Search

Recent advancements in large language models (LLMs) have significantly enhanced their knowledge and generative capabilities, leading to a surge of interest in leveraging LLMs for high-quality data synthesis. However, synthetic data generation via prompting LLMs remains challenging due to LLMs' limited understanding of target data distributions and the complexity of prompt engineering, especially for structured formatted data. To address these issues, we introduce DiffLM, a controllable data synthesis framework based on variational autoencoder (VAE), which further (1) leverages diffusion models to reserve more information of original distribution and format structure in the learned latent distribution and (2) decouples the learning of target distribution knowledge from the LLM's generative objectives via a plug-and-play latent feature injection module. As we observed significant discrepancies between the VAE's latent representations and the real data distribution, the latent diffusion module is introduced into our framework to learn a fully expressive latent distribution. Evaluations on seven real-world datasets with structured formatted data (i.e., Tabular, Code and Tool data) demonstrate that DiffLM generates high-quality data, with performance on downstream tasks surpassing that of real data by 2-7 percent in certain cases. The data and code will be publicly available upon completion of internal review., Comment: 17 pages, 8 figures

Accurately predicting the dynamic responses of building structures under seismic loads is essential for ensuring structural safety and minimizing potential damage. This critical aspect of structural analysis allows engineers to evaluate how structures perform under various loading conditions, facilitating informed design and safety decisions. Traditional methods, which rely on complex finite element models often struggle with balancing computational efficiency and accuracy. To address this challenge, we introduce SeisGPT, a data-driven, large physics-informed model that leverages deep neural networks based on the Generative Pre-trained Transformer (GPT) architecture. SeisGPT is designed to predict, in real-time the dynamic behavior of building structures under seismic forces. Trained on a diverse corpus of seismic data and structural engineering principles, it instantly generates predictive responses, including displacement, acceleration, and inter-story drift, with high accuracy and computational efficiency. Its adaptability across various building typologies and seismic intensities makes this framework a valuable tool for designing robust structures and assessing seismic risk. Through comprehensive validation, this approach exhibits superior performance, offering engineers and researchers a powerful tool for assessing seismic response and informing resilient design strategies. This innovative framework represents a significant advancement in seismic engineering practice, with potential applications in mitigating seismic hazards and enhancing structural resilience., Comment: 23 pages, 6 figures

Magnetically induced ferroelectrics in multiferroics provide an optimal approach to pursuit intrinsically strong magnetoelectricity. However, the complex antiferromagnetism, faint magnetically induced polarization, and low working temperatures make their magnetoelectric performance incompetent from the applications demands. Here, a family of two-dimensional $5d$ halides Os$X_2$ monolayers is predicted to be ferroelectric and ferromagnetic above room temperature. More interestingly, benefiting from the strong spin-orbital coupling and high-spin state of Os$^{2+}$ ion, the magnetically induced ferroelectric polarization can reach $5.9$ $\mu$C/cm$^2$, a record-large value in type-II multiferroics. The magnetoelectric effect, that is, controlling ferroelectric polarization by magnetic field has been demonstrated, and magnetically driven ferrovalley also emerges in this system. This work provides an effective way to solve the main defects of type-II multiferroics., Comment: 7 pages, 4 figures

The wireless communication technologies have fundamentally revolutionized industrial operations. The operation of the automated equipment is conducted in a closed-loop manner, where the status of devices is collected and sent to the control center through the uplink channel, and the control center sends the calculated control commands back to the devices via downlink communication. However, existing studies neglect the interdependent relationship between uplink and downlink communications, and there is an absence of a unified approach to model the communication, sensing, and control within the loop. This can lead to inaccurate performance assessments, ultimately hindering the ability to provide guidance for the design of practical systems. Therefore, this paper introduces an integrated closed-loop model that encompasses sensing, communication, and control functionalities, while addressing the coupling effects between uplink and downlink communications. Through the analysis of system convergence, an inequality pertaining to the performances of sensing, communication, and control is derived. Additionally, a joint optimization algorithm for control and resource allocation is proposed. Simulation results are presented to offer an intuitive understanding of the impact of system parameters. The findings of this paper unveil the intricate correlation among sensing, communication, and control, providing insights for the optimal design of industrial closed-loop systems., Comment: 12 pages, 6 figures

Identifying dynamical system (DS) is a vital task in science and engineering. Traditional methods require numerous calls to the DS solver, rendering likelihood-based or least-squares inference frameworks impractical. For efficient parameter inference, two state-of-the-art techniques are the kernel method for modeling and the "one-step framework" for jointly inferring unknown parameters and hyperparameters. The kernel method is a quick and straightforward technique, but it cannot estimate solutions and their derivatives, which must strictly adhere to physical laws. We propose a model-embedded "one-step" Bayesian framework for joint inference of unknown parameters and hyperparameters by maximizing the marginal likelihood. This approach models the solution and its derivatives using Gaussian process regression (GPR), taking into account smoothness and continuity properties, and treats differential equations as constraints that can be naturally integrated into the Bayesian framework in the linear case. Additionally, we prove the convergence of the model-embedded Gaussian process regression (ME-GPR) for theoretical development. Motivated by Taylor expansion, we introduce a piecewise first-order linearization strategy to handle nonlinear dynamic systems. We derive estimates and confidence intervals, demonstrating that they exhibit low bias and good coverage properties for both simulated models and real data., Comment: 24 pages, 3 figures, 5 tables

Preference Optimization (PO), is gaining popularity as an alternative choice of Proximal Policy Optimization (PPO) for aligning Large Language Models (LLMs). Recent research on aligning LLMs iteratively with synthetic or partially synthetic data shows promising results in scaling up PO training for both academic settings and proprietary trained models such as Llama3. Despite its success, our study shows that the length exploitation issue present in PO is even more severe in Iterative Preference Optimization (IPO) due to the iterative nature of the process. In this work, we study iterative preference optimization with synthetic data. We share the findings and analysis along the way of building the iterative preference optimization pipeline. More specifically, we discuss the length exploitation issue during iterative preference optimization and propose our training objective for iterative preference optimization, namely Agreement-aware Iterative Preference Optimization (AIPO). To demonstrate the effectiveness of our method, we conduct comprehensive experiments and achieve state-of-the-art performance on MT-Bench, AlpacaEval 2.0, and Arena-Hard. Our implementation and model checkpoints will be made available at https://github.com/bytedance/AIPO.

Bug report management is a costly software maintenance process comprised of several challenging tasks. Given the UI-driven nature of mobile apps, bugs typically manifest through the UI, hence the identification of buggy UI screens and UI components (Buggy UI Localization) is important to localizing the buggy behavior and eventually fixing it. However, this task is challenging as developers must reason about bug descriptions (which are often low-quality), and the visual or code-based representations of UI screens. This paper is the first to investigate the feasibility of automating the task of Buggy UI Localization through a comprehensive study that evaluates the capabilities of one textual and two multi-modal deep learning (DL) techniques and one textual unsupervised technique. We evaluate such techniques at two levels of granularity, Buggy UI Screen and UI Component localization. Our results illustrate the individual strengths of models that make use of different representations, wherein models that incorporate visual information perform better for UI screen localization, and models that operate on textual screen information perform better for UI component localization -- highlighting the need for a localization approach that blends the benefits of both types of techniques. Furthermore, we study whether Buggy UI Localization can improve traditional buggy code localization, and find that incorporating localized buggy UIs leads to improvements of 9%-12% in Hits@10., Comment: 11 pages and 6 figures. To appear in ISSTA'24

Electric field control of magnetism in solids, i.e. the converse magnetoelectricity, is highly desired for applications of scalable energy-efficient logic devices. However, it is not only a technical challenge but also a scientific paradox, since in principle the electric and magnetic degrees of freedom obey distinct rules of symmetries. Despite the great progresses obtained in the community of multiferroics during the past decades, the success of magnetoelectricity remains on its way and more alternative approaches with conceptual revolution are urgently needed. Here, by introducing the concept of topology into multiferroics, an exotic magnetoelectric double-leaf Riemann-surface is unveiled based on the mechanism of spin-dependent $d-p$ hybridization in a two-dimensional magnet: GdI$_2$ monolayer. Protected by the topology, a $180^\circ$ spin reversal can be precisely achieved by an electric cycle, leading to a robust and dissipationless converse magnetoelectric function. Such a topological magnetoelectricity allows the nontrivial manipulation of magnetization by AC electric field. In this category, more candidate materials with better performance are designed targetedly, which pave the road to the potential applications with topological magnetoelectrics., Comment: 7 pages. 5 figures

The last decade has seen a renewed exploration of semiconductor materials for X-ray detection, foremost focusing on lead-based perovskites and other metal halides as direct-conversion materials and scintillators. However, the reported performance characteristics are often incomplete or misleading in assessing the practical utility of materials. This Perspective offers guidelines for choosing, estimating and presenting the relevant figures of merit. We also provide ready-to-used tools for calculating these figures of merit: MATLAB application, Mathcad worksheet and a website. The X-ray detectors for medical imaging are at focus for their increasing societal value and since they bring about the most stringent requirements as the image shall be acquired at as low as reasonably attainable (i.e. ALARA principle) dose received by the patient.

The ternary chalcogenide Bi2Rh3Se2 was found to be a charge density wave (CDW) superconductor with a 2*2 periodicity. The key questions regarding the underlying mechanism of CDW state and its interplay with lattice and electronic properties remains to be explored. Here, based on the systematic Raman scattering investigations on single crystalline Bi2Rh3Se2, we observed the fingerprinting feature of CDW state, a collective amplitude mode at 39 cm-1. The temperature evolution of Raman shift and line width for this amplitude mode can be well described by the critical behavior of two-dimensional (2D) Ising model, suggesting the interlayer interactions of Bi2Rh3Se2 is negligible when CDW state is formed, as a consequence, the quantum fluctuations play an important role at low temperature. Moreover, temperature dependence of Raman shift for Ag9 mode deviates significantly from the expected anharmonic behavior when approaching the CDW transition temperature 240 K, demonstrated that strong electron-phonon coupling plays a key role in the formation of CDW. Our results reveal that Bi2Rh3Se2 is an intriguing quasi-2D system to explore electronic quantum phase transition and modulate the correlations between CDW and superconductivity.

Colonoscopy videos provide richer information in polyp segmentation for rectal cancer diagnosis. However, the endoscope's fast moving and close-up observing make the current methods suffer from large spatial incoherence and continuous low-quality frames, and thus yield limited segmentation accuracy. In this context, we focus on robust video polyp segmentation by enhancing the adjacent feature consistency and rebuilding the reliable polyp representation. To achieve this goal, we in this paper propose SALI network, a hybrid of Short-term Alignment Module (SAM) and Long-term Interaction Module (LIM). The SAM learns spatial-aligned features of adjacent frames via deformable convolution and further harmonizes them to capture more stable short-term polyp representation. In case of low-quality frames, the LIM stores the historical polyp representations as a long-term memory bank, and explores the retrospective relations to interactively rebuild more reliable polyp features for the current segmentation. Combing SAM and LIM, the SALI network of video segmentation shows a great robustness to the spatial variations and low-visual cues. Benchmark on the large-scale SUNSEG verifies the superiority of SALI over the current state-of-the-arts by improving Dice by 2.1%, 2.5%, 4.1% and 1.9%, for the four test sub-sets, respectively. Codes are at https://github.com/Scatteredrain/SALI., Comment: Accepted to MICCAI 2024. Code and models: https://github.com/Scatteredrain/SALI

With the launch of ChatGPT, large language models (LLMs) have attracted global attention. In the realm of article writing, LLMs have witnessed extensive utilization, giving rise to concerns related to intellectual property protection, personal privacy, and academic integrity. In response, AI-text detection has emerged to distinguish between human and machine-generated content. However, recent research indicates that these detection systems often lack robustness and struggle to effectively differentiate perturbed texts. Currently, there is a lack of systematic evaluations regarding detection performance in real-world applications, and a comprehensive examination of perturbation techniques and detector robustness is also absent. To bridge this gap, our work simulates real-world scenarios in both informal and professional writing, exploring the out-of-the-box performance of current detectors. Additionally, we have constructed 12 black-box text perturbation methods to assess the robustness of current detection models across various perturbation granularities. Furthermore, through adversarial learning experiments, we investigate the impact of perturbation data augmentation on the robustness of AI-text detectors. We have released our code and data at https://github.com/zhouying20/ai-text-detector-evaluation., Comment: Accepted by ACL 2024, Main Conference

In many empirical settings, directly observing a treatment variable may be infeasible although an error-prone surrogate measurement of the latter will often be available. Causal inference based solely on the surrogate measurement is particularly challenging without validation data. We propose a method that obviates the need for validation data by carefully incorporating the surrogate measurement with a proxy of the hidden treatment to obtain nonparametric identification of several causal effects of interest, including the population average treatment effect, the effect of treatment on the treated, quantile treatment effects, and causal effects under marginal structural models. For inference, we provide general semiparametric theory for causal effects identified using our approach and derive a large class of semiparametric efficient estimators with an appealing multiple robustness property. A significant obstacle to our approach is the estimation of nuisance functions which involve the hidden treatment therefore preventing the direct use of standard machine learning algorithms, which we resolve by introducing a novel semiparametric EM algorithm. We examine the finite-sample performance of our method using simulations and an application which aims to estimate the causal effect of Alzheimer's disease on hippocampal volume using data from the Alzheimer's Disease Neuroimaging Initiative.

Photometric constraint is indispensable for self-supervised monocular depth estimation. It involves warping a source image onto a target view using estimated depth&pose, and then minimizing the difference between the warped and target images. However, the endoscopic built-in light causes significant brightness fluctuations, and thus makes the photometric constraint unreliable. Previous efforts only mitigate this relying on extra models to calibrate image brightness. In this paper, we propose MonoPCC to address the brightness inconsistency radically by reshaping the photometric constraint into a cycle form. Instead of only warping the source image, MonoPCC constructs a closed loop consisting of two opposite forward-backward warping paths: from target to source and then back to target. Thus, the target image finally receives an image cycle-warped from itself, which naturally makes the constraint invariant to brightness changes. Moreover, MonoPCC transplants the source image's phase-frequency into the intermediate warped image to avoid structure lost, and also stabilizes the training via an exponential moving average (EMA) strategy to avoid frequent changes in the forward warping. The comprehensive and extensive experimental results on four endoscopic datasets demonstrate that our proposed MonoPCC shows a great robustness to the brightness inconsistency, and exceeds other state-of-the-arts by reducing the absolute relative error by at least 7.27%, 9.38%, 9.90% and 3.17%, respectively., Comment: 11 pages, 10 figures

With the development of large language models (LLMs), detecting whether text is generated by a machine becomes increasingly challenging in the face of malicious use cases like the spread of false information, protection of intellectual property, and prevention of academic plagiarism. While well-trained text detectors have demonstrated promising performance on unseen test data, recent research suggests that these detectors have vulnerabilities when dealing with adversarial attacks such as paraphrasing. In this paper, we propose a framework for a broader class of adversarial attacks, designed to perform minor perturbations in machine-generated content to evade detection. We consider two attack settings: white-box and black-box, and employ adversarial learning in dynamic scenarios to assess the potential enhancement of the current detection model's robustness against such attacks. The empirical results reveal that the current detection models can be compromised in as little as 10 seconds, leading to the misclassification of machine-generated text as human-written content. Furthermore, we explore the prospect of improving the model's robustness over iterative adversarial learning. Although some improvements in model robustness are observed, practical applications still face significant challenges. These findings shed light on the future development of AI-text detectors, emphasizing the need for more accurate and robust detection methods., Comment: Accepted by COLING 2024

We propose MonoBox, an innovative box-supervised segmentation method constrained by monotonicity to liberate its training from the user-unfriendly box-tightness assumption. In contrast to conventional box-supervised segmentation, where the box edges must precisely touch the target boundaries, MonoBox leverages imprecisely-annotated boxes to achieve robust pixel-wise segmentation. The 'linchpin' is that, within the noisy zones around box edges, MonoBox discards the traditional misguiding multiple-instance learning loss, and instead optimizes a carefully-designed objective, termed monotonicity constraint. Along directions transitioning from the foreground to background, this new constraint steers responses to adhere to a trend of monotonically decreasing values. Consequently, the originally unreliable learning within the noisy zones is transformed into a correct and effective monotonicity optimization. Moreover, an adaptive label correction is introduced, enabling MonoBox to enhance the tightness of box annotations using predicted masks from the previous epoch and dynamically shrink the noisy zones as training progresses. We verify MonoBox in the box-supervised segmentation task of polyps, where satisfying box-tightness is challenging due to the vague boundaries between the polyp and normal tissues. Experiments on both public synthetic and in-house real noisy datasets demonstrate that MonoBox exceeds other anti-noise state-of-the-arts by improving Dice by at least 5.5% and 3.3%, respectively. Codes are at https://github.com/Huster-Hq/MonoBox.

Strategic Traffic Management Initiatives (TMIs) such as Ground Delay Programs (GDPs) play a crucial role in mitigating operational costs associated with demand-capacity imbalances. However, GDPs can only be planned (e.g., duration, delay assignments) with confidence if the future capacities at constrained resources (i.e., airports) are predictable. In reality, such future capacities are uncertain, and predictive models may provide forecasts that are vulnerable to errors and distribution shifts. Motivated by the goal of planning optimal GDPs that are \emph{distributionally robust} against airport capacity prediction errors, we study a fully integrated learning-driven optimization framework. We design a deep learning-based prediction model capable of forecasting arrival and departure capacity distributions across a network of airports. We then integrate the forecasts into a distributionally robust formulation of the multi-airport ground holding problem (\textsc{dr-MAGHP}). We show how \textsc{dr-MAGHP} can outperform stochastic optimization when distribution shifts occur, and conclude with future research directions to improve both the learning and optimization stages., Comment: 8 pages, 6 figures

Large language models (LLMs) have exhibited remarkable capabilities in text generation tasks. However, the utilization of these models carries inherent risks, including but not limited to plagiarism, the dissemination of fake news, and issues in educational exercises. Although several detectors have been proposed to address these concerns, their effectiveness against adversarial perturbations, specifically in the context of student essay writing, remains largely unexplored. This paper aims to bridge this gap by constructing AIG-ASAP, an AI-generated student essay dataset, employing a range of text perturbation methods that are expected to generate high-quality essays while evading detection. Through empirical experiments, we assess the performance of current AIGC detectors on the AIG-ASAP dataset. The results reveal that the existing detectors can be easily circumvented using straightforward automatic adversarial attacks. Specifically, we explore word substitution and sentence substitution perturbation methods that effectively evade detection while maintaining the quality of the generated essays. This highlights the urgent need for more accurate and robust methods to detect AI-generated student essays in the education domain., Comment: Accepted by EMNLP 2023 Main conference, Oral Presentation

Multimodal learning has exhibited a significant advantage in affective analysis tasks owing to the comprehensive information of various modalities, particularly the complementary information. Thus, many emerging studies focus on disentangling the modality-invariant and modality-specific representations from input data and then fusing them for prediction. However, our study shows that modality-specific representations may contain information that is irrelevant or conflicting with the tasks, which downgrades the effectiveness of learned multimodal representations. We revisit the disentanglement issue, and propose a novel triple disentanglement approach, TriDiRA, which disentangles the modality-invariant, effective modality-specific and ineffective modality-specific representations from input data. By fusing only the modality-invariant and effective modality-specific representations, TriDiRA can significantly alleviate the impact of irrelevant and conflicting information across modalities during model training. Extensive experiments conducted on four benchmark datasets demonstrate the effectiveness and generalization of our triple disentanglement, which outperforms SOTA methods., Comment: 14 pages, 6 figures

Quantum materials whose atoms are arranged on a lattice of corner-sharing triangles, $\textit{i.e.}$, the kagome lattice, have recently emerged as a captivating platform for investigating exotic correlated and topological electronic phenomena. Here, we combine ultra-low temperature angle-resolved photoemission spectroscopy (ARPES) with scanning tunneling microscopy and density functional theory calculations to reveal the fascinating electronic structure of the bilayer-distorted kagome material $\textit{Ln}$Ti${_3}$Bi${_4}$, where $\textit{Ln}$ stands for Nd and Yb. Distinct from other kagome materials, $\textit{Ln}$Ti${_3}$Bi${_4}$ exhibits two-fold, rather than six-fold, symmetries, stemming from the distorted kagome lattice, which leads to a unique electronic structure. Combining experiment and theory we map out the electronic structure and discover double flat bands as well as multiple van Hove singularities (VHSs), with one VHS exhibiting higher-order characteristics near the Fermi level. Notably, in the magnetic version NdTi${_3}$Bi${_4}$, the ultra-low base temperature ARPES measurements unveil an unconventional band splitting in the band dispersions which is induced by the ferromagnetic ordering. These findings reveal the potential of bilayer-distorted kagome metals $\textit{Ln}$Ti${_3}$Bi${_4}$ as a promising platform for exploring novel emergent phases of matter at the intersection of strong correlation and magnetism.

Anomalous Hall effect (AHE), one of the most important electronic transport phenomena, generally appears in ferromagnetic materials but is rare in materials without magnetic elements. Here, we present a study of La$_3$MgBi$_5$, whose band structure carries multitype Dirac fermions. Although magnetic elements are absent in La$_3$MgBi$_5$, clear signals of AHE can be observed. In particular, the anomalous Hall conductivity is extremely large, reaching 42,356 $\Omega^{-1}$ cm$^{-1}$ with an anomalous Hall angle of 8.8 %, the largest one that has been observed in the current AHE systems. The AHE is suggested to originate from the combination of skew scattering and Berry curvature. Another unique property discovered in La$_3$MgBi$_5$ is the axial diamagnetism. The diamagnetism is significantly enhanced and dominates the magnetization in the axial directions, which is the result of restricted motion of the Dirac fermion at Fermi level. Our findings not only establish La$_3$MgBi$_5$ as a suitable platform to study AHE and quantum transport, but also indicate the great potential of 315-type Bi-based materials for exploring novel physical properties.