LLMs jointly rerank the generated text segment and select the best one, which is the fused text generation in one step. Extensive experiments are conducted across a variety of benchmark datasets. On \emph{GSM8K}, \emph{Cool-Fusion} increases accuracy from three strong source LLMs by a significant 8\%-17.8\%.

Reconstructing textureless areas in MVS poses challenges due to the absence of reliable pixel correspondences within fixed patch. Although certain methods employ patch deformation to expand the receptive field, their patches mistakenly skip depth edges to calculate areas with depth discontinuity, thereby causing ambiguity. Consequently, we introduce Multi-granularity Segmentation Prior Multi-View Stereo (MSP-MVS). Specifically, we first propose multi-granularity segmentation prior by integrating multi-granularity depth edges to restrict patch deformation within homogeneous areas. Moreover, we present anchor equidistribution that bring deformed patches with more uniformly distributed anchors to ensure an adequate coverage of their own homogeneous areas. Furthermore, we introduce iterative local search optimization to represent larger patch with sparse representative candidates, significantly boosting the expressive capacity for each patch. The state-of-the-art results on ETH3D and Tanks & Temples benchmarks demonstrate the effectiveness and robust generalization ability of our proposed method., Comment: After a thorough internal review, we identified a significant error in the experimental design described in the Multi-granularity Segmentation Prior Section of our paper, which impacts the accuracy of the data analysis and conclusions. We are in the process of correcting these errors and will submit an updated version in due course

Recently, Handwritten Mathematical Expression Recognition (HMER) has gained considerable attention in pattern recognition for its diverse applications in document understanding. Current methods typically approach HMER as an image-to-sequence generation task within an autoregressive (AR) encoder-decoder framework. However, these approaches suffer from several drawbacks: 1) a lack of overall language context, limiting information utilization beyond the current decoding step; 2) error accumulation during AR decoding; and 3) slow decoding speed. To tackle these problems, this paper makes the first attempt to build a novel bottom-up Non-AutoRegressive Modeling approach for HMER, called NAMER. NAMER comprises a Visual Aware Tokenizer (VAT) and a Parallel Graph Decoder (PGD). Initially, the VAT tokenizes visible symbols and local relations at a coarse level. Subsequently, the PGD refines all tokens and establishes connectivities in parallel, leveraging comprehensive visual and linguistic contexts. Experiments on CROHME 2014/2016/2019 and HME100K datasets demonstrate that NAMER not only outperforms the current state-of-the-art (SOTA) methods on ExpRate by 1.93%/2.35%/1.49%/0.62%, but also achieves significant speedups of 13.7x and 6.7x faster in decoding time and overall FPS, proving the effectiveness and efficiency of NAMER., Comment: Accepted by ECCV 2024

The existing large scale weak lensing surveys typically reserve the best seeing conditions for a certain optical band to minimize shape measurement errors and maximize the number of usable background galaxies. This is because most popular shear measurement methods contain explicit or implicit thresholds on the galaxy-to-PSF (point spread function) size ratio, below which their shape measurement errors increase abruptly. Using the DECaLS data, we have previously demonstrated that the Fourier\_Quad method performs very well on poorly resolved galaxy images in general. It is therefore a ready tool for shear measurement with multi-band images regardless of their seeing conditions. In this paper, we apply the Fourier\_Quad pipeline on the multi-band images from the third public data release of the Hyper Suprime-Cam Subaru Strategic Program. We show that the shear catalogs from the five optical bands (g/r/i/z/y) all pass the field-distortion test with very high accuracy. Using the LOWZ and CMASS galaxies as foreground lenses, we show that the errorbar in the galaxy-galaxy lensing measurement can be decreased by factors around 15\% by combining shear catalogs from different bands. This indicates that it is worthful to do multi-bands shear measurements for a better shear statistics., Comment: 14 pages, 12 figures

Tele-correlation refers to the correlation of galaxy shapes with large angular separations (e.g., $>100$ degrees). Since there are no astrophysical reasons causing such a correlation on cosmological scales, any detected tele-correlation could disclose systematic effects in shear-shear correlation measurement. If the shear estimators are measured on single exposures, we show that the field distortion (FD) signal associated with the galaxy position on the CCD can be retained and used in tele-correlation to help us directly calibrate the multiplicative and additive biases in shear-shear correlations. We use the DECaLS shear catalog produced by the Fourier\_Quad pipeline to demonstrate this idea. To our surprise, we find that significant multiplicative biases can arise (up to more than 10\%) due to redshift binning of the galaxies. Correction for this bias leads to about 1$\sigma$ increase of the best-fit value of $S_8$ from $0.760^{+0.015}_{-0.017}$ to $0.777^{+0.016}_{-0.019}$ in our tomography study.

Scheduling real-time tasks that utilize GPUs with analyzable guarantees poses a significant challenge due to the intricate interaction between CPU and GPU resources, as well as the complex GPU hardware and software stack. While much research has been conducted in the real-time research community, several limitations persist, including the absence or limited availability of GPU-level preemption, extended blocking times, and/or the need for extensive modifications to program code. In this paper, we propose GCAPS, a GPU Context-Aware Preemptive Scheduling approach for real-time GPU tasks. Our approach exerts control over GPU context scheduling at the device driver level and enables preemption of GPU execution based on task priorities by simply adding one-line macros to GPU segment boundaries. In addition, we provide a comprehensive response time analysis of GPU-using tasks for both our proposed approach as well as the default Nvidia GPU driver scheduling that follows a work-conserving round-robin policy. Through empirical evaluations and case studies, we demonstrate the effectiveness of the proposed approaches in improving taskset schedulability and response time. The results highlight significant improvements over prior work as well as the default scheduling approach, with up to 40% higher schedulability, while also achieving predictable worst-case behavior on Nvidia Jetson embedded platforms., Comment: Accepted by ECRTS 2024. arXiv admin note: substantial text overlap with arXiv:2401.16529

Silver-based chalcohalide anti-perovskites (CAP), Ag$_{3}$BC (B = S, Se; C = Cl, Br, I), represent an emerging family of energy materials with intriguing optoelectronic, vibrational and ionic transport properties. However, the structural features and phase stability of CAP remain poorly investigated to date, hindering their fundamental understanding and potential integration into technological applications. Here we employ theoretical first-principles methods based on density functional theory to fill this knowledge gap. Through crystal structure prediction techniques, ab initio molecular dynamics simulations, and quasi-harmonic free energy calculations, we unveil a series of previously overlooked energetically competitive phases and temperature-induced phase transitions for all CAP. Specifically, we identify a new cubic $P2_{1}3$ structure as the stable phase of all CAP containing S both at zero temperature and $T \neq 0$ K conditions. Consequently, our calculations suggest that the cubic $Pm\overline{3}m$ phase identified in room-temperature X-ray diffraction experiments is likely to be metastable. Furthermore, for CAP containing Se, we propose different orthorhombic ($Pca2_{1}$ and $P2_{1}2_{1}2_{1}$) and cubic ($I2_{1}3$) structures as the ground-state phases and reveal several phase transformations induced by temperature. This theoretical investigation not only identifies new candidate ground-state phases and solid-solid phase transformations for all CAP but also provides insights into potential stability issues affecting these highly anharmonic superionic materials., Comment: 14 pages, 12 figures

Most current deep learning models equivariant to $O(n)$ or $SO(n)$ either consider mostly scalar information such as distances and angles or have a very high computational complexity. In this work, we test a few novel message passing graph neural networks (GNNs) based on Clifford multivectors, structured similarly to other prevalent equivariant models in geometric deep learning. Our approach leverages efficient invariant scalar features while simultaneously performing expressive learning on multivector representations, particularly through the use of the equivariant geometric product operator. By integrating these elements, our methods outperform established efficient baseline models on an N-Body simulation task and protein denoising task while maintaining a high efficiency. In particular, we push the state-of-the-art error on the N-body dataset to 0.0035 (averaged over 3 runs); an 8% improvement over recent methods. Our implementation is available on Github.

As the current initialization method in the state-of-the-art Stereo Visual-Inertial SLAM framework, ORB-SLAM3 has limitations. Its success depends on the performance of the pure stereo SLAM system and is based on the underlying assumption that pure visual SLAM can accurately estimate the camera trajectory, which is essential for inertial parameter estimation. Meanwhile, the further improved initialization method for ORB-SLAM3, known as Stereo-NEC, is time-consuming due to applying keypoint tracking to estimate gyroscope bias with normal epipolar constraints. To address the limitations of previous methods, this paper proposes a method aimed at enhancing translation accuracy during the initialization stage. The fundamental concept of our method is to improve the translation estimate with a 3 Degree-of-Freedom (DoF) Bundle Adjustment (BA), independently, while the rotation estimate is fixed, instead of using ORB-SLAM3's 6-DoF BA. Additionally, the rotation estimate will be updated by considering IMU measurements and gyroscope bias, unlike ORB-SLAM3's rotation, which is directly obtained from stereo visual odometry and may yield inferior results when operating in challenging scenarios. We also conduct extensive evaluations on the public benchmark, the EuRoC dataset, demonstrating that our method excels in accuracy.

Reconstruction of the point spread function (PSF) plays an important role in many areas of astronomy, including photometry, astrometry, galaxy morphology, and shear measurement. The atmospheric and instrumental effects are the two main contributors to the PSF, both of which may exhibit complex spatial features. Current PSF reconstruction schemes typically rely on individual exposures, and its ability of reproducing the complicated features of the PSF distribution is therefore limited by the number of stars. Interestingly, in conventional methods, after stacking the model residuals of the PSF ellipticities and (relative) sizes from a large number of exposures, one can often observe some stable and nontrivial spatial patterns on the entire focal plane, which could be quite detrimental to, e.g., weak lensing measurements. These PSF residual patterns are caused by instrumental effects as they consistently appear in different exposures. Taking this as an advantage, we propose a multi-layer PSF reconstruction method to remove such PSF residuals, the second and third layers of which make use of all available exposures together. We test our method on the i-band data of the second release of Hyper Suprime-Cam. Our method successfully eliminates most of the PSF residuals. Using the Fourier\_Quad shear measurement method, we further test the performance of the resulting PSF fields on shear recovery using the field distortion effect. The PSF residuals have strong correlations with the shear residuals, and our new multi-layer PSF reconstruction method can remove most of such systematic errors related to PSF, leading to much smaller shear biases.

Multi-camera SLAM systems offer a plethora of advantages, primarily stemming from their capacity to amalgamate information from a broader field of view, thereby resulting in heightened robustness and improved localization accuracy. In this research, we present a significant extension and refinement of the state-of-the-art stereo SLAM system, known as ORB-SLAM2, with the objective of attaining even higher precision. To accomplish this objective, we commence by mapping measurements from all cameras onto a virtual camera termed BundledFrame. This virtual camera is meticulously engineered to seamlessly adapt to multi-camera configurations, facilitating the effective fusion of data captured from multiple cameras. Additionally, we harness extrinsic parameters in the bundle adjustment (BA) process to achieve precise trajectory estimation.Furthermore, we conduct an extensive analysis of the role of bundle adjustment (BA) in the context of multi-camera scenarios, delving into its impact on tracking, local mapping, and global optimization. Our experimental evaluation entails comprehensive comparisons between ground truth data and the state-of-the-art SLAM system. To rigorously assess the system's performance, we utilize the EuRoC datasets. The consistent results of our evaluations demonstrate the superior accuracy of our system in comparison to existing approaches.

Phenotype-driven gene prioritization is a critical process in the diagnosis of rare genetic disorders for identifying and ranking potential disease-causing genes based on observed physical traits or phenotypes. While traditional approaches rely on curated knowledge graphs with phenotype-gene relations, recent advancements in large language models have opened doors to the potential of AI predictions through extensive training on diverse corpora and complex models. This study conducted a comprehensive evaluation of five large language models, including two Generative Pre-trained Transformers series, and three Llama2 series, assessing their performance across three key metrics: task completeness, gene prediction accuracy, and adherence to required output structures. Various experiments explored combinations of models, prompts, input types, and task difficulty levels. Our findings reveal that even the best-performing LLM, GPT-4, achieved an accuracy of 16.0%, which still lags behind traditional bioinformatics tools. Prediction accuracy increased with the parameter/model size. A similar increasing trend was observed for the task completion rate, with complicated prompts more likely to increase task completeness in models smaller than GPT-4. However, complicated prompts are more likely to decrease the structure compliance rate, but no prompt effects on GPT-4. Compared to HPO term-based input, LLM was also able to achieve better than random prediction accuracy by taking free-text input, but slightly lower than with the HPO input. Bias analysis showed that certain genes, such as MECP2, CDKL5, and SCN1A, are more likely to be top-ranked, potentially explaining the variances observed across different datasets. This study provides valuable insights into the integration of LLMs within genomic analysis, contributing to the ongoing discussion on the utilization of advanced LLMs in clinical workflows., Comment: 56 pages, 6 figures, 6 tables, 2 supplementary tables

Despite the promising future of autonomous robots, several key issues currently remain that can lead to compromised performance and safety. One such issue is latency, where we find that even the latest embedded platforms from NVIDIA fail to execute intelligence tasks (e.g., object detection) of autonomous vehicles in a real-time fashion. One remedy to this problem is the promising paradigm of edge computing. Through collaboration with our industry partner, we identify key prohibitive limitations of the current edge mindset: (1) servers are not distributed enough and thus, are not close enough to vehicles, (2) current proposed edge solutions do not provide substantially better performance and extra information specific to autonomous vehicles to warrant their cost to the user, and (3) the state-of-the-art solutions are not compatible with popular frameworks used in autonomous systems, particularly the Robot Operating System (ROS). To remedy these issues, we provide Genie, an encapsulation technique that can enable transparent caching in ROS in a non-intrusive way (i.e., without modifying the source code), can build the cache in a distributed manner (in contrast to traditional central caching methods), and can construct a collective three-dimensional object map to provide substantially better latency (even on low-power edge servers) and higher quality data to all vehicles in a certain locality. We fully implement our design on state-of-the-art industry-adopted embedded and edge platforms, using the prominent autonomous driving software Autoware, and find that Genie can enhance the latency of Autoware Vision Detector by 82% on average, enable object reusability 31% of the time on average and as much as 67% for the incoming requests, and boost the confidence in its object map considerably over time., Comment: Submitted to ICRA 2025

While Large Language Models (LLMs) demonstrate impressive capabilities in text generation, we find that their ability has yet to be generalized to music, humanity's creative language. We introduce ChatMusician, an open-source LLM that integrates intrinsic musical abilities. It is based on continual pre-training and finetuning LLaMA2 on a text-compatible music representation, ABC notation, and the music is treated as a second language. ChatMusician can understand and generate music with a pure text tokenizer without any external multi-modal neural structures or tokenizers. Interestingly, endowing musical abilities does not harm language abilities, even achieving a slightly higher MMLU score. Our model is capable of composing well-structured, full-length music, conditioned on texts, chords, melodies, motifs, musical forms, etc, surpassing GPT-4 baseline. On our meticulously curated college-level music understanding benchmark, MusicTheoryBench, ChatMusician surpasses LLaMA2 and GPT-3.5 on zero-shot setting by a noticeable margin. Our work reveals that LLMs can be an excellent compressor for music, but there remains significant territory to be conquered. We release our 4B token music-language corpora MusicPile, the collected MusicTheoryBench, code, model and demo in GitHub., Comment: GitHub: https://shanghaicannon.github.io/ChatMusician/

We introduce Clifford Group Equivariant Simplicial Message Passing Networks, a method for steerable E(n)-equivariant message passing on simplicial complexes. Our method integrates the expressivity of Clifford group-equivariant layers with simplicial message passing, which is topologically more intricate than regular graph message passing. Clifford algebras include higher-order objects such as bivectors and trivectors, which express geometric features (e.g., areas, volumes) derived from vectors. Using this knowledge, we represent simplex features through geometric products of their vertices. To achieve efficient simplicial message passing, we share the parameters of the message network across different dimensions. Additionally, we restrict the final message to an aggregation of the incoming messages from different dimensions, leading to what we term shared simplicial message passing. Experimental results show that our method is able to outperform both equivariant and simplicial graph neural networks on a variety of geometric tasks.

The Segment Anything Model (SAM), a profound vision foundation model pre-trained on a large-scale dataset, breaks the boundaries of general segmentation and sparks various downstream applications. This paper introduces Hi-SAM, a unified model leveraging SAM for hierarchical text segmentation. Hi-SAM excels in text segmentation across four hierarchies, including stroke, word, text-line, and paragraph, while realizing layout analysis as well. Specifically, we first turn SAM into a high-quality text stroke segmentation (TSS) model through a parameter-efficient fine-tuning approach. We use this TSS model to iteratively generate the text stroke labels in a semi-automatical manner, unifying labels across the four text hierarchies in the HierText dataset. Subsequently, with these complete labels, we launch the end-to-end trainable Hi-SAM based on the TSS architecture with a customized hierarchical mask decoder. During inference, Hi-SAM offers both automatic mask generation (AMG) mode and promptable segmentation mode. In terms of the AMG mode, Hi-SAM segments text stroke foreground masks initially, then samples foreground points for hierarchical text mask generation and achieves layout analysis in passing. As for the promptable mode, Hi-SAM provides word, text-line, and paragraph masks with a single point click. Experimental results show the state-of-the-art performance of our TSS model: 84.86% fgIOU on Total-Text and 88.96% fgIOU on TextSeg for text stroke segmentation. Moreover, compared to the previous specialist for joint hierarchical detection and layout analysis on HierText, Hi-SAM achieves significant improvements: 4.73% PQ and 5.39% F1 on the text-line level, 5.49% PQ and 7.39% F1 on the paragraph level layout analysis, requiring 20x fewer training epochs. The code is available at https://github.com/ymy-k/Hi-SAM., Comment: GitHub repository: https://github.com/ymy-k/Hi-SAM

Scheduling real-time tasks that utilize GPUs with analyzable guarantees poses a significant challenge due to the intricate interaction between CPU and GPU resources, as well as the complex GPU hardware and software stack. While much research has been conducted in the real-time research community, several limitations persist, including the absence or limited availability of preemption, extended blocking times, and/or the need for extensive modifications to program code. In this paper, we propose two novel techniques, namely the kernel thread and IOCTL-based approaches, to enable preemptive priority-based scheduling for real-time GPU tasks. Our approaches exert control over GPU context scheduling at the device driver level and enable preemptive GPU scheduling based on task priorities. The kernel thread-based approach achieves this without requiring modifications to user-level programs, while the IOCTL-based approach needs only a single macro at the boundaries of GPU access segments. In addition, we provide a comprehensive response time analysis that takes into account overlaps between different task segments, mitigating pessimism in worst-case estimates. Through empirical evaluations and case studies, we demonstrate the effectiveness of the proposed approaches in improving taskset schedulability and timeliness of real-time tasks. The results highlight significant improvements over prior work, with up to 40\% higher schedulability, while also achieving predictable worst-case behavior on Nvidia Jetson embedded platforms.

In recent times, a plethora of Large Code Generation Models (LCGMs) have been proposed, showcasing significant potential in assisting developers with complex programming tasks. Benchmarking LCGMs necessitates the creation of a set of diverse programming problems, and each problem comprises the prompt (including the task description), canonical solution, and test inputs. The existing methods for constructing such a problem set can be categorized into two main types: manual methods and perturbation-based methods. However, manual methods demand high effort and lack scalability, while also risking data integrity due to LCGMs' potentially contaminated data collection, and perturbation-based approaches mainly generate semantically homogeneous problems with the same canonical solutions and introduce typos that can be easily auto-corrected by IDE, making them ineffective and unrealistic. In this work, we propose the idea of programming problem merging (PPM) and provide two implementation of this idea, we utilize our tool on two widely-used datasets and compare it against nine baseline methods using eight code generation models. The results demonstrate the effectiveness of our tool in generating more challenging, diverse, and natural programming problems, comparing to the baselines., Comment: This paper has been accepted to The ACM International Conference on the Foundations of Software Engineering FSE 2024

The language control network is vital among language-related networks responsible for solving the problem of multiple language switching. Researchers have expressed concerns about the instability of the language control network when exposed to external influences (e.g., Long-term second language learning). However, some studies have suggested that the language control network is stable. Therefore, whether the language control network is stable or not remains unclear. In the present study, we directly evaluated the stability and instability of the language control network using resting-state functional magnetic resonance imaging (rs-fMRI). We employed cohorts of Chinese first-year college students majoring in English who underwent second language (L2) acquisition courses at a university and those who did not. Two resting-state fMRI scans were acquired approximately 1 year apart. We found that the language control network was both moderately stable and unstable. We further investigated the morphological coexistence patterns of stability and instability within the language control network. First, we extracted connections representing stability and plasticity from the entire network. We then evaluated whether the coexistence patterns were modular (stability and instability involve different brain regions) or non-modular (stability and plasticity involve the same brain regions but have unique connectivity patterns). We found that both stability and instability coexisted in a non-modular pattern. Compared with the non-English major group, the English major group has a more non-modular coexistence pattern.. These findings provide preliminary evidence of the coexistence of stability and instability in the language control network.

Large Language Models (LLMs) have been widely employed in programming language analysis to enhance human productivity. Yet, their reliability can be compromised by various code distribution shifts, leading to inconsistent outputs. While probabilistic methods are known to mitigate such impact through uncertainty calibration and estimation, their efficacy in the language domain remains underexplored compared to their application in image-based tasks. In this work, we first introduce a large-scale benchmark dataset, incorporating three realistic patterns of code distribution shifts at varying intensities. Then we thoroughly investigate state-of-the-art probabilistic methods applied to CodeLlama using these shifted code snippets. We observe that these methods generally improve the uncertainty awareness of CodeLlama, with increased calibration quality and higher uncertainty estimation~(UE) precision. However, our study further reveals varied performance dynamics across different criteria (e.g., calibration error vs misclassification detection) and trade-off between efficacy and efficiency, highlighting necessary methodological selection tailored to specific contexts., Comment: 16 pages, 12 figures

Individuals with suspected rare genetic disorders often undergo multiple clinical evaluations, imaging studies, laboratory tests and genetic tests, to find a possible answer over a prolonged period of time. Addressing this "diagnostic odyssey" thus has substantial clinical, psychosocial, and economic benefits. Many rare genetic diseases have distinctive facial features, which can be used by artificial intelligence algorithms to facilitate clinical diagnosis, in prioritizing candidate diseases to be further examined by lab tests or genetic assays, or in helping the phenotype-driven reinterpretation of genome/exome sequencing data. Existing methods using frontal facial photos were built on conventional Convolutional Neural Networks (CNNs), rely exclusively on facial images, and cannot capture non-facial phenotypic traits and demographic information essential for guiding accurate diagnoses. Here we introduce GestaltMML, a multimodal machine learning (MML) approach solely based on the Transformer architecture. It integrates facial images, demographic information (age, sex, ethnicity), and clinical notes (optionally, a list of Human Phenotype Ontology terms) to improve prediction accuracy. Furthermore, we also evaluated GestaltMML on a diverse range of datasets, including 528 diseases from the GestaltMatcher Database, several in-house datasets of Beckwith-Wiedemann syndrome (BWS, over-growth syndrome with distinct facial features), Sotos syndrome (overgrowth syndrome with overlapping features with BWS), NAA10-related neurodevelopmental syndrome, Cornelia de Lange syndrome (multiple malformation syndrome), and KBG syndrome (multiple malformation syndrome). Our results suggest that GestaltMML effectively incorporates multiple modalities of data, greatly narrowing candidate genetic diagnoses of rare diseases and may facilitate the reinterpretation of genome/exome sequencing data., Comment: Significant revisions

Recent developments in balancing the usefulness and safety of Large Language Models (LLMs) have raised a critical question: Are mainstream NLP tasks adequately aligned with safety consideration? Our study, focusing on safety-sensitive documents obtained through adversarial attacks, reveals significant disparities in the safety alignment of various NLP tasks. For instance, LLMs can effectively summarize malicious long documents but often refuse to translate them. This discrepancy highlights a previously unidentified vulnerability: attacks exploiting tasks with weaker safety alignment, like summarization, can potentially compromise the integrity of tasks traditionally deemed more robust, such as translation and question-answering (QA). Moreover, the concurrent use of multiple NLP tasks with lesser safety alignment increases the risk of LLMs inadvertently processing harmful content. We demonstrate these vulnerabilities in various safety-aligned LLMs, particularly Llama2 models, Gemini and GPT-4, indicating an urgent need for strengthening safety alignments across a broad spectrum of NLP tasks., Comment: Accepted to ACL2024 main

Since the release of ChatGPT, generative models have achieved tremendous success and become the de facto approach for various NLP tasks. However, its application in the field of input methods remains under-explored. Many neural network approaches have been applied to the construction of Chinese input method engines(IMEs).Previous research often assumed that the input pinyin was correct and focused on Pinyin-to-character(P2C) task, which significantly falls short of meeting users' demands. Moreover, previous research could not leverage user feedback to optimize the model and provide personalized results. In this study, we propose a novel Generative Input paradigm named GeneInput. It uses prompts to handle all input scenarios and other intelligent auxiliary input functions, optimizing the model with user feedback to deliver personalized results. The results demonstrate that we have achieved state-of-the-art performance for the first time in the Full-mode Key-sequence to Characters(FK2C) task. We propose a novel reward model training method that eliminates the need for additional manual annotations and the performance surpasses GPT-4 in tasks involving intelligent association and conversational assistance. Compared to traditional paradigms, GeneInput not only demonstrates superior performance but also exhibits enhanced robustness, scalability, and online learning capabilities.

Recently, heatmap regression methods based on 1D landmark representations have shown prominent performance on locating facial landmarks. However, previous methods ignored to make deep explorations on the good potentials of 1D landmark representations for sequential and structural modeling of multiple landmarks to track facial landmarks. To address this limitation, we propose a Transformer architecture, namely 1DFormer, which learns informative 1D landmark representations by capturing the dynamic and the geometric patterns of landmarks via token communications in both temporal and spatial dimensions for facial landmark tracking. For temporal modeling, we propose a recurrent token mixing mechanism, an axis-landmark-positional embedding mechanism, as well as a confidence-enhanced multi-head attention mechanism to adaptively and robustly embed long-term landmark dynamics into their 1D representations; for structure modeling, we design intra-group and inter-group structure modeling mechanisms to encode the component-level as well as global-level facial structure patterns as a refinement for the 1D representations of landmarks through token communications in the spatial dimension via 1D convolutional layers. Experimental results on the 300VW and the TF databases show that 1DFormer successfully models the long-range sequential patterns as well as the inherent facial structures to learn informative 1D representations of landmark sequences, and achieves state-of-the-art performance on facial landmark tracking.

Cosmic shear statistics, such as the two-point correlation function (2PCF), can be evaluated with the PDF-SYM method instead of the traditional weighted-sum approach. It makes use of the full PDF information of the shear estimators, and does not require weightings on the shear estimators, which can in principle introduce additional systematic biases. This work presents our constraints on $S_8$ and $\Omega_m$ from the shear-shear correlations using the PDF-SYM method. The data we use is from the z-band images of the Dark Energy Camera Legacy Survey (DECaLS), which covers about 10000 deg$^2$ with more than 100 million galaxies. The shear catalog is produced by the Fourier\_Quad method, and well tested on the real data itself with the field-distortion effect. Our main approach is called quasi-2D as we do use the photo-$z$ information of each individual galaxy, but without dividing the galaxies into redshift bins. We mainly use galaxy pairs within the redshift interval between 0.2 and 1.3, and the angular range from $4.7$ to $180$ arcmin. Our analysis yields $S_8=0.762 \pm 0.026$ and $\Omega_{\rm m}=0.234 \pm 0.075$, with the baryon effects and the intrinsic alignments included. The results are robust against redshift uncertainties. We check the consistency of our results by deriving the cosmological constraints from auto-correlations of $\gamma_1$ and $\gamma_2$ separately, and find that they are consistent with each other, but the constraints from the $\gamma_1$ component is much weaker than that from $\gamma_2$. It implies a much worse data quality of $\gamma_1$, which is likely due to additional shear uncertainties caused by CCD electronics (according to the survey strategy of DECaLS). We also perform a pure 2D analysis, which gives $S_8=0.81^{+0.03}_{-0.04}$ and $\Omega_{\rm m}=0.25^{+0.06}_{-0.05}$. Our findings demonstrate the potential of the PDF-SYM method for precision cosmology., Comment: 10 figures. Published on SCIENCE CHINA Physics, Mechanics & Astronomy

The early afterglow of a Gamma-ray burst (GRB) can provide critical information on the jet and progenitor of the GRB. The extreme brightness of GRB 221009A allows us to probe its early afterglow in unprecedented detail. In this letter, we report comprehensive observation results of the early afterglow of GRB 221009A (from $T_0$+660 s to $T_0$+1860 s, where $T_0$ is the \textit{Insight}-HXMT/HE trigger time) in X/$\gamma$-ray energy band (from 20 keV to 20 MeV) by \textit{Insight}-HXMT/HE, GECAM-C and \textit{Fermi}/GBM. We find that the spectrum of the early afterglow in 20 keV-20 MeV could be well described by a cutoff power-law with an extra power-law which dominates the low and high energy bands respectively. The cutoff power-law $E_{\rm peak}$ is $\sim$ 30 keV and the power-law photon index is $\sim$ 1.8 throughout the early afterglow phase. By fitting the light curves in different energy bands, we find that a significant achromatic break (from keV to TeV) is required at $T_0$ + 1246$^{+27}_{-26}$ s (i.e. 1021 s since the afterglow starting time $T_{\rm AG}$=$T_0$+225 s), providing compelling evidence of a jet break. Interestingly, both the pre-break and post-break decay slopes vary with energy, and these two slopes become closer in the lower energy band, making the break less identifiable. Intriguingly, the spectrum of the early afterglow experienced a slight hardening before the break and a softening after the break. These results provide new insights into the understanding of this remarkable GRB., Comment: Accepted for publication in ApJ Letters on 19-Jan-2024, 11 pages, 7 figures and 2 tables

Recent studies have demonstrated that large language models (LLMs) store massive factual knowledge within their parameters. But existing LLMs are prone to hallucinate unintended text due to false or outdated knowledge. Since retraining LLMs is resource intensive, there has been a growing interest in the concept of model editing. Despite the emergence of benchmarks and approaches, these unidirectional editing and evaluation have failed to explore the reversal curse. Intuitively, if "The capital of France is" is edited to be a counterfact "London" within a model, then it should be able to naturally reason and recall the reverse fact, i.e., "London is the capital of" followed by "France" instead of "England". In this paper, we study bidirectional language model editing, aiming to provide rigorous model editing evaluation to assess if edited LLMs can recall the editing knowledge bidirectionally. A new evaluation metric of reversibility is introduced, and a benchmark dubbed as Bidirectional Assessment for Knowledge Editing (BAKE) is constructed to evaluate the reversibility of edited models in recalling knowledge in the reverse direction of editing. We surprisingly observe that while current editing methods and LLMs can effectively recall editing facts in the direction of editing, they suffer serious deficiencies when evaluated in the reverse direction. To mitigate the reversal curse, a method named Bidirectionally Inversible Relationship moDeling (BIRD) is proposed. A set of editing objectives that incorporate bidirectional relationships between subject and object into the updated model weights are designed. Experiments show that BIRD improves the performance of four representative LLMs of different sizes via question answering and judgement.

The prosperity of deep neural networks (DNNs) is largely benefited from open-source datasets, based on which users can evaluate and improve their methods. In this paper, we revisit backdoor-based dataset ownership verification (DOV), which is currently the only feasible approach to protect the copyright of open-source datasets. We reveal that these methods are fundamentally harmful given that they could introduce malicious misclassification behaviors to watermarked DNNs by the adversaries. In this paper, we design DOV from another perspective by making watermarked models (trained on the protected dataset) correctly classify some `hard' samples that will be misclassified by the benign model. Our method is inspired by the generalization property of DNNs, where we find a \emph{hardly-generalized domain} for the original dataset (as its \emph{domain watermark}). It can be easily learned with the protected dataset containing modified samples. Specifically, we formulate the domain generation as a bi-level optimization and propose to optimize a set of visually-indistinguishable clean-label modified data with similar effects to domain-watermarked samples from the hardly-generalized domain to ensure watermark stealthiness. We also design a hypothesis-test-guided ownership verification via our domain watermark and provide the theoretical analyses of our method. Extensive experiments on three benchmark datasets are conducted, which verify the effectiveness of our method and its resistance to potential adaptive methods. The code for reproducing main experiments is available at \url{https://github.com/JunfengGo/Domain-Watermark}., Comment: This paper is accepted by NeurIPS 2023. The first two authors contributed equally to this work. 30 pages