Search

In the Architecture Engineering & Construction (AEC) industry, how design behaviors impact design quality remains unclear. This study proposes a novel approach, which, for the first time, identifies and quantitatively describes the relationship between design behaviors and quality of design based on Building Information Modeling (BIM). Real-time collection and log mining are integrated to collect raw data of design behaviors. Feature engineering and various machine learning models are then utilized for quantitative modeling and interpretation. Results confirm an existing quantifiable relationship which can be learned by various models. The best-performing model using Extremely Random Trees achieved an R2 value of 0.88 on the test set. Behavioral features related to designer's skill level and changes of design intentions are identified to have significant impacts on design quality. These findings deepen our understanding of the design process and help forming BIM designs with better quality.

DarkSHINE is a newly proposed fixed-target experiment initiative to search for the invisible decay of Dark Photon via missing energy/momentum signatures, based on the high repetition rate electron beam to be deployed/delivered by the Shanghai High repetition rate XFEL and Extreme light facility (SHINE). This report elaborates the baseline design of DarkSHINE experiment by introducing the physics goals, experimental setups, details of each sub-detector system technical designs, signal and backgground modelings, expected search sensitivities and future prospects, which mark an important step towards the further prototyping and technical demonstrations.

Depth completion using lightweight time-of-flight (ToF) depth sensors is attractive due to their low cost. However, lightweight ToF sensors usually have a limited field of view (FOV) compared with cameras. Thus, only pixels in the zone area of the image can be associated with depth signals. Previous methods fail to propagate depth features from the zone area to the outside-zone area effectively, thus suffering from degraded depth completion performance outside the zone. To this end, this paper proposes the CFPNet to achieve cross-zone feature propagation from the zone area to the outside-zone area with two novel modules. The first is a direct-attention-based propagation module (DAPM), which enforces direct cross-zone feature acquisition. The second is a large-kernel-based propagation module (LKPM), which realizes cross-zone feature propagation by utilizing convolution layers with kernel sizes up to 31. CFPNet achieves state-of-the-art (SOTA) depth completion performance by combining these two modules properly, as verified by extensive experimental results on the ZJU-L5 dataset. The code will be made public.

Jailbreak attack can be used to access the vulnerabilities of Large Language Models (LLMs) by inducing LLMs to generate the harmful content. And the most common method of the attack is to construct semantically ambiguous prompts to confuse and mislead the LLMs. To access the security and reveal the intrinsic relation between the input prompt and the output for LLMs, the distribution of attention weight is introduced to analyze the underlying reasons. By using statistical analysis methods, some novel metrics are defined to better describe the distribution of attention weight, such as the Attention Intensity on Sensitive Words (Attn_SensWords), the Attention-based Contextual Dependency Score (Attn_DepScore) and Attention Dispersion Entropy (Attn_Entropy). By leveraging the distinct characteristics of these metrics, the beam search algorithm and inspired by the military strategy "Feint and Attack", an effective jailbreak attack strategy named as Attention-Based Attack (ABA) is proposed. In the ABA, nested attack prompts are employed to divert the attention distribution of the LLMs. In this manner, more harmless parts of the input can be used to attract the attention of the LLMs. In addition, motivated by ABA, an effective defense strategy called as Attention-Based Defense (ABD) is also put forward. Compared with ABA, the ABD can be used to enhance the robustness of LLMs by calibrating the attention distribution of the input prompt. Some comparative experiments have been given to demonstrate the effectiveness of ABA and ABD. Therefore, both ABA and ABD can be used to access the security of the LLMs. The comparative experiment results also give a logical explanation that the distribution of attention weight can bring great influence on the output for LLMs.

In 1927, during the fifth Solvay Conference, Einstein and Bohr described a double-slit interferometer with a "movable slit" that can detect the momentum recoil of one photon. Here, we report a faithful realization of the Einstein-Bohr interferometer using a single atom in an optical tweezer, cooled to the motional ground state in three dimensions. The single atom has an intrinsic momentum uncertainty comparable to a single photon, which serves as a movable slit obeying the minimum Heisenberg uncertainty principle. The atom's momentum wavefunction is dynamically tunable by the tweezer laser power, which enables observation of an interferometric visibility reduction at a shallower trap, demonstrating the quantum nature of this interferometer. We further identify classical noise due to atom heating and precession, illustrating a quantum-to-classical transition., Comment: 18 pages, 4 figures

Federated instruction tuning enables multiple clients to collaboratively fine-tune a shared large language model (LLM) that can follow humans' instructions without directly sharing raw data. However, existing literature impractically requires that all the clients readily hold instruction-tuning data (i.e., structured instruction-response pairs), which necessitates massive human annotations since clients' data is usually unstructured text instead. Addressing this, we propose a novel and flexible framework FedIT-U2S, which can automatically transform unstructured corpus into structured data for federated instruction tuning. FedIT-U2S consists two key steps: (1) few-shot instruction-tuning data generation, where each unstructured data piece together with several examples is combined to prompt an LLM in generating an instruction-response pair. To further enhance the flexibility, a retrieval-based example selection technique is proposed, where the examples are automatically selected based on the relatedness between the client's data piece and example pool, bypassing the need of determining examples in advance. (2) A typical federated instruction tuning process based on the generated data. Overall, FedIT-U2S can be applied to diverse scenarios as long as the client holds valuable text corpus, broadening the application scope of federated instruction tuning. We conduct a series of experiments on three domains (medicine, knowledge, and math), showing that our proposed FedIT-U2S can consistently and significantly brings improvement over the base LLM., Comment: 11 pages, work in progress

With significant advances in classifying and cataloguing topological matter, the focus of topological physics has shifted towards quantum control, particularly the creation and manipulation of topological phases of matter. Floquet engineering, the concept of tailoring a system by periodic fields, offers a powerful tool to manipulate electronic properties of condensed systems, and even to create exotic non-equilibrium topological states that are impossibly present in equilibrium scenarios. In this perspective, we give a brief review of recent progress in theoretical investigations of Floquet engineering topological states from effective models towards realistic materials. We show that light irradiation can realize various desired topological states through the introduction of symmetry breaking, such as first- and higher-order Weyl fermions, quadrupole topological insulator with periodic driving and disorder, quantum anomalous Hall effects with a tunable Chern number, as well as beyond. Moreover, based on first-principles calculations and Floquet theorem, we show several realistic material candidates proposed as potential hosts for promising Floquet topological states, facilitating their verification in experiments. We believe that our perspective on Floquet engineering of topological states will advance further studies of rich exotic light-induced phenomena in condensed matter physics., Comment: 40 pages, 11 figures

Foundation models that bridge vision and language have made significant progress, inspiring numerous life-enriching applications. However, their potential for misuse to introduce new threats remains largely unexplored. This paper reveals that vision-language models (VLMs) can be exploited to overcome longstanding limitations in gradient inversion attacks (GIAs) within federated learning (FL), where an FL server reconstructs private data samples from gradients shared by victim clients. Current GIAs face challenges in reconstructing high-resolution images, especially when the victim has a large local data batch. While focusing reconstruction on valuable samples rather than the entire batch is promising, existing methods lack the flexibility to allow attackers to specify their target data. In this paper, we introduce Geminio, the first approach to transform GIAs into semantically meaningful, targeted attacks. Geminio enables a brand new privacy attack experience: attackers can describe, in natural language, the types of data they consider valuable, and Geminio will prioritize reconstruction to focus on those high-value samples. This is achieved by leveraging a pretrained VLM to guide the optimization of a malicious global model that, when shared with and optimized by a victim, retains only gradients of samples that match the attacker-specified query. Extensive experiments demonstrate Geminio's effectiveness in pinpointing and reconstructing targeted samples, with high success rates across complex datasets under FL and large batch sizes and showing resilience against existing defenses.

The control of legged robots, particularly humanoid and quadruped robots, presents significant challenges due to their high-dimensional and nonlinear dynamics. While linear systems can be effectively controlled using methods like Model Predictive Control (MPC), the control of nonlinear systems remains complex. One promising solution is the Koopman Operator, which approximates nonlinear dynamics with a linear model, enabling the use of proven linear control techniques. However, achieving accurate linearization through data-driven methods is difficult due to issues like approximation error, domain shifts, and the limitations of fixed linear state-space representations. These challenges restrict the scalability of Koopman-based approaches. This paper addresses these challenges by proposing a continual learning algorithm designed to iteratively refine Koopman dynamics for high-dimensional legged robots. The key idea is to progressively expand the dataset and latent space dimension, enabling the learned Koopman dynamics to converge towards accurate approximations of the true system dynamics. Theoretical analysis shows that the linear approximation error of our method converges monotonically. Experimental results demonstrate that our method achieves high control performance on robots like Unitree G1/H1/A1/Go2 and ANYmal D, across various terrains using simple linear MPC controllers. This work is the first to successfully apply linearized Koopman dynamics for locomotion control of high-dimensional legged robots, enabling a scalable model-based control solution.

This paper presents an a priori error analysis of the Deep Mixed Residual method (MIM) for solving high-order elliptic equations with non-homogeneous boundary conditions, including Dirichlet, Neumann, and Robin conditions. We examine MIM with two types of loss functions, referred to as first-order and second-order least squares systems. By providing boundedness and coercivity analysis, we leverage C\'{e}a's Lemma to decompose the total error into the approximation, generalization, and optimization errors. Utilizing the Barron space theory and Rademacher complexity, an a priori error is derived regarding the training samples and network size that are exempt from the curse of dimensionality. Our results reveal that MIM significantly reduces the regularity requirements for activation functions compared to the deep Ritz method, implying the effectiveness of MIM in solving high-order equations., Comment: 40 pages, none figures

Multimodal Large Language Models (MLLMs) can enhance trustworthiness by aligning with human preferences. As human preference labeling is laborious, recent works employ evaluation models for assessing MLLMs' responses, using the model-based assessments to automate preference dataset construction. This approach, however, faces challenges with MLLMs' lengthy and compositional responses, which often require diverse reasoning skills that a single evaluation model may not fully possess. Additionally, most existing methods rely on closed-source models as evaluators. To address limitations, we propose DecompGen, a decomposable framework that uses an ensemble of open-sourced expert models. DecompGen breaks down each response into atomic verification tasks, assigning each task to an appropriate expert model to generate fine-grained assessments. The DecompGen feedback is used to automatically construct our preference dataset, DGPref. MLLMs aligned with DGPref via preference learning show improvements in trustworthiness, demonstrating the effectiveness of DecompGen.

We report the discovery of five galaxy candidates at redshifts between $15.9

Commercial video generation models have exhibited realistic, high-fidelity results but are still restricted to limited access. One crucial obstacle for large-scale applications is the expensive training and inference cost. In this paper, we argue that videos contain much more redundant information than images, thus can be encoded by very few motion latents based on a content image. Towards this goal, we design an image-conditioned VAE to encode a video to an extremely compressed motion latent space. This magic Reducio charm enables 64x reduction of latents compared to a common 2D VAE, without sacrificing the quality. Training diffusion models on such a compact representation easily allows for generating 1K resolution videos. We then adopt a two-stage video generation paradigm, which performs text-to-image and text-image-to-video sequentially. Extensive experiments show that our Reducio-DiT achieves strong performance in evaluation, though trained with limited GPU resources. More importantly, our method significantly boost the efficiency of video LDMs both in training and inference. We train Reducio-DiT in around 3.2K training hours in total and generate a 16-frame 1024*1024 video clip within 15.5 seconds on a single A100 GPU. Code released at https://github.com/microsoft/Reducio-VAE ., Comment: Code available at https://github.com/microsoft/Reducio-VAE

The data-driven method for infrared small target detection (IRSTD) has achieved promising results. However, due to the small scale of infrared small target datasets and the limited number of pixels occupied by the targets themselves, it is a challenging task for deep learning methods to directly learn from these samples. Utilizing human expert knowledge to assist deep learning methods in better learning is worthy of exploration. To effectively guide the model to focus on targets' spatial features, this paper proposes the Local Contrast Attention Enhanced infrared small target detection Network (LCAE-Net), combining prior knowledge with data-driven deep learning methods. LCAE-Net is a U-shaped neural network model which consists of two developed modules: a Local Contrast Enhancement (LCE) module and a Channel Attention Enhancement (CAE) module. The LCE module takes advantages of prior knowledge, leveraging handcrafted convolution operator to acquire Local Contrast Attention (LCA), which could realize background suppression while enhance the potential target region, thus guiding the neural network to pay more attention to potential infrared small targets' location information. To effectively utilize the response information throughout downsampling progresses, the CAE module is proposed to achieve the information fusion among feature maps' different channels. Experimental results indicate that our LCAE-Net outperforms existing state-of-the-art methods on the three public datasets NUDT-SIRST, NUAA-SIRST, and IRSTD-1K, and its detection speed could reach up to 70 fps. Meanwhile, our model has a parameter count and Floating-Point Operations (FLOPs) of 1.945M and 4.862G respectively, which is suitable for deployment on edge devices., Comment: 16 pages, 8 figures

We construct a minimal lattice model to provide an orbital description of lowest and first Landau levels. With the maximally localized Wannier functions with $s$, $p_-$, $p_+$ orbital characteristics, a three-orbital model is developed, where the lowest two Chern bands are flat with $\mathcal{C}=1$. This model can be viewed as consecutive band inversions between these Wannier states at $\Gamma$ and $K$ in momentum space, which adiabatically connects the atomic insulator limit to Landau level physics. Interestingly, many-body exact diagonalization and entanglement spectrum analysis suggest that the Abelian states can appear in the 1/3-filled lowest Chern band, while the signatures of the non-Abelian states are found in the half-filled first Chern band. This construction can be further extended to realize flat Chern bands resembling the higher Landau levels. Our results offer a new perspective to understand the lattice analogue of Landau levels, potentially enabling the realization of the fascinating topological phenomena at higher temperatures., Comment: 7 pages, 4 figures

Asymmetric relational data is increasingly prevalent across diverse fields, underscoring the need for directed network models to address the complex challenges posed by their unique structures. Unlike undirected models, directed models can capture reciprocity, the tendency of nodes to form mutual links. In this work, we address a fundamental question: what is the effective sample size for modeling reciprocity? We examine this by analyzing the Bernoulli model with reciprocity, allowing for varying sparsity levels between non-reciprocal and reciprocal effects. We then extend this framework to a model that incorporates node-specific heterogeneity and link-specific reciprocity using covariates. Our findings reveal intriguing interplays between non-reciprocal and reciprocal effects in sparse networks. We propose a straightforward inference procedure based on maximum likelihood estimation that operates without prior knowledge of sparsity levels, whether covariates are included or not.

We present the in-lab and on-sky performance for the upgraded 90 GHz focal plane of the Cosmology Large Angular Scale Surveyor (CLASS), which had four of its seven detector wafers updated during the austral winter of 2022. The update aimed to improve the transition-edge-sensor (TES) stability and bias range and to realize the high optical efficiency of the sensor design. Modifications included revised circuit terminations, electrical contact between the TES superconductor and the normal metal providing the bulk of the bolometer's heat capacity, and additional filtering on the TES bias lines. The upgrade was successful: 94% of detectors are stable down to 15% of the normal resistance, providing a wide overlapping range of bias voltages for all TESs on a wafer. The median telescope efficiency improved from $0.42^{+0.15}_{-0.22}$ to $0.60^{+0.10}_{-0.32}$ (68% quantiles). For the four upgraded wafers alone, median telescope efficiency increased to $0.65^{+0.06}_{-0.06}$. Given our efficiency estimate for the receiver optics, this telescope efficiency implies a detector efficiency exceeding $0.90$. The overall noise-equivalent temperature of the 90 GHz focal plane improved from 19 $\mu$K$\sqrt{s}$ to 11.3 $\mu$K$\sqrt{s}$., Comment: 18 pages, 10 figures.Submitted to ApJS

Clifford circuits can be utilized to disentangle quantum state with polynomial cost, thanks to the Gottesman-Knill theorem. Based on this idea, Clifford Circuits Augmented Matrix Product States (CAMPS) method, which is a seamless integration of Clifford circuits within the DMRG algorithm, was proposed recently and was shown to be able to reduce entanglement in various quantum systems. In this work, we further explore the power of CAMPS method in critical spin chains described by conformal field theories (CFTs) in the scaling limit. We find that the variationally optimized disentangler corresponds to {\it duality} transformations, which significantly reduce the entanglement entropy in the ground state. For critical quantum Ising spin chain governed by the Ising CFT with self-duality, the Clifford circuits found by CAMPS coincide with the duality transformation, e.g., the Kramer-Wannier self-duality in the critical Ising chain. It reduces the entanglement entropy by mapping the free conformal boundary condition to the fixed one. In the more general case of XXZ chain, the CAMPS gives rise to a duality transformation mapping the model to the quantum Ashkin-Teller spin chain. Our results highlight the potential of CAMPS as a versatile tool for uncovering hidden dualities and simplifying the entanglement structure of critical quantum systems., Comment: 6 pages, 4 figures

We introduce OrigamiPlot, an open-source R package and Shiny web application designed to enhance the visualization of multivariate data. This package implements the origami plot, a novel visualization technique proposed by Duan et al. in 2023, which improves upon traditional radar charts by ensuring that the area of the connected region is invariant to the ordering of attributes, addressing a key limitation of radar charts. The software facilitates multivariate decision-making by supporting comparisons across multiple objects and attributes, offering customizable features such as auxiliary axes and weighted attributes for enhanced clarity. Through the R package and user-friendly Shiny interface, researchers can efficiently create and customize plots without requiring extensive programming knowledge. Demonstrated using network meta-analysis as a real-world example, OrigamiPlot proves to be a versatile tool for visualizing multivariate data across various fields. This package opens new opportunities for simplifying decision-making processes with complex data.

Despite the success of text retrieval in many NLP tasks, code retrieval remains a largely underexplored area. Most text retrieval systems are tailored for natural language queries, often neglecting the specific challenges of retrieving code. This gap leaves existing models unable to effectively capture the diversity of programming languages and tasks across different domains, highlighting the need for more focused research in code retrieval. To address this, we introduce CodeXEmbed, a family of large-scale code embedding models ranging from 400M to 7B parameters. Our novel training pipeline unifies multiple programming languages and transforms various code-related tasks into a common retrieval framework, enhancing model generalizability and retrieval performance. Our 7B model sets a new state-of-the-art (SOTA) in code retrieval, outperforming the previous leading model, Voyage-Code, by over 20% on CoIR benchmark. In addition to excelling in code retrieval, our models demonstrate competitive performance on the widely adopted BeIR text retrieval benchmark, offering versatility across domains. Experimental results demonstrate that improving retrieval performance significantly enhances end-to-end Retrieval-Augmented Generation (RAG) performance for code-related tasks.

This article introduces the ManiSkill-ViTac Challenge 2025, which focuses on learning contact-rich manipulation skills using both tactile and visual sensing. Expanding upon the 2024 challenge, ManiSkill-ViTac 2025 includes 3 independent tracks: tactile manipulation, tactile-vision fusion manipulation, and tactile sensor structure design. The challenge aims to push the boundaries of robotic manipulation skills, emphasizing the integration of tactile and visual data to enhance performance in complex, real-world tasks. Participants will be evaluated using standardized metrics across both simulated and real-world environments, spurring innovations in sensor design and significantly advancing the field of vision-tactile fusion in robotics., Comment: Challenge webpage: https://ai-workshops.github.io/maniskill-vitac-challenge-2025/

In inertial confinement fusion (ICF), affected by non-steady ablation and various physical mechanisms, we extend the classical buoyancy-drag (BD) model into an ablative version for evaluating and controlling nonlinear ablative Rayleigh-Taylor instability (ARTI) in real space. The application of our ablative BD model in the nonlinear phase lies in a single adjustable coefficient influenced by initial perturbations, linear growth rate and terminal velocity. After validating the effectiveness and sensitivity of this model through simulations, we propose a strategy to shift the dominant mode away from the "most dangerous mode", which depends on initial perturbations. Our findings suggest that the "most dangerous mode" may clarify gain differences among targets of similar qualities and provide guidance for target manufacturing and pulse optimization in proximity to the ignition cliff.

Recently, Text-to-Image (T2I) synthesis technology has made tremendous strides. Numerous representative T2I models have emerged and achieved promising application outcomes, such as DALL-E, Stable Diffusion, Imagen, etc. In practice, it has become increasingly popular for model developers to selectively adopt various pre-trained text encoders and conditional diffusion models from third-party platforms, integrating them to build customized (personalized) T2I models. However, such an adoption approach is vulnerable to backdoor attacks. In this work, we propose a Combinational Backdoor Attack against Customized T2I models (CBACT2I) targeting this application scenario. Different from previous backdoor attacks against T2I models, CBACT2I embeds the backdoor into the text encoder and the conditional diffusion model separately. The customized T2I model exhibits backdoor behaviors only when the backdoor text encoder is used in combination with the backdoor conditional diffusion model. These properties make CBACT2I more stealthy and flexible than prior backdoor attacks against T2I models. Extensive experiments demonstrate the effectiveness of CBACT2I with different backdoor triggers and different backdoor targets on the open-sourced Stable Diffusion model. This work reveals the backdoor vulnerabilities of customized T2I models and urges countermeasures to mitigate backdoor threats in this scenario.

Graph clustering is an unsupervised machine learning method that partitions the nodes in a graph into different groups. Despite achieving significant progress in exploiting both attributed and structured data information, graph clustering methods often face practical challenges related to data isolation. Moreover, the absence of collaborative methods for graph clustering limits their effectiveness. In this paper, we propose a collaborative graph clustering framework for attributed graphs, supporting attributed graph clustering over vertically partitioned data with different participants holding distinct features of the same data. Our method leverages a novel technique that reduces the sample space, improving the efficiency of the attributed graph clustering method. Furthermore, we compare our method to its centralized counterpart under a proximity condition, demonstrating that the successful local results of each participant contribute to the overall success of the collaboration. We fully implement our approach and evaluate its utility and efficiency by conducting experiments on four public datasets. The results demonstrate that our method achieves comparable accuracy levels to centralized attributed graph clustering methods. Our collaborative graph clustering framework provides an efficient and effective solution for graph clustering challenges related to data isolation., Comment: 16 pages, 3 figures

Efficient acquisition of real-world embodied data has been increasingly critical. However, large-scale demonstrations captured by remote operation tend to take extremely high costs and fail to scale up the data size in an efficient manner. Sampling the episodes under a simulated environment is a promising way for large-scale collection while existing simulators fail to high-fidelity modeling on texture and physics. To address these limitations, we introduce the RoboGSim, a real2sim2real robotic simulator, powered by 3D Gaussian Splatting and the physics engine. RoboGSim mainly includes four parts: Gaussian Reconstructor, Digital Twins Builder, Scene Composer, and Interactive Engine. It can synthesize the simulated data with novel views, objects, trajectories, and scenes. RoboGSim also provides an online, reproducible, and safe evaluation for different manipulation policies. The real2sim and sim2real transfer experiments show a high consistency in the texture and physics. Moreover, the effectiveness of synthetic data is validated under the real-world manipulated tasks. We hope RoboGSim serves as a closed-loop simulator for fair comparison on policy learning. More information can be found on our project page https://robogsim.github.io/ .

Transitioning from quantum computation on physical qubits to quantum computation on encoded, logical qubits can improve the error rate of operations, and will be essential for realizing valuable quantum computational advantages. Using a neutral atom quantum processor with 256 qubits, each an individual Ytterbium atom, we demonstrate the entanglement of 24 logical qubits using the distance-two [[4,2,2]] code, simultaneously detecting errors and correcting for lost qubits. We also implement the Bernstein-Vazirani algorithm with up to 28 logical qubits encoded in the [[4,1,2]] code, showing better-than-physical error rates. We demonstrate fault-tolerant quantum computation in our approach, guided by the proposal of Gottesman (2016), by performing repeated loss correction for both structured and random circuits encoded in the [[4,2,2]] code. Finally, since distance-two codes can correct qubit loss, but not other errors, we show repeated loss and error correction using the distance-three [[9,1,3]] Bacon-Shor code. These results begin to clear a path for achieving scientific quantum advantage with a programmable neutral atom quantum processor., Comment: 17 pages, 16 figures

In recent years, with the significant evolution of multi-modal large models, many recommender researchers realized the potential of multi-modal information for user interest modeling. In industry, a wide-used modeling architecture is a cascading paradigm: (1) first pre-training a multi-modal model to provide omnipotent representations for downstream services; (2) The downstream recommendation model takes the multi-modal representation as additional input to fit real user-item behaviours. Although such paradigm achieves remarkable improvements, however, there still exist two problems that limit model performance: (1) Representation Unmatching: The pre-trained multi-modal model is always supervised by the classic NLP/CV tasks, while the recommendation models are supervised by real user-item interaction. As a result, the two fundamentally different tasks' goals were relatively separate, and there was a lack of consistent objective on their representations; (2) Representation Unlearning: The generated multi-modal representations are always stored in cache store and serve as extra fixed input of recommendation model, thus could not be updated by recommendation model gradient, further unfriendly for downstream training. Inspired by the two difficulties challenges in downstream tasks usage, we introduce a quantitative multi-modal framework to customize the specialized and trainable multi-modal information for different downstream models., Comment: Work in progress

Nuclear magnetic resonance instruments are becoming available to the do-it-yourself community. The challenges encountered in the endeavor to build a magnetic resonance imaging instrument from scratch were confronted in a four-day hackathon at Singapore University of Technology and Design in spring 2024. One day was devoted to educational lectures and three days to system construction and testing. Seventy young researchers from all parts of the world formed six teams focusing on magnet, gradient coil, RF coil, console, system integration, and design, which together produced a working MRI instrument in three days. The different steps, encountered challenges, and their solutions are reported., Comment: 49 pages, 23 figures, comments welcome (this paper is meant to be useful to people constructing their own MRI systems)

Graph Neural Networks (GNNs) have been widely employed for semi-supervised node classification tasks on graphs. However, the performance of GNNs is significantly affected by label noise, that is, a small amount of incorrectly labeled nodes can substantially misguide model training. Mainstream solutions define node classification with label noise (NCLN) as a reliable labeling task, often introducing node similarity with quadratic computational complexity to more accurately assess label reliability. To this end, in this paper, we introduce the Label Ensemble Graph Neural Network (LEGNN), a lower complexity method for robust GNNs training against label noise. LEGNN reframes NCLN as a label ensemble task, gathering informative multiple labels instead of constructing a single reliable label, avoiding high-complexity computations for reliability assessment. Specifically, LEGNN conducts a two-step process: bootstrapping neighboring contexts and robust learning with gathered multiple labels. In the former step, we apply random neighbor masks for each node and gather the predicted labels as a high-probability label set. This mitigates the impact of inaccurately labeled neighbors and diversifies the label set. In the latter step, we utilize a partial label learning based strategy to aggregate the high-probability label information for model training. Additionally, we symmetrically gather a low-probability label set to counteract potential noise from the bootstrapped high-probability label set. Extensive experiments on six datasets demonstrate that LEGNN achieves outstanding performance while ensuring efficiency. Moreover, it exhibits good scalability on dataset with over one hundred thousand nodes and one million edges.

The advancement of Multimodal Large Language Models (MLLMs) has enabled significant progress in multimodal understanding, expanding their capacity to analyze video content. However, existing evaluation benchmarks for MLLMs primarily focus on abstract video comprehension, lacking a detailed assessment of their ability to understand video compositions, the nuanced interpretation of how visual elements combine and interact within highly compiled video contexts. We introduce VidComposition, a new benchmark specifically designed to evaluate the video composition understanding capabilities of MLLMs using carefully curated compiled videos and cinematic-level annotations. VidComposition includes 982 videos with 1706 multiple-choice questions, covering various compositional aspects such as camera movement, angle, shot size, narrative structure, character actions and emotions, etc. Our comprehensive evaluation of 33 open-source and proprietary MLLMs reveals a significant performance gap between human and model capabilities. This highlights the limitations of current MLLMs in understanding complex, compiled video compositions and offers insights into areas for further improvement. The leaderboard and evaluation code are available at https://yunlong10.github.io/VidComposition/.

Context-aware methods have achieved remarkable advancements in supervised scene text recognition by leveraging semantic priors from words. Considering the heterogeneity of text and background in STR, we propose that such contextual priors can be reinterpreted as the relations between textual elements, serving as effective self-supervised labels for representation learning. However, textual relations are restricted to the finite size of the dataset due to lexical dependencies, which causes over-fitting problem, thus compromising the representation quality. To address this, our work introduces a unified framework of Relational Contrastive Learning and Masked Image Modeling for STR (RCMSTR), which explicitly models the enriched textual relations. For the RCL branch, we first introduce the relational rearrangement module to cultivate new relations on the fly. Based on this, we further conduct relational contrastive learning to model the intra- and inter-hierarchical relations for frames, sub-words and words. On the other hand, MIM can naturally boost the context information via masking, where we find that the block masking strategy is more effective for STR. For the effective integration of RCL and MIM, we also introduce a novel decoupling design aimed at mitigating the impact of masked images on contrastive learning. Additionally, to enhance the compatibility of MIM with CNNs, we propose the adoption of sparse convolutions and directly sharing the weights with dense convolutions in training. The proposed RCMSTR demonstrates superior performance in various evaluation protocols for different STR-related downstream tasks, outperforming the existing state-of-the-art self-supervised STR techniques. Ablation studies and qualitative experimental results further validate the effectiveness of our method. The code and pre-trained models will be available at https://github.com/ThunderVVV/RCMSTR ., Comment: arXiv admin note: text overlap with arXiv:2308.00508

Despite the outstanding performance of Large language models (LLMs) in diverse tasks, they are vulnerable to jailbreak attacks, wherein adversarial prompts are crafted to bypass their security mechanisms and elicit unexpected responses.Although jailbreak attacks are prevalent, the understanding of their underlying mechanisms remains limited. Recent studies have explain typical jailbreaking behavior (e.g., the degree to which the model refuses to respond) of LLMs by analyzing the representation shifts in their latent space caused by jailbreak prompts or identifying key neurons that contribute to the success of these attacks. However, these studies neither explore diverse jailbreak patterns nor provide a fine-grained explanation from the failure of circuit to the changes of representational, leaving significant gaps in uncovering the jailbreak mechanism. In this paper, we propose JailbreakLens, an interpretation framework that analyzes jailbreak mechanisms from both representation (which reveals how jailbreaks alter the model's harmfulness perception) and circuit perspectives (which uncovers the causes of these deceptions by identifying key circuits contributing to the vulnerability), tracking their evolution throughout the entire response generation process. We then conduct an in-depth evaluation of jailbreak behavior on four mainstream LLMs under seven jailbreak strategies. Our evaluation finds that jailbreak prompts amplify components that reinforce affirmative responses while suppressing those that produce refusal. Although this manipulation shifts model representations toward safe clusters to deceive the LLM, leading it to provide detailed responses instead of refusals, it still produce abnormal activation which can be caught in the circuit analysis., Comment: 18 pages, 10 figures

Recent advancements of generative AI have significantly promoted content creation and editing, where prevailing studies further extend this exciting progress to video editing. In doing so, these studies mainly transfer the inherent motion patterns from the source videos to the edited ones, where results with inferior consistency to user prompts are often observed, due to the lack of particular alignments between the delivered motions and edited contents. To address this limitation, we present a shape-consistent video editing method, namely StableV2V, in this paper. Our method decomposes the entire editing pipeline into several sequential procedures, where it edits the first video frame, then establishes an alignment between the delivered motions and user prompts, and eventually propagates the edited contents to all other frames based on such alignment. Furthermore, we curate a testing benchmark, namely DAVIS-Edit, for a comprehensive evaluation of video editing, considering various types of prompts and difficulties. Experimental results and analyses illustrate the outperforming performance, visual consistency, and inference efficiency of our method compared to existing state-of-the-art studies., Comment: Project page: https://alonzoleeeooo.github.io/StableV2V, code: https://github.com/AlonzoLeeeooo/StableV2V, model weights: https://huggingface.co/AlonzoLeeeooo/StableV2V, dataset (DAVIS-Edit): https://huggingface.co/datasets/AlonzoLeeeooo/DAVIS-Edit

We study the three-body Fermi liquid effects in the SU($N$) Anderson impurity model in the strong interaction limit where the occupation number $N_d^{}$ of the impurity levels varies over the range of $0

Euclidean representation learning methods have achieved commendable results in image fusion tasks, which can be attributed to their clear advantages in handling with linear space. However, data collected from a realistic scene usually have a non-Euclidean structure, where Euclidean metric might be limited in representing the true data relationships, degrading fusion performance. To address this issue, a novel SPD (symmetric positive definite) manifold learning framework is proposed for multi-modal image fusion, named SPDFusion, which extends the image fusion approach from the Euclidean space to the SPD manifolds. Specifically, we encode images according to the Riemannian geometry to exploit their intrinsic statistical correlations, thereby aligning with human visual perception. Actually, the SPD matrix underpins our network learning, with a cross-modal fusion strategy employed to harness modality-specific dependencies and augment complementary information. Subsequently, an attention module is designed to process the learned weight matrix, facilitating the weighting of spatial global correlation semantics via SPD matrix multiplication. Based on this, we design an end-to-end fusion network based on cross-modal manifold learning. Extensive experiments on public datasets demonstrate that our framework exhibits superior performance compared to the current state-of-the-art methods., Comment: 14 pages, 12 figures

The emergence and growing popularity of multimodal large language models (MLLMs) have significant potential to enhance various aspects of daily life, from improving communication to facilitating learning and problem-solving. Mobile phones, as essential daily companions, represent the most effective and accessible deployment platform for MLLMs, enabling seamless integration into everyday tasks. However, deploying MLLMs on mobile phones presents challenges due to limitations in memory size and computational capability, making it difficult to achieve smooth and real-time processing without extensive optimization. In this paper, we present BlueLM-V-3B, an algorithm and system co-design approach specifically tailored for the efficient deployment of MLLMs on mobile platforms. To be specific, we redesign the dynamic resolution scheme adopted by mainstream MLLMs and implement system optimization for hardware-aware deployment to optimize model inference on mobile phones. BlueLM-V-3B boasts the following key highlights: (1) Small Size: BlueLM-V-3B features a language model with 2.7B parameters and a vision encoder with 400M parameters. (2) Fast Speed: BlueLM-V-3B achieves a generation speed of 24.4 token/s on the MediaTek Dimensity 9300 processor with 4-bit LLM weight quantization. (3) Strong Performance: BlueLM-V-3B has attained the highest average score of 66.1 on the OpenCompass benchmark among models with $\leq$ 4B parameters and surpassed a series of models with much larger parameter sizes (e.g., MiniCPM-V-2.6, InternVL2-8B)., Comment: 21 pages

Dense prediction is a critical task in computer vision. However, previous methods often require extensive computational resources, which hinders their real-world application. In this paper, we propose BiDense, a generalized binary neural network (BNN) designed for efficient and accurate dense prediction tasks. BiDense incorporates two key techniques: the Distribution-adaptive Binarizer (DAB) and the Channel-adaptive Full-precision Bypass (CFB). The DAB adaptively calculates thresholds and scaling factors for binarization, effectively retaining more information within BNNs. Meanwhile, the CFB facilitates full-precision bypassing for binary convolutional layers undergoing various channel size transformations, which enhances the propagation of real-valued signals and minimizes information loss. By leveraging these techniques, BiDense preserves more real-valued information, enabling more accurate and detailed dense predictions in BNNs. Extensive experiments demonstrate that our framework achieves performance levels comparable to full-precision models while significantly reducing memory usage and computational costs.

Recently, the altermagnets, a new kind of colinear antiferromagnet with zero net magnetization and momentum-dependent spin-splitting of bands, have sparked great interest. Despite simple magnetic structures, these altermagnets exhibit intriguing and intricate dependence of AHE on the N\'eel vector, in contrast to the conventional perpendicular configuration of Hall current with magnetization in ferromagnets. However, the relationship between the AHE and the N\'eel vector remains largely elusive. Here, we propose an "extrinsic parameter" method and further reveal diverse unconventional anomalous Hall textures in the N\'eel vector space, dubbed anomalous-Hall N\'eel textures (AHNTs), for altermagnets. Notably, we find that AHNTs resemble the spin textures in momentum space, and identify 10 types across four categories of AHNTs in altermagnets. Meanwhile, we examine our key discoveries in prototypical altermagnets. Our work can offer a methodology for detecting N\'eel vectors via anomalous Hall transport, and provide useful guidelines for designing electronic and optoelectronic devices based on altermagnets.

In large-scale content recommendation systems, retrieval serves as the initial stage in the pipeline, responsible for selecting thousands of candidate items from billions of options to pass on to ranking modules. Traditionally, the dominant retrieval method has been Embedding-Based Retrieval (EBR) using a Deep Neural Network (DNN) dual-tower structure. However, applying transformer in retrieval tasks has been the focus of recent research, though real-world industrial deployment still presents significant challenges. In this paper, we introduce KuaiFormer, a novel transformer-based retrieval framework deployed in a large-scale content recommendation system. KuaiFormer fundamentally redefines the retrieval process by shifting from conventional score estimation tasks (such as click-through rate estimate) to a transformer-driven Next Action Prediction paradigm. This shift enables more effective real-time interest acquisition and multi-interest extraction, significantly enhancing retrieval performance. KuaiFormer has been successfully integrated into Kuaishou App's short-video recommendation system since May 2024, serving over 400 million daily active users and resulting in a marked increase in average daily usage time of Kuaishou users. We provide insights into both the technical and business aspects of deploying transformer in large-scale recommendation systems, addressing practical challenges encountered during industrial implementation. Our findings offer valuable guidance for engineers and researchers aiming to leverage transformer models to optimize large-scale content recommendation systems.

As the information from diverse disciplines continues to integrate during the whole life cycle of an Architecture, Engineering, and Construction (AEC) project, the BIM (Building Information Model/Modeling) becomes increasingly large. This condition will cause users difficulty in acquiring the information they truly desire on a mobile device with limited space for interaction. To improve the value of the big data of BIM, an approach to intelligent data retrieval and representation for cloud BIM applications based on natural language processing was proposed. First, strategies for data storage and query acceleration based on the popular cloud-based database were explored to handle the large amount of BIM data. Then, the concepts keyword and constraint were proposed to capture the key objects and their specifications in a natural-language-based sentence that expresses the requirements of the user. Keywords and constraints can be mapped to IFC entities or properties through the International Framework for Dictionaries (IFD). The relationship between the user's requirement and the IFC-based data model was established by path finding in a graph generated from the IFC schema, enabling data retrieval and analysis. Finally, the analyzed and summarized results of BIM data were represented based on the structure of the retrieved data. A prototype application was developed to validate the proposed approach on the data collected during the construction of the terminal of Kunming Airport, the largest single building in China. With this approach, users can significantly benefit from requesting for information and the value of BIM will be enhanced.

This paper explores the use of autonomous unmanned vehicles for supporting power grid operations. With built-in batteries and the capability to carry additional battery energy storage, the rising number of autonomous vehicles can represent a substantial amount of capacity that is currently underutilized in the power grid. Unlike traditional electric vehicles which require drivers, the operations of autonomous vehicles can be performed without human intervention. To guide idle vehicles to support power grids autonomously, we propose a tractable optimization-based method for effectively integrating these ``mobile batteries'' into grid operations. During real-time operations, the vehicles are strategically routed to target locations to help maintain system power balance and reduce operating costs. Numerical studies have confirmed both the validity and scalability of the proposed algorithm for efficiently integrating autonomous vehicles into routine power system operations.

Machine learning (ML) defenses protect against various risks to security, privacy, and fairness. Real-life models need simultaneous protection against multiple different risks which necessitates combining multiple defenses. But combining defenses with conflicting interactions in an ML model can be ineffective, incurring a significant drop in the effectiveness of one or more defenses being combined. Practitioners need a way to determine if a given combination can be effective. Experimentally identifying effective combinations can be time-consuming and expensive, particularly when multiple defenses need to be combined. We need an inexpensive, easy-to-use combination technique to identify effective combinations. Ideally, a combination technique should be (a) accurate (correctly identifies whether a combination is effective or not), (b) scalable (allows combining multiple defenses), (c) non-invasive (requires no change to the defenses being combined), and (d) general (is applicable to different types of defenses). Prior works have identified several ad-hoc techniques but none satisfy all the requirements above. We propose a principled combination technique, Def\Con, to identify effective defense combinations. Def\Con meets all requirements, achieving 90% accuracy on eight combinations explored in prior work and 81% in 30 previously unexplored combinations that we empirically evaluate in this paper.

The exclusivity principle (E-principle) states that the sum of probabilities of pairwise exclusive events cannot exceed 1. Unlike other principles proposed to characterize quantum correlations, its intrinsically non-bipartite formulation enables its application in more general contextuality scenarios. Although equivalent to the no-signalling condition for any bipartite Bell scenario, this equivalence breaks down for two independent copies of the same scenario. Such violations of the E-principle due to multiple copies, known as its activation effect, was studied in [Nat Commun 4, 2263 (2013)] for the nonlocal extremal boxes of $(2,m,2)$, $(2,2,d)$, and $(3,2,2)$ Bell scenarios. The authors mapped the problem of exhibiting activation effects to finding certain cliques inside joint exclusivity graphs. In this work, we refine the joint exclusivity structure to be an edge-colored exclusivity multigraph. This allows us to draw a novel connection to Ramsey theory, which guarantees the existence of certain monochromatic subgraphs in sufficiently large edge-colored cliques, providing a powerful tool for ruling out E-principle violations. We then exploit this connection, drawing on Ramsey-theoretic results to study violations of the E-principle by multiple copies of the contextual extremal boxes of $n$-cycle scenarios, called $n$-cycle PR boxes, for the usual ($n=4$) PR box we show that the known E-principle violation of $5/4$ is the maximal achievable using two copies, and that this same upper bound applies to two copies of the KCBS ($n = 5$) PR box. We then prove that $n \geq 6$-cycle PR boxes, unlike the extremal boxes of the aforementioned Bell scenarios, do not exhibit activation effects with two or three copies. Finally, for any number of independent copies $k$, we establish a lower bound on $n$ above which $n$-cycle PR boxes do not exhibit activation effects with $k$ copies., Comment: 17 pages, 14 figures

With the continuous development of large language models (LLMs), transformer-based models have made groundbreaking advances in numerous natural language processing (NLP) tasks, leading to the emergence of a series of agents that use LLMs as their control hub. While LLMs have achieved success in various tasks, they face numerous security and privacy threats, which become even more severe in the agent scenarios. To enhance the reliability of LLM-based applications, a range of research has emerged to assess and mitigate these risks from different perspectives. To help researchers gain a comprehensive understanding of various risks, this survey collects and analyzes the different threats faced by these agents. To address the challenges posed by previous taxonomies in handling cross-module and cross-stage threats, we propose a novel taxonomy framework based on the sources and impacts. Additionally, we identify six key features of LLM-based agents, based on which we summarize the current research progress and analyze their limitations. Subsequently, we select four representative agents as case studies to analyze the risks they may face in practical use. Finally, based on the aforementioned analyses, we propose future research directions from the perspectives of data, methodology, and policy, respectively.

Understanding complex collaboration processes is essential for the success of construction projects. However, there is still a lack of efficient methods for timely collection and analysis of collaborative networks. Therefore, an integrated framework consisting three parts, namely, system updating for data collection, data preprocessing, and social network analysis, is proposed for the twinning and mining collaborative network of a construction project. First, a system updating strategy for automatic data collection is introduced. Centrality measures are then utilized to identify key players, including hubs and brokers. Meanwhile, information sharing frequency (ISF) and association rule mining are introduced to discover collaborative patterns, that is, frequently collaborating users (FCUs) and associations between information flows and task levels. Finally, the proposed framework is validated and demonstrated in a large-scale project. The results show that key players, FCUs, and associations between information flows and task levels were successfully discovered, providing a deep understanding of collaboration and communication for decision-making processes. This research contributes to the body of knowledge by: 1) introducing ISF and Apriori-based association mining algorithm to identify FCUs and information flow patterns in collaboration; 2) establishing a new data-driven framework to map and analyze fine-grained collaborative networks automatically. It is also shown that people tend to form small groups to handle certain levels or types of tasks more efficiently. Other researchers and industrial practitioners may use this work as a foundation to further improve the efficiency of collaboration and communication.

This paper introduces OpenLS-DGF, an adaptive logic synthesis dataset generation framework, to enhance machine learning~(ML) applications within the logic synthesis process. Previous dataset generation flows were tailored for specific tasks or lacked integrated machine learning capabilities. While OpenLS-DGF supports various machine learning tasks by encapsulating the three fundamental steps of logic synthesis: Boolean representation, logic optimization, and technology mapping. It preserves the original information in both Verilog and machine-learning-friendly GraphML formats. The verilog files offer semi-customizable capabilities, enabling researchers to insert additional steps and incrementally refine the generated dataset. Furthermore, OpenLS-DGF includes an adaptive circuit engine that facilitates the final dataset management and downstream tasks. The generated OpenLS-D-v1 dataset comprises 46 combinational designs from established benchmarks, totaling over 966,000 Boolean circuits. OpenLS-D-v1 supports integrating new data features, making it more versatile for new challenges. This paper demonstrates the versatility of OpenLS-D-v1 through four distinct downstream tasks: circuit classification, circuit ranking, quality of results (QoR) prediction, and probability prediction. Each task is chosen to represent essential steps of logic synthesis, and the experimental results show the generated dataset from OpenLS-DGF achieves prominent diversity and applicability. The source code and datasets are available at https://github.com/Logic-Factory/ACE/blob/master/OpenLS-DGF/readme.md., Comment: 14 pages

To ensure secure and reliable communication in wireless systems, authenticating the identities of numerous nodes is imperative. Traditional cryptography-based authentication methods suffer from issues such as low compatibility, reliability, and high complexity. Physical-Layer Authentication (PLA) is emerging as a promising complement due to its exploitation of unique properties in wireless environments. Recently, Machine Learning (ML)-based PLA has gained attention for its intelligence, adaptability, universality, and scalability compared to non-ML approaches. However, a comprehensive overview of state-of-the-art ML-based PLA and its foundational aspects is lacking. This paper presents a comprehensive survey of characteristics and technologies that can be used in the ML-based PLA. We categorize existing ML-based PLA schemes into two main types: multi-device identification and attack detection schemes. In deep learning-based multi-device identification schemes, Deep Neural Networks are employed to train models, avoiding complex processing and expert feature transformation. Deep learning-based multi-device identification schemes are further subdivided, with schemes based on Convolutional Neural Networks being extensively researched. In ML-based attack detection schemes, receivers utilize intelligent ML techniques to set detection thresholds automatically, eliminating the need for manual calculation or knowledge of channel models. ML-based attack detection schemes are categorized into three sub-types: Supervised Learning, Unsupervised Learning, and Reinforcement Learning. Additionally, we summarize open-source datasets used for PLA, encompassing Radio Frequency fingerprints and channel fingerprints. Finally, this paper outlines future research directions to guide researchers in related fields., Comment: 111 pages, 9 figures

We demonstrate the effect of pulse shaping in momentum selective atomic Bragg diffraction. We compare temporal square pulses, which produce sidelobes in momentum space, with other shapes which can produce more nearly square momentum distributions. We produce pulses that simultaneously address two sets of velocity classes and demonstrate that we can control the differential phase imprinted on them in a way that is insensitive to laser phase fluctuations. Our work marks a significant step forward in testing Bell inequalities using massive particles entangled in momentum.